Fact-checked by Grok 2 weeks ago

Central processing unit

A central processing unit (CPU), also known as a or , is the primary electronic circuitry in a computer system responsible for executing instructions from programs by performing arithmetic, logical, control, and operations. Often referred to as the "" of the computer, the CPU interprets and processes from , enabling all computational tasks such as calculations, data comparisons, and task coordination. In modern systems, CPUs are typically implemented as integrated circuits on a single chip, containing billions of transistors that allow for high-speed operation measured in gigahertz (GHz). The core architecture of a CPU consists of three main components: the arithmetic logic unit (ALU), the control unit (CU), and registers. The ALU performs basic arithmetic operations like addition and subtraction, as well as logical operations such as comparisons (e.g., greater-than or equal-to). The control unit directs the flow of data and instructions by fetching them from , decoding them, and coordinating execution with other hardware components. Registers serve as high-speed temporary storage locations within the CPU for holding instructions, addresses, and intermediate results during processing. This , which separates processing from , underpins most general-purpose computers today. The concept of a CPU traces its origins to 19th-century mechanical designs, such as Charles Babbage's , which proposed a "mill" for computation controlled by punched cards and supported by a store. Theoretical foundations were advanced by Alan Turing's 1936 , which formalized the stored-program concept essential to modern CPUs. The first electronic stored-program computer, the Manchester "Baby" in 1948, demonstrated practical implementation, evolving from vacuum tube-based systems to integrated circuits. A pivotal milestone occurred in 1971 with the , the first commercially available single-chip featuring 2,300 s and capable of 92,000 instructions per second. Subsequent generations introduced multi-core designs, 64-bit processing, and increased densities, enabling contemporary CPUs to handle complex multitasking in devices from smartphones to supercomputers.

Overview

Definition and function

A central processing unit (CPU), often simply called a , is the primary electronic circuitry in a computer that executes instructions from programs by performing the fundamental fetch-decode-execute cycle, thereby transforming input data into output results. This process enables the CPU to serve as the "brain" of the system, managing computations and coordinating overall operations. The CPU's core functions encompass arithmetic operations, such as addition, subtraction, multiplication, and division, handled by dedicated hardware; logical decisions, including comparisons, AND/OR/NOT operations, and bitwise manipulations; data movement, which involves transferring information between internal registers, , and ; and management, such as jumping to conditional branches or looping through sequences of instructions to direct program execution. These functions collectively allow the CPU to process complex tasks efficiently, from simple calculations to orchestrating multitasking environments. The CPU interacts closely with system to fetch instructions and operands, relying on high-speed caches and registers for quick access, while communicating with (I/O) devices—such as keyboards, displays, and storage drives—through standardized buses that transmit address signals (to locate ), signals (to carry information), and signals (to synchronize operations). This bus-mediated ensures seamless flow across the computer system, preventing bottlenecks in instruction processing. The term "central processing unit" and its underlying concept trace their origins to the , first articulated in von Neumann's 1945 "First Draft of a Report on the ," which proposed a stored-program design featuring a for and functions separate from and I/O. In a von Neumann machine, the CPU occupies a central position, linked bidirectionally to main for program and data storage, and to peripheral I/O devices, all interconnected via shared buses as depicted in this simplified :
+----------------+       Buses (Address, [Data](/page/Data), [Control](/page/Control))       +----------------+
|                | <---------------------------------------> |                |
|     CPU        |                                           |   Main Memory  |
|                |                                           | ([RAM](/page/Ram)/[Storage](/page/Storage))  |
+----------------+                                           +----------------+
         ^                                                           ^
      +----------------+                                        +----------------+
      |     I/O        |                                        |   Secondary    |
      |   Devices      |                                        |   Storage      |
      | (e.g., [Keyboard](/page/Keyboard),|                                        |   (e.g., Disk) |
      |   [Display](/page/Display))     |                                        +----------------+
      +----------------+
This architecture underscores the CPU's role in sequentially processing instructions while accessing shared resources.

Types and classifications

Central processing units (CPUs) are classified by their (ISA), which defines the set of instructions available to programmers and influences performance, power efficiency, and complexity. Complex Instruction Set Computing (CISC) architectures, such as x86 used in many desktop and server processors, feature a large number of instructions that can perform multiple operations in a single command, allowing for more compact code but requiring more complex hardware to decode and execute them. In contrast, Reduced Instruction Set Computing (RISC) architectures, exemplified by in mobile and systems, employ a smaller set of simpler, fixed-length instructions that execute in fewer clock cycles, promoting pipelining efficiency and lower power consumption at the cost of potentially longer code sequences. This dichotomy emerged in the to balance software density against hardware simplicity. CPUs are further categorized by their level of , reflecting how elements are packaged for different applications. Discrete CPUs consist of standalone chips that with separate and peripheral components via external buses, commonly found in early computers and high-performance servers for modularity and upgradability. integrate the full CPU functionality onto a single , enabling compact designs since the 1970s and dominating general-purpose computing. processors adapt microprocessor designs for resource-limited environments like appliances and automotive systems, prioritizing low power and real-time performance over raw speed. System-on-chip () variants extend this by combining the CPU core with , interfaces, and accelerators on one die, optimizing for mobile devices and where space and efficiency are critical. Special-purpose classifications address CPUs tailored for domain-specific tasks beyond general computing. Digital signal processors (DSPs) incorporate hardware optimized for mathematical operations on signals, such as multiply-accumulate instructions for audio and image processing, achieving higher throughput than general CPUs for repetitive numerical workloads. Graphics processing units (GPUs), often serving as co-processors, feature thousands of simpler cores for parallel computations, excelling in vectorized tasks like rendering and but relying on a host CPU for orchestration. Application-specific integrated circuits () customize CPU-like logic for singular functions, such as cryptocurrency mining, offering superior efficiency and speed for fixed algorithms at the expense of flexibility. These variants complement general-purpose CPUs in heterogeneous systems to handle specialized workloads. The evolution from single-core to multi-core and many-core designs has transformed CPU scalability to exploit parallelism amid from clock speed increases. Single-core processors dominated until the mid-2000s, limited by power walls and heat dissipation, prompting the shift to multi-core architectures where multiple processing units share resources to boost throughput for threaded applications. Homogeneous multi-core systems employ identical cores for uniform task distribution, as in most consumer CPUs, ensuring balanced performance but underutilizing specialized needs. Heterogeneous designs integrate diverse cores—such as high-performance and energy-efficient ones—on the same chip, adapting dynamically to workloads like for better overall efficiency. Many-core processors, with dozens or hundreds of cores, extend this for data-center and applications, emphasizing massive parallelism over per-core complexity. Classification metrics often revolve around ISA levels and microarchitecture, providing a framework to evaluate design trade-offs. The ISA level specifies the abstract interface, including opcodes, registers, and addressing modes, which remains stable across implementations to ensure software compatibility. Microarchitecture, the internal realization of the ISA, varies implementations like out-of-order execution or branch prediction to optimize for speed, power, or area, allowing the same ISA (e.g., x86) to support diverse hardware generations. These layers enable abstraction, where ISA defines "what" instructions do, while microarchitecture handles "how" they execute efficiently.

Historical development

Early computational devices

The development of early computational devices laid the foundational concepts for modern central processing units (CPUs) by introducing mechanical, electromechanical, and early electronic mechanisms for performing calculations and controlling operations. In 1837, conceived the , a mechanical general-purpose computer designed to execute programmable instructions using punched cards for input and control. This device featured a "mill" analogous to an (ALU) for performing operations and a "store" for holding data and instructions, representing an early conceptual stored-program system, though it was never fully built due to technological limitations of the era. Advancing to electromechanical designs, completed the Z3 in 1941, recognized as the first functional, freely programmable digital computer based on logic and using approximately 2,300 electromechanical relays for switching operations. The Z3 employed arithmetic to perform floating-point calculations and was programmed via punched film tape, enabling it to solve complex engineering problems like aerodynamic simulations, but its relay-based architecture limited speed and reliability compared to later electronic systems. The vacuum tube era marked a shift to electronic computing, exemplified by the ENIAC (Electronic Numerical Integrator and Computer), completed in 1945 as the first general-purpose electronic digital computer, utilizing 18,000 vacuum tubes for high-speed arithmetic and logic functions. Programming ENIAC required manual reconfiguration through plugboards and switches rather than stored instructions, allowing it to perform 5,000 additions per second but occupying 1,800 square feet, consuming 150 kilowatts of power, and suffering frequent tube failures that reduced reliability. Building on these efforts, the EDSAC (Electronic Delay Storage Automatic Calculator) ran its first program in May 1949 at the University of Cambridge, introducing practical stored-program execution where instructions and data resided in the same mercury delay-line memory, facilitated by an initial orders assembler for symbolic programming. A pivotal concept emerged in John von Neumann's 1945 report on the , articulating the stored-program principle where programs and data are stored interchangeably in high-speed memory using representation, enabling flexible computation and . arithmetic, as implemented in devices like the Z3 and , provided an efficient basis for digital logic by representing numbers with two states (0 and 1), contrasting earlier decimal systems. These early machines, however, were hampered by enormous size, excessive power demands, heat generation, and low reliability—such as vacuum tubes burning out every few hours—necessitating the eventual transition to more efficient technology.

Transistor and integrated circuit eras

The invention of the in 1947 at Bell Laboratories marked a pivotal shift from vacuum tubes to semiconductor-based electronics in computing. On December 16, 1947, physicists and Walter Brattain demonstrated the first using , achieving current amplification without the fragility and heat issues of vacuum tubes. , their colleague, refined this into the more practical junction transistor by early 1948, enabling the creation of smaller, more reliable logic gates essential for digital circuits. This breakthrough dramatically reduced the size, power consumption, and failure rates of computational components, paving the way for transistorized computers that outperformed their vacuum-tube predecessors in reliability and efficiency. The first transistorized computers emerged in the mid-1950s, replacing discrete vacuum tubes with individual s for logic functions. In 1954, Bell Laboratories completed (Transistor Digital Computer) under a U.S. contract, becoming the world's first fully transistorized computer with approximately 700 point-contact transistors and over 10,000 diodes handling airborne tasks. This system demonstrated transistors' viability for applications, operating at speeds up to 1 MHz while consuming far less than tube-based machines. By 1959, introduced the 7090, a large-scale scientific computer using approximately 50,000 discrete transistors for its core logic, marking one of the earliest commercial successes of transistor technology in . These machines highlighted transistors' ability to enable faster clock rates—such as 1 MHz for and approximately 0.46 MHz for the 7090—and modular designs, influencing subsequent CPU architectures. A major advancement came with the (IC), which combined multiple and components on a single , further miniaturizing CPU designs. In September 1958, at fabricated the first IC prototype—a germanium integrating a , resistors, and capacitors—demonstrating monolithic construction that eliminated discrete wiring and reduced costs. In 1959, at developed the planar process, using passivation and to create stable, mass-producible ICs on a flat surface, which improved yield and scalability. These innovations slashed component counts, enabling denser logic integration and lower manufacturing expenses, fundamentally transforming CPU fabrication from hand-wired assemblies to automated processes. The 1960s saw the rise of small-scale integration (SSI), where packed 10 to 100 transistors per chip, allowing CPUs to incorporate multiple logic functions without relying on discrete components. This era's computers, such as the 1966 Apollo Guidance Computer (AGC) developed by and for , utilized around 5,600 SSI to achieve compact, radiation-hardened processing for navigation. The AGC's design, with its 16-bit word length and 2.048 MHz clock, exemplified SSI's impact by fitting essential and control logic into a 70-pound unit, far smaller than equivalent vacuum-tube systems. By the 1970s, medium-scale integration (MSI) with 100 to 1,000 transistors per chip and large-scale integration (LSI) exceeding 1,000 transistors enabled even greater density, supporting complex CPU subsystems on fewer chips. MSI/LSI chips facilitated preliminary CPU designs, such as custom logic arrays for calculators that prefigured the Intel 4004, by integrating registers, ALUs, and control units to boost performance while cutting assembly costs. These advancements increased transistor counts to thousands per system, allowing clock speeds to reach several MHz and laying the groundwork for fully integrated processors.

Microprocessor introduction

The advent of the marked a pivotal shift in by integrating the central processing unit (CPU) onto a single chip, enabling compact, affordable, and versatile computational devices. The first commercially available , the , was developed in 1971 by a team led by Ted Hoff and at Intel Corporation, initially designed for the 141-PF calculator. This 4-bit processor contained approximately 2,300 transistors and operated at a clock speed of 740 kHz, performing basic arithmetic and logic operations while interfacing with custom chips for a specific application. By consolidating CPU functions into one , the 4004 reduced manufacturing complexity and costs compared to earlier multi-chip designs, laying the groundwork for programmable general-purpose . Subsequent developments rapidly expanded microprocessor capabilities and applications. In 1972, Intel introduced the 8008, the world's first 8-bit programmable with around 3,500 transistors, targeted at embedded control systems like the Computer Terminal Corporation's Datapoint 2200. This was followed by the more powerful in 1974, featuring improved instruction sets and higher performance, which powered the MITS —the first commercially successful kit released in 1975 and instrumental in sparking the home computing revolution. Advancing to 16-bit architectures, the debuted in 1978 as the foundation of the x86 , enabling broader memory addressing and compatibility that persists in modern processors. Concurrently, Motorola's 68000, introduced in 1979, offered a 16/32-bit design with 68,000 transistors and was selected for Apple's Macintosh computer in 1984, influencing and systems. These innovations profoundly democratized by transitioning from bespoke, expensive custom logic circuits to standardized, programmable instruction set architectures (ISAs), drastically lowering costs and enabling widespread adoption in personal computers, embedded devices, and . The microprocessor's rise facilitated the personal boom of the and , making powerful accessible beyond mainframes and minicomputers to hobbyists, businesses, and eventually households. This accessibility was amplified by , Gordon E. Moore's 1965 observation that the number of transistors on an would roughly double every year (later revised to every two years), driving exponential improvements in performance and density that sustained the proliferation of microprocessors. Overall, microprocessors transformed from a specialized enterprise tool into a ubiquitous technology, powering diverse applications from desktop PCs to microcontrollers in appliances.

Instruction processing

Fetch phase

The fetch phase initiates the instruction cycle in a central processing unit (CPU) by retrieving the next from , setting the stage for subsequent processing stages. This phase assumes that a has been loaded into and relies on the (PC), a dedicated that stores the of the forthcoming . The PC, as one of the CPU's essential s, ensures sequential execution by pointing to the correct location in . In the fetch process, the CPU transmits the address held in the PC over the address bus to the system, while simultaneously sending a read signal via the to request the . The targeted —typically the in modern designs or main if the cache misses—responds by placing the onto the data bus, allowing the CPU to load it into the (). Following the transfer, the PC increments by the length of the fetched , usually , to prepare for the next cycle. This interaction with the minimizes through cache hierarchies, where faster on-chip s serve most fetches to avoid slower main access. To enhance efficiency in pipelined processors, the fetch phase often incorporates instruction prefetching, where the CPU anticipates and retrieves multiple sequential instructions ahead of time, buffering them to sustain pipeline flow and mask . Data integrity during fetching is maintained via mechanisms, such as bits for basic checks or error-correcting codes () that identify and repair single-bit errors in retrieved instructions from or . In pipelined systems, ECC processing may introduce minor delays, but it ensures reliable operation by correcting errors on-the-fly without halting the fetch entirely. As an illustrative example in a basic , the fetch phase loads the complete —including the and any immediate operands—directly into the IR from a unified space shared with data. The output of this phase, the raw in the IR, directly feeds into the subsequent decode phase for interpretation.

Decode phase

In the decode phase of the CPU's processing cycle, the retrieves the from the (IR) and analyzes its binary encoding to determine the intended operation and required resources. The portion of the , typically the leading bits, is decoded to identify the specific command, such as addition, branching, or load/store, enabling the generation of control signals that configure the processor's and activate relevant hardware units like the (ALU) or registers. This process ensures the CPU understands the semantics of the before proceeding to execution. Operand handling occurs concurrently during decoding, where the CPU parses the remaining bits of the to classify operands as immediate values (constants embedded directly in the ), register identifiers (specifying sources or destinations in the register file), or memory references (requiring address computation). For memory-based operands, the decode stage calculates the effective address using specified addressing modes, such as register indirect or , by combining base registers, offsets, and immediates as dictated by the format. Register operands are read from the register file in this stage to prepare data for subsequent operations, while immediate values are sign-extended if necessary to match the processor's word size. In Complex Instruction Set Computing (CISC) architectures, such as x86, the decode phase often employs to handle the complexity of variable-length, multifaceted instructions. The indexes into a microcode control store, which translates the macro-instruction into a sequence of primitive micro-operations (micro-ops) that simpler hardware can execute, effectively breaking down operations like string manipulation or into ALU accesses, shifts, and conditional branches. This microcode-driven approach allows CISC processors to support a rich set without exponentially increasing in the decoder. Pipelined CPUs integrate detection into the decode stage to maintain throughput by identifying potential s early. hazards, particularly read-after-write () dependencies where an requires results from a prior uncompleted , are detected by comparing source specifiers in the current with destination registers in stages ahead. Upon detection, the controller may the fetch and decode stages by inserting bubbles (no-operation cycles) or enable forwarding from later registers to the without delay. hazards from es are also evaluated here, with the computed and logic consulted to decide whether to continue sequential fetching. The decode phase culminates in the emission of control signals that set up the execute stage, directing data routing between registers, the ALU, and interfaces while specifying operation types (e.g., add, compare) and write-back destinations. These signals, generated by hardwired logic in Reduced Instruction Set Computing (RISC) designs or via in CISC, ensure precise orchestration of resources without performing any or access. The unit's role in producing these signals is fundamental to seamless instruction flow.

Execute phase

In the execute phase, the CPU performs the operation indicated by the decoded instruction, utilizing its functional units to process data and update system state. The arithmetic logic unit (ALU) executes arithmetic operations, such as addition or subtraction, and logical operations, such as AND, OR, or bitwise shifts, on operands sourced from registers. The results of these computations are stored in designated registers or written to memory locations, completing the data processing for the instruction. For instructions like , the execute phase evaluates status flags—such as zero (Z), carry (C), (V), and negative (N)—generated from prior ALU operations to determine the next execution path. If a is satisfied, the (PC) is loaded with the target address; otherwise, the PC increments by the length, typically 4 bytes in 32-bit architectures, to advance to the sequential next . Input/output (I/O) instructions during this phase facilitate data transfer between CPU registers and peripheral devices, often through port-mapped I/O or memory-mapped I/O mechanisms, enabling communication with external hardware like keyboards or displays. At the conclusion of the execute phase, the PC is finalized to address the subsequent instruction, and updated status flags are stored in the flags register to influence future conditional operations. For instance, in an ADD adding the contents of two registers (e.g., R1 and ), the ALU computes the sum, writes it to the destination (e.g., R0), and sets flags like the if the result equals zero or the if an occurs during .

Core components

Control unit

The (CU) is a fundamental component of the central processing unit (CPU) that directs the processor's operations by generating timing and control signals to coordinate the fetch-decode-execute cycle. It interprets instructions fetched from , decodes their requirements, and orchestrates the necessary actions across the CPU's subsystems to ensure orderly execution of . By acting as the processor's , the CU manages data flow between components without performing computations itself, enabling the CPU to complex sequences of operations efficiently. Control units are designed in two main architectures: hardwired and microprogrammed. Hardwired control employs circuits and finite state machines to produce control signals directly from the instruction opcode and current state, providing rapid response times since signals are generated without accesses, though modifications require hardware redesigns. Microprogrammed control, conversely, stores sequences of microinstructions in a dedicated control (typically ), where a sequencer fetches and executes these microinstructions to generate signals; this method offers flexibility for implementing intricate instruction sets and facilitating or updates via , albeit with performance overhead from the microinstruction fetch cycle. Key functions of the include , handling, and sequencing logic. It leverages the CPU's system clock to produce synchronized timing pulses, ensuring that fetch, decode, and execute phases occur in across components during each machine cycle. For handling, the CU detects incoming requests from peripherals or software, suspends the current stream by saving in registers, and redirects execution to a handler routine before resuming normal operation upon completion. Sequencing logic, implemented via machines in hardwired designs or sequencers in microprogrammed ones, determines the progression of control signals based on type, processor flags, and prior outcomes, maintaining the integrity of the pipeline. The evolution of control units reflects advancements in processor complexity, progressing from rudimentary state machines in early architectures that handled basic sequential instructions to sophisticated schedulers in modern superscalar designs. These advanced units manage dynamic instruction dispatching and resource arbitration to support and , significantly enhancing throughput in environments. The interacts seamlessly with other CPU elements by issuing enable signals to the ALU for operation selection, controlling access for data loading and storage, and driving and control buses for external and I/O coordination. These signals are essential in the decode phase to interpret opcodes and route resources accordingly.

Arithmetic logic unit

The arithmetic logic unit (ALU) is the computational core of a central processing unit (CPU) responsible for performing arithmetic and logical operations on data. It processes binary inputs from registers or memory, executing functions such as addition, subtraction, and bitwise manipulations to produce results that are stored back or used for further computations. The ALU receives two operands, typically denoted as A and B, and applies a selected operation (op) to generate an output result, expressed fundamentally as \text{result} = A \oplus B, where \oplus represents the decoded function. This unit forms the basis for all numerical and logical processing in the CPU, enabling everything from simple calculations to complex algorithm execution. Key components of the ALU include adders for arithmetic summation, shifters for bit manipulation, and logic gates such as AND, OR, and XOR for bitwise operations. Adders, often implemented as full adders, handle carry propagation across bits, while shifters perform left or right shifts to multiply or divide by powers of two. A flag register, or status register, accompanies these components to store condition flags like zero (indicating a result of all zeros), carry (from the most significant bit), overflow (for signed arithmetic exceeding representable range), and sign (the most significant bit of the result). These flags support conditional branching and error detection in program flow. The ALU supports a range of operations, including integer arithmetic such as addition, subtraction, multiplication, and division, as well as bitwise logic (AND, OR, XOR) and comparisons (equality, greater than, less than). Subtraction is typically implemented by inverting one operand and adding one, using two's complement representation for signed integers. Comparisons generate flags without storing the full result, often by performing subtraction and checking the zero or sign flags. Floating-point operations, requiring higher precision and specialized handling, are generally delegated to a separate floating-point unit (FPU), though some designs integrate basic extensions into the ALU. ALU design primarily employs for instantaneous operation selection via multiplexers, contrasting with used in pipelined or state-dependent extensions. To accelerate , carry-lookahead adders (CLAs) precompute carry bits using generate (G_i = A_i B_i) and propagate (P_i = A_i \oplus B_i) signals, allowing parallel carry generation across bits rather than sequential ripple propagation. This reduces delay in wide ALUs, where the carry-out for bit i is C_i = G_i + P_i C_{i-1}. Data width varies from 8-bit in early microprocessors to 64-bit or more in modern CPUs, matching the processor's word size for efficient handling of large integers; for example, 64-bit ALUs support operations on operands up to $2^{64} - 1 in unsigned form. Control signals from the CPU's dictate the operation, ensuring the ALU executes the appropriate function on demand.

Registers and addressing units

The register file serves as a small, high-speed array of storage locations integrated into the central processing unit (CPU), designed to hold operands and intermediate results during instruction execution. It consists primarily of general-purpose registers (GPRs), which are versatile storage units accessible by most instructions for data manipulation, and special-purpose registers dedicated to system functions. In the ARMv8-A architecture's AArch64 execution state, there are 31 GPRs (X0 through X30), each 64 bits wide, providing flexible temporary storage for operands in data-processing operations. Similarly, the x86-64 architecture defines 16 GPRs (RAX, RBX, RCX, RDX, RSI, RDI, RBP, RSP, and R8 through R15), each also 64 bits, enabling efficient handling of integer and pointer operations in 64-bit environments. Special registers include the program counter (PC), which stores the memory address of the next instruction to fetch; the stack pointer (SP), which maintains the address of the top of the call stack for managing subroutine calls, returns, and local variables; and the instruction register (IR), a temporary holding area for the currently fetched and decoded instruction within the pipeline. In ARMv8-A, the SP is a dedicated 64-bit register (often aliased with X31 in certain contexts), while the PC operates as a special register not directly writable like GPRs, and the IR is internal to the decode stage. Addressing modes determine how instructions specify the location of , enabling flexible to in , immediate values, or without requiring separate address-calculation . Common modes include immediate addressing, where the operand value is embedded directly in the word; direct addressing, using an absolute specified in the ; indirect addressing, where a holds the of the operand; and indexed addressing, which adds an offset (immediate or from another ) to a base to compute the effective address. These computations are performed using dedicated circuits within the CPU, allowing efficient operand location during the decode and execute phases; for instance, in architectures, load/store support offset, pre-indexed, and post-indexed modes that leverage -based indexing for array or operations. The (AGU) is specialized hardware that accelerates these calculations by computing effective addresses in parallel, often as a separate stage or port to reduce latency in . In x86 processors, the AGU handles complex addressing like base-plus-index-plus-scale for vectorized , integrating seamlessly with load/store units to minimize dependencies on the . Within the CPU pipeline, the plays a critical role in managing data flow and resolving hazards, particularly through techniques like to eliminate false dependencies. dynamically maps architectural s (visible to software) to a larger pool of physical s, preventing write-after-read (WAR) and write-after-write (WAW) hazards by allowing without stalling for register availability. This approach, common in superscalar processors, expands the effective register capacity beyond the architectural limit—such as providing 128 or more physical registers in modern designs to support dozens of in-flight s—while the core remains compact. The PC , for example, increments automatically after each fetch to point to the subsequent address. Overall, the offers limited on-chip storage, typically 128 to 512 bytes for GPRs and specials in baseline configurations (e.g., 16 × 64-bit GPRs in yield 128 bytes), in stark contrast to the gigabytes available in off-chip main memory, prioritizing speed over capacity to sustain high throughput.

Memory and performance subsystems

Memory management unit

The Memory Management Unit (MMU) is a dedicated hardware component within the central processing unit responsible for translating virtual addresses generated by software into physical addresses in main memory, while enforcing to isolate processes and prevent unauthorized access. This translation enables systems, allowing programs to operate in a contiguous abstracted from the physical layout. The MMU achieves this by consulting page tables—data structures maintained by the operating system that map virtual page numbers (VPNs) to physical page numbers (PPNs)—and combining the PPN with the page offset from the virtual address to form the . Core functions of the MMU include performing page table lookups during memory accesses and caching these mappings in a (TLB), a small, high-speed typically organized as fully associative with entries for recent translations to minimize . On a TLB hit, the MMU retrieves the physical address directly in one or two cycles; on a miss, it initiates a page table walk, traversing the hierarchy (e.g., one to four levels in multilevel paging) to fetch the page table entry (PTE), which includes validity bits, protection flags, and the PPN. For protection, the MMU supports paging, which divides memory into fixed-size pages (commonly 4 KB) for non-contiguous allocation and isolation, and segmentation, which uses variable-sized segments defined by base and bound registers to group related code or data logically while checking bounds to avert overflows. These mechanisms operate in distinct modes—user mode for application code and kernel mode for the OS—ensuring processes cannot access each other's memory or kernel structures; violations trigger exceptions. If a PTE indicates an invalid page or permission breach (e.g., write to read-only), the MMU generates a , invoking the OS handler to load the page from disk or terminate the process. MMUs were first integrated into CPUs in the , with the PDP-11 series from establishing a foundational architecture that divided the 16-bit into eight 8 KB segments for mapping and protection across , , and modes, enabling expansion beyond 64 KB while supporting and data spaces. This design influenced subsequent systems by embedding translation hardware directly on-chip, reducing reliance on external components. However, address translation incurs overhead, as each access may require multiple cycles for walks—typically 1–4 references in single-level paging, escalating to dozens of cycles on TLB misses and contributing up to 16% of execution time in scale-out workloads due to interference and walk latency. In modern designs, MMUs incorporate optimizations like support for large pages (e.g., 2 or 1 ) to cover more with fewer TLB entries, reducing miss rates and walk frequency, though limited hardware coverage can increase overhead by up to 54% in others. For virtualization, nested paging—implemented via Intel's Extended Page Tables (EPT) or AMD's Nested Page Tables (nPT)—enables two-dimensional from guest virtual addresses through guest physical to host physical addresses, avoiding frequent traps but introducing up to 6× more PTE lookups (24 accesses versus 4 in native paging) and 2.4× slowdowns; mitigations include huge pages to alleviate this in environments.

Cache hierarchy

The cache hierarchy refers to the multi-level structure of small, fast on-chip memories integrated into modern CPUs to bridge the speed gap between the processor core and slower off-chip main memory. These caches store frequently accessed data and instructions, leveraging the principle of to minimize average access latency. Temporal locality exploits the tendency of programs to reuse recently accessed data soon after, while spatial locality capitalizes on accesses to nearby memory locations in the near future. This organization allows the CPU to achieve high performance by serving most requests from fast local storage rather than slower (DRAM). Caches are typically organized into three levels: L1, L2, and L3, all implemented using (SRAM), which provides lower and higher speed than DRAM but at the cost of smaller capacity and higher density requirements. The L1 cache is the smallest and fastest, positioned closest to each , and is usually split into separate (L1i) and (L1d) caches to allow simultaneous for fetching and execution. Typical L1 sizes range from 32 KB to 64 KB per core, enabling sub-nanosecond access times. The L2 cache is larger and unified, holding both instructions and data, often dedicated per core or shared among a small , with sizes commonly between 256 KB and 2 MB per core. The L3 cache, or last-level cache, is the largest in the , shared across all cores on the chip, and can reach 100 MB or more in multi-core designs to capture broader working sets. Cache organization relies on associativity to map memory blocks (cache lines, typically 64 bytes) to storage locations, balancing hit rate, access speed, and hardware complexity. In a direct-mapped cache, each block maps to exactly one location, offering simplicity and low latency but suffering from conflict misses when multiple blocks compete for the same slot. Set-associative caches divide the cache into sets, allowing multiple blocks (ways, often 2–8) per set, which reduces conflicts while keeping lookup feasible via parallel tag comparisons. Fully associative caches permit any block to map anywhere, maximizing flexibility and hit rates but requiring slower, more power-hungry searches across the entire cache. Most modern CPU caches employ set-associative designs, with L1 often 8-way and L3 16–20-way, to optimize for both speed and efficiency. Cache management involves policies for handling writes and ensuring consistency across cores. Write-through policies update both the cache and main memory on every write, ensuring immediate consistency but increasing memory bandwidth usage. Write-back policies defer memory updates until the cache line is evicted or flushed, reducing traffic through coalescing writes but risking data loss if not managed carefully, and are prevalent in performance-oriented designs. In multi-core systems, cache coherence protocols like MESI (Modified, Exclusive, Shared, Invalid) maintain a consistent view of data by tracking line states and invalidating or updating copies on shared bus snooping or directory-based mechanisms. The MESI protocol, developed at the University of Illinois, enables efficient write-back operation by distinguishing dirty (Modified) lines from clean shared ones, minimizing unnecessary traffic. A cache miss, where requested data is absent, imposes a significant penalty by stalling the core until data is fetched from a lower level or main , often taking tens to hundreds of cycles depending on the level. This can degrade performance substantially in latency-sensitive workloads, as the CPU waits idle during the transfer. To mitigate misses, hardware prefetching mechanisms anticipate future accesses based on patterns like sequential strides, proactively loading data into the ahead of , thereby hiding and improving throughput without explicit software intervention.

Clock and power mechanisms

The clock signal in a central processing unit (CPU) is generated by a quartz crystal oscillator, which vibrates at a precise frequency to produce periodic electrical cycles that synchronize the processor's operations. These oscillators typically operate in the gigahertz (GHz) range for modern CPUs, providing the timing reference for fetching, decoding, and executing instructions across multiple pipeline stages. The throughput of instructions is quantified using cycles per instruction (CPI), a metric that measures the average number of clock cycles required to complete one instruction, influencing overall performance as lower CPI values indicate higher efficiency. To balance performance and , CPUs employ techniques such as dynamic voltage and frequency scaling (DVFS), which dynamically adjust the operating frequency and supply voltage based on demands. DVFS reduces power consumption by lowering voltage and frequency during periods of low activity, as power dissipation scales quadratically with voltage and linearly with frequency, enabling significant savings—up to 18% in some systems—without proportionally impacting throughput. This mechanism is particularly vital in battery-powered or thermally constrained environments, where it prevents excessive heat generation while maintaining computational adequacy. In contrast to synchronous designs reliant on a global clock, asynchronous CPU architectures use clockless logic that employs handshaking protocols between circuit components to coordinate data flow and completion signals. These designs activate only the necessary logic gates on demand, reducing average power usage and eliminating clock-related overheads like and . A notable example is the AMULET series, which implements the ARM instruction set asynchronously and achieves rapid start-stop capabilities with lower energy per operation compared to clocked equivalents. Power delivery to the CPU is managed through voltage regulator modules (VRMs) integrated on the , which convert and stabilize the input voltage from the power supply to the precise levels required by the processor , often below 1V for modern nodes. VRMs consist of buck converters with multiple phases to handle high currents efficiently, ensuring stable operation under varying loads. To mitigate overheating, CPUs incorporate throttling, a mechanism that automatically reduces clock or activity when internal temperatures exceed safe thresholds, thereby capping power draw and preserving reliability. Key metrics for CPU power management include (TDP), which specifies the maximum heat dissipation a is engineered to produce under typical workloads, guiding cooling design. For instance, many desktop CPUs, such as certain models, have a TDP of 65W, balancing performance for general tasks while allowing for adequate thermal headroom in standard chassis configurations.

Parallelism and optimization

Instruction-level parallelism

Instruction-level parallelism (ILP) refers to the ability of a to execute multiple from a single concurrently, thereby increasing throughput without requiring multiple threads. This parallelism is exploited through techniques that overlap or reorder execution while preserving the program's semantic correctness. Key methods include pipelining, superscalar execution, and out-of-order processing, each addressing different aspects of instruction dependencies and resource utilization. Pipelining divides instruction execution into sequential stages—such as fetch, decode, execute, access, and write-back—allowing overlapping operations where a subsequent begins before the previous one completes. This overlap increases throughput to approach one per clock in ideal conditions, though hazards can reduce efficiency. For instance, in a five-stage , the fetch stage of n+1 occurs simultaneously with the decode stage of n. Seminal work on of pipelined architectures highlights how this technique scales performance by reducing the cycle time per . Superscalar architectures extend pipelining by incorporating multiple execution pipelines or issue units, enabling the simultaneous dispatch of independent instructions to different functional units in the same cycle. The , introduced in 1993, was the first superscalar implementation for the x86 architecture, featuring two integer pipelines that could process up to two , along with a . This design doubled the potential throughput over scalar processors like the 80486. Out-of-order execution further enhances ILP by dynamically reordering instructions at runtime, using structures like to dispatch ready instructions to functional units regardless of their original sequence. Pioneered by in the , this approach employs and a common data bus to track dependencies and forward results efficiently, preventing stalls from data hazards. In modern implementations, buffer instructions and operands, allowing execution as soon as sources are available. Dependencies between instructions limit ILP and manifest as hazards: true data hazards include read-after-write (), write-after-read (), and write-after-write (WAW), while false hazards arise from name dependencies without actual data conflicts. RAW hazards occur when an instruction reads a register before a prior write completes, potentially causing incorrect values; WAR and WAW involve write ordering issues. Hardware handles true RAW hazards via bypassing (forwarding) paths that route results directly from execution units to dependent instructions, bypassing the register file and reducing stalls by up to two cycles in a typical . False hazards, such as WAR and WAW, are resolved through , which maps architectural registers to physical ones to eliminate name conflicts without altering data flow. Control hazards from branches disrupt pipeline flow by altering the , leading to misprediction penalties of 10–20 cycles in deep . Branch mitigates this by speculatively fetching instructions based on predicted outcomes; static methods use fixed rules (e.g., always taken for backward branches), while dynamic schemes adapt using history tables. A common dynamic predictor employs a 2-bit saturating counter per , incrementing toward "taken" (11) on taken branches and toward "not taken" (00) otherwise, achieving 90–95% accuracy in typical workloads by that resists single flips. The seminal two-level adaptive predictor, building on such counters, indexes pattern history tables with global branch outcomes for correlated . Very long instruction word (VLIW) architectures represent an alternative for ILP, where compilers explicitly pack multiple independent operations into long instructions for parallel execution, offloading scheduling from hardware to software. The Intel Itanium processor (2001) adopted an explicitly parallel instruction computing (EPIC) variant of VLIW, bundling three 41-bit operations into 128-bit instructions with dependency hints, enabling up to six operations per cycle but relying heavily on compiler optimization. Despite these advances, ILP faces fundamental limits from sequential code portions and dependency chains, as quantified by , which states that the speedup from parallelizing a fraction p of a program is bounded by $1 / (1 - p + p/N), where N is the degree of parallelism; even high ILP yields diminishing returns if sequential bottlenecks persist.

Thread-level and data parallelism

Thread-level parallelism enables a CPU to execute multiple independent threads concurrently, improving throughput on multi-threaded applications by distributing workloads across processing units. (SMP) is a foundational approach where multiple identical processors share a common memory space and operating system, allowing threads to run in parallel while maintaining coherence through hardware mechanisms like cache snooping. Hyper-Threading Technology, introduced by in 2002 with the processor family, implements () to simulate additional logical cores on a single physical , duplicating architectural states such as registers while sharing execution resources to boost utilization without significantly increasing die size or power consumption. Multi-core processors advanced thread-level parallelism by integrating multiple independent cores on a single die, with AMD's in 2005 marking an early commercial desktop implementation of true dual- architecture, enabling across distinct processing units for enhanced in consumer workloads. Modern scalability in multi-core designs is achieved through architectures, as pioneered by in and families, where smaller modular dies are interconnected to form high-core-count processors, improving yield, reducing costs, and allowing flexible scaling for and desktop applications. Data parallelism complements thread-level approaches by processing multiple data elements simultaneously using a single instruction, primarily through (Single Instruction, Multiple Data) extensions. In x86 architectures, Intel's (SSE), introduced with the Pentium III, provide 128-bit vector operations for packed floating-point and integer data, while (AVX) extend this to 256-bit vectors for more efficient parallel computations in and scientific tasks. ARM processors employ as their SIMD extension, offering 128-bit vector registers to handle parallel operations on integers, floating-point, and fixed-point data, accelerating tasks like and inference on mobile and embedded systems. To coordinate parallel execution, synchronization mechanisms such as locks ensure for shared resources, preventing race conditions by allowing only one to access critical sections at a time, while barriers enforce collective waiting until all threads reach a designated point before proceeding. In large-scale multi-core systems, (NUMA) architectures introduce latency variations based on memory proximity to cores, requiring NUMA-aware to minimize remote access overheads and optimize thread placement for coherent data sharing across nodes. These parallelism techniques find application in workloads like , where threads handle independent scene elements across cores, and AI training, where data batches are processed in parallel via SIMD and multi-threading to accelerate model optimization on multi-core CPUs.

Asynchronous and specialized designs

Asynchronous CPU designs, also known as clockless or self-timed , operate without a global , relying instead on local handshaking protocols to coordinate data flow between components. This event-driven approach allows circuits to activate only when needed, potentially reducing power consumption compared to synchronous designs that continuously toggle regardless of activity. Pioneering work in this area includes the Caltech Asynchronous (), developed in the late 1980s as the first single-chip asynchronous , which demonstrated self-timed logic for instruction execution and using bundled-data signaling. These designs offer several advantages, including lower (EMI) due to the absence of periodic clock edges that generate noise, and adaptive operating speeds that adjust dynamically to workload and process variations without fixed clock rates. However, they face challenges in , as ensuring correct ordering and completion of operations across distributed components requires complex mechanisms, which can increase design overhead and verification difficulty. In pipelines, asynchronous elements can enhance flexibility by stages from a rigid clock, allowing data-dependent timing similar to self-timed logic while maintaining compatibility with synchronous cores; for instance, asynchronous arbiters and buffers have been integrated to handle variable without stalling the entire . Specialized CPU designs extend these principles to application-specific architectures, such as reconfigurable processors that incorporate FPGA-like logic blocks directly into the CPU fabric for on-the-fly customization of compute units, enabling efficient handling of domain-specific tasks like without external accelerators. processors represent another , mimicking neural structures with asynchronous, event-based computation; IBM's TrueNorth chip, released in 2014, features 1 million neurons and 256 million synapses in a scalable, low-power that processes asynchronously, consuming just 65 mW while supporting AI inference. Modern examples include RISC-V-based systems with custom ISA extensions for asynchronous peripherals, such as interfaces supporting parallel memory accesses via handshaking protocols, which allow the CPU to interface with event-driven I/O without clock synchronization overhead.

Operating modes and evaluation

Privilege and security modes

Privilege and security modes in central processing units (CPUs) define hierarchical execution states that isolate user applications from system resources, preventing unauthorized access and ensuring operational stability. These modes trace their origins to the 1970s by , which implemented kernel mode for full hardware control and user mode to restrict access to privileged instructions and memory, laying the foundation for modern protection mechanisms. In x86 architectures, privilege is structured into four rings (0–3), with ring 0 (kernel/supervisor mode) granting unrestricted access to hardware and instructions, while ring 3 (user mode) limits operations to prevent interference with the system; intermediate rings 1 and 2 are rarely used in contemporary designs. The current privilege level is tracked via the code segment register, and access violations trigger exceptions like general protection faults. ARM processors employ a similar model with four exception levels: EL0 for unprivileged user applications, EL1 for operating system kernels, EL2 for hypervisors, and EL3 for the most privileged secure firmware, where higher levels supersede lower ones in authority. Mode transitions occur through system calls, interrupts, and exceptions to enable secure interactions between privilege levels. In x86, a system call via SYSENTER or SYSCALL instructions from ring 3 invokes a software , prompting the CPU to save the user state, switch to ring 0 using the , and execute a handler; returns use IRET or SYSRET to restore the prior mode. Hardware interrupts from peripherals follow the same path, vectoring through the to handlers without allowing direct user escalation. In , synchronous exceptions from system calls elevate the exception level, with asynchronous interrupts or faults similarly routing to higher-level handlers, and returns via ERET instructions that demote the level while preserving isolation. Hardware enforces these protections via the , which checks during address translation and generates traps for violations, such as unauthorized memory access from lower rings. Security enhancements include the No eXecute (NX) bit in x86 processors, which designates pages as non-executable to block malicious code execution from data regions, a feature integrated since early 64-bit CPUs to mitigate exploits. CPUs also facilitate (ASLR) through virtual addressing, enabling OSes to randomize memory layouts for added exploit resistance. extensions like VT-x introduce VMX root mode for hypervisors (highest ) and non-root mode for guest OSes, supporting nested controls via VM entry and exit operations. Contemporary advancements extend these modes with hardware-isolated environments, such as , which create secure enclaves—protected memory regions accessible only to authorized code, shielding sensitive computations from even kernel-level access and enhancing confidentiality in multi-tenant systems. More recent advancements include Intel's Trust Domain Extensions (TDX), introduced in 2021, which enable hardware-isolated virtual machines with memory encryption and attestation for , and AMD's Secure Encrypted Virtualization with Secure Nested Paging (SEV-SNP), enhanced in 2022, providing similar protections against hypervisor attacks.

Performance metrics

Performance metrics for central processing units (CPUs) provide quantitative measures to evaluate computational capabilities, focusing on speed, efficiency, and resource utilization across various workloads. Key metrics include millions of (MIPS), which quantifies the number of machine instructions a CPU can execute in one second, offering a basic indicator of processing power. However, MIPS has limitations as it does not account for complexity or workload variations, potentially misleading comparisons between architectures. (FLOPS) measures the rate of arithmetic calculations on real numbers, essential for scientific and graphics-intensive tasks, with higher values indicating better handling of floating-point computations. The (SPEC) benchmarks, such as SPEC CPU 2017, assess compute-intensive performance through suites of integer and floating-point workloads, providing standardized scores for processor comparison across hardware platforms. Several factors influence these metrics, including instructions per cycle (IPC), which represents the average number of instructions executed per clock cycle and reflects architectural efficiency in utilizing stages. , the time to complete specific operations like memory access, and throughput, the rate of processing data streams, are also critical, varying based on CPU , cache design, and application demands. These factors are workload-dependent; for instance, branch-intensive programs may suffer from prediction errors, reducing effective IPC. Hardware performance counters enable detailed profiling by tracking events such as cycle counts, retired instructions, and branch mispredictions directly from CPU hardware registers. Tools like Linux perf utilize these counters to analyze bottlenecks, offering insights into real-time execution dynamics without relying solely on aggregate metrics. Benchmark suites like CoreMark evaluate embedded and general-purpose CPU performance through a mix of list processing, matrix manipulation, and state machine tasks, yielding scores that normalize for core efficiency. Geekbench provides cross-platform assessments of single-core and multi-core capabilities via workloads simulating everyday computing, such as image processing and machine learning inference. Real-world benchmarks, derived from actual applications like video encoding, better capture practical performance than synthetic ones, which may overemphasize peak theoretical rates but overlook overheads like I/O waits. Comparisons often distinguish single-thread performance, emphasizing per-core speed for sequential tasks, from multi-thread scores, which scale with core count for parallel workloads; for example, modern CPUs show 2-4x gains in multi-thread metrics over predecessors due to increased cores. Generational improvements, such as those in the mid-2010s in CPUs over seven years (as analyzed in ), have delivered up to 3x uplifts at similar power levels through enhanced and ism.

Enhancement techniques

Overclocking refers to the practice of increasing a CPU's clock beyond its manufacturer-specified limits to achieve higher . This is typically accomplished by adjusting the multiplier or base clock (BCLK) in the system's / settings, where the core is calculated as BCLK multiplied by the multiplier. For unlocked processors, such as Intel's K-series or AMD's non-locked models, users can raise the multiplier directly, while BCLK adjustments affect the entire system, including and PCIe devices, requiring careful tuning to maintain stability. Effective demands enhanced cooling solutions, such as air coolers, liquid cooling, or extreme methods like for record attempts, to dissipate the additional heat generated. However, pushing frequencies too high can lead to system instability, , accelerated , or permanent damage due to and thermal stress. Overclocking originated in the and but gained widespread popularity in the early with Intel's processors, when enthusiasts began modifying jumpers to select higher clock speeds on compatible boards, effectively turning slower CPUs into faster variants without hardware alterations. This era marked the birth of as a , driven by the desire to extend the life of aging amid rapid advancements in processor speeds. By the mid-1990s, techniques evolved to include simple multiplier unlocks on chips like the for 486 systems, laying the groundwork for modern practices. To validate an overclock's stability, users commonly employ stress testing software like Prime95, which performs intensive mathematical computations to push the CPU to its limits and detect errors or crashes. Tests are run for extended periods, often 24 hours or more, under high-priority settings to simulate worst-case loads and ensure reliability without real-world application mismatches. Undervolting complements by reducing the CPU's operating voltage below stock levels, which lowers power consumption and heat output while maintaining or even improving through better thermal headroom. This technique is particularly beneficial for , extending life in laptops and reducing costs in desktops, without sacrificing clock speeds if the quality allows stable operation at lower voltages. Proper undervolting requires tools to avoid instability from insufficient voltage, but it enhances overall system by minimizing thermal degradation. In contemporary designs, automated enhancement features like and AMD Precision Boost provide safe, dynamic within predefined power and thermal limits. opportunistically increases core frequencies above the base clock during lighter workloads, balancing performance with by scaling based on available power budget. Similarly, 's Precision Boost 2 adjusts clock speeds in real-time across processors, leveraging per-core sensors to maximize frequency while adhering to safe operational envelopes. These technologies represent manufacturer-sanctioned enhancements, eliminating much of the risk associated with manual .

References

  1. [1]
    How The Computer Works: The CPU and Memory
    The central processing unit (CPU), is a highly complex, extensive set of electronic circuitry that executes stored program instructions.
  2. [2]
    CPU, GPU, ROM, and RAM - E 115 - NC State University
    Central Processing Unit (CPU)​​ The CPU is often called the “brain” of the computer. It performs all the basic calculations and logic operations (like adding ...
  3. [3]
    5.5. Building a Processor - Dive Into Systems
    The CPU is constructed from basic arithmetic/logic, storage, and control circuit building blocks. Its main functional components are the arithmetic logic ...
  4. [4]
    Components of the CPU - Dr. Mike Murphy
    Mar 29, 2022 · The CPU is actually comprised of several different components, including the Control Unit, ALU, and interfaces to memory and I/O devices.
  5. [5]
    Instructing the CPU - Stephen Marz
    The most basic central processing unit can be divided into three parts: (1) control unit, (2) arithmetic and logic unit (ALU), and (3) register file.<|control11|><|separator|>
  6. [6]
    6.2. The Processor — CS160 Reader - Chemeketa CS
    This basic model of a computer as a processing unit that takes input, communicates with a memory and produces output is known as the von Nuemann architecture.
  7. [7]
    The Modern History of Computing
    Dec 18, 2000 · The Analytical Engine was to have had a memory store and a central processing unit (or 'mill') and would have been able to select from among ...
  8. [8]
    [PDF] The Birth, Evolution and Future of Microprocessor
    Intel 4004 was the first commercially available single-chip microprocessor in history. It was a 4-bit CPU designed for usage in calculators, designed for " ...
  9. [9]
    What is a Central Processing Unit (CPU)? - IBM
    The central processing unit (CPU) is the invisible manager inside the computer where data input is transformed into information output.
  10. [10]
    The central processing unit (CPU): Its components and functionality
    Jul 23, 2020 · The term central processing unit originated way back in the mists of computer time when a single massive cabinet contained the circuitry ...
  11. [11]
    [PDF] First draft report on the EDVAC by John von Neumann - MIT
    IEEE Annals of the History of Computing, Vol. 15, No. 4, 1993 •. 27. Page 2. First Draft of a Report on the EDVAC by. John von Neumann. Contract No. W-670-ORD ...
  12. [12]
    Complex Instruction Set Computer Architecture - ScienceDirect.com
    Processor architectures are classified as either a Reduced Instruction Set Computer (RISC) or as a Complex Instruction Set Computer (CISC).
  13. [13]
    [PDF] RISC and CISC - Semantic Scholar
    Comparison of RISC & CISC in details, encompassing the addressing modes, evolution, definitions and characteristics, is made. Comparison of RISC & CISC in ...
  14. [14]
    Revisiting the RISC vs. CISC debate on contemporary ARM and x86 ...
    The main advantage of the RISC processors is simplicity; these processors contain less transistors and take less silicon die area, which makes its power ...
  15. [15]
    (PDF) CPU vs. SOC - The battle for the future of computing
    Jun 7, 2016 · SoCs are the next step after CPUs. Eventually, SoCs will almost completely consume CPUs. We are already seeing this with AMD's Llano and Intel's Ivy Bridge ...
  16. [16]
    System-on-Chip - an overview | ScienceDirect Topics
    A system on chip (SoC) is defined as an integrated circuit that incorporates a microprocessor and additional components necessary to perform specific functions ...
  17. [17]
    Hardware Architectures for Real-Time Medical Imaging - MDPI
    This paper focuses on the evolution and the application of different hardware architectures (namely, CPU, GPU, DSP, FPGA, and ASIC) in medical imaging
  18. [18]
    Special Purpose Processor - an overview | ScienceDirect Topics
    Special purpose processors refer to processing devices designed for specific tasks, distinct from general-purpose microprocessors, and may include ...Missing: scholarly | Show results with:scholarly
  19. [19]
    (PDF) Fifty years of microprocessor evolution: from single CPU to ...
    Aug 6, 2025 · A comparison between single and multiple-core processor is given in Table 4. ... Therefore, heterogeneous multi-core systems 1 (HMCS) emerged.
  20. [20]
    The multicore architecture - ScienceDirect.com
    Multicore architectures can be heterogeneous or homogeneous. In homogeneous architectures, as the name suggests, all the cores on the device are the same. In ...Missing: evolution | Show results with:evolution
  21. [21]
    Microarchitecture - an overview | ScienceDirect Topics
    A single ISA, such as RISC-V, MIPS, ARM, or x86, can have many different microarchitectures, each offering distinct trade-offs in performance, cost, and ...
  22. [22]
    [PDF] Processor Microarchitecture: An Implementation Perspective
    Processor microarchitectures can be classified along multiple orthogonal dimensions. ... A Sub-1W to 2W Low-Power ISA Processor for Mobile Internet Devices.
  23. [23]
    The Modern History of Computing
    Dec 18, 2000 · Babbage's proposed Difference Engine was a special-purpose digital computing machine for the automatic production of mathematical tables (such ...
  24. [24]
    KONRAD ZUSE 1910-1995 - National Academy of Engineering
    Zuse is widely credited with the creation of the first functioning, freely programmable, and fully automatic digital computer.
  25. [25]
    ENIAC - Penn Engineering
    Originally announced on February 14, 1946, the Electronic Numerical Integrator and Computer (ENIAC), was the first general-purpose electronic computer.
  26. [26]
    EDSAC - Clemson University
    The EDSAC (electronic delay storage automatic calculator) performed its first calculation at Cambridge University, England, in May 1949.
  27. [27]
    1947: Invention of the Point-Contact Transistor | The Silicon Engine
    John Bardeen & Walter Brattain achieve transistor action in a germanium point-contact device in December 1947. Bardeen, Brattain, and Shockley ( ...
  28. [28]
    1948: Conception of the Junction Transistor | The Silicon Engine
    After Bardeen and Brattain's December 1947 invention of the point-contact transistor (1947 Milestone), Bell Labs physicist William Shockley began a month of ...
  29. [29]
    How the First Transistor Worked - IEEE Spectrum
    Nov 20, 2022 · The first recorded instance of a working transistor was the legendary point-contact device built at AT&T Bell Telephone Laboratories in the fall of 1947.
  30. [30]
    1953: Transistorized Computers Emerge | The Silicon Engine
    In 1953, a transistorized computer prototype was demonstrated. In 1954, TRADIC was built, and in 1956, TX-0 and ETL Mark III were created. By 1960, new designs ...
  31. [31]
    Bell Labs TRADIC Computer - 102627244 - CHM
    From this program came the first all transistor computer, TRADIC, announced by Bell Telephone Laboratories in 1955.
  32. [32]
    The History of IEEE and Electrotechnologies
    Computers and Computing. 1959. IBM 7090, one of the first fully transistorized computers. 1952. John Von Neumann with his experimental IAS computer. By the late ...
  33. [33]
    [PDF] The development of the most popular computer of the 1960s and the ...
    In 1958, IBM was already design- ing several large-scale computers using SMS technology: the scientific IBM 7090 (delivered in November 1959), the commercial ...
  34. [34]
    The chip that changed the world | TI.com - Texas Instruments
    When Jack Kilby invented the first integrated circuit (IC) at Texas Instruments in 1958 ... integrated into two integrated circuits so small they fit ...
  35. [35]
    The Men Who Made the Microchip - IEEE Spectrum
    At Fairchild in 1959, Noyce conceived and patented the ideas that were to ... To fabricate microchips using the planar process required precise optical masks ...
  36. [36]
    IEEE Milestone: Semiconductor Planar Process and Integrated Circuit
    Jun 6, 2019 · At Fairchild Semiconductor, Jean Hoerni's revolutionary planar process inspired Robert Noyce's vision of interconnecting multiple elements on a ...
  37. [37]
    ICs Rocket to Success - CHM Revolution - Computer History Museum
    Requiring 200,000 Fairchild Micrologic circuits, the Apollo Guidance Computer project was the largest user of ICs through 1965. View Artifact Detail · Eldon C.
  38. [38]
    Silicon Chips Take Man to the Moon - Computer History Museum
    Jul 17, 2019 · The Apollo guidance, navigation, and control computer (AGC) evolved from an earlier MIT/IL Mars probe study for the Air Force. In 1960 Ramon ...
  39. [39]
    [PDF] Understanding microprocessors
    Thru the 1970's-LSI to Very Large Scale Integration (1,000 to 50,000 Gates) ... • Medium-scale computers have been reduced in cost by 300 times as a result ...
  40. [40]
    [PDF] Fairchild Symbol Computer
    9 The advent of LSI chips and memory later fostered the development of the single-chip CPU at Intel.10–12. The earlier computer development project in Fair-.
  41. [41]
    Announcing a New Era of Integrated Electronics - Intel
    ... Intel 4004, became the first general-purpose microprocessor. Hoff, Federico Faggin (who took over design leadership from Hoff) and Stan Mazor at Intel ...
  42. [42]
    How Ted Hoff Invented the First Microprocessor - IEEE Spectrum
    Teaming up with Stanley Mazor and Federico Faggin, he created the first commercial microprocessor, the Intel 4004.
  43. [43]
    1971: Microprocessor Integrates CPU Function onto a Single Chip
    Intel 4004 Microprocessor, interview with Ted Hoff & Stan Mazor (2006-09-20) · Intel 4004 Microprocessor oral history panel - Faggin, Feeny, Hoff, Masatoshi ...
  44. [44]
    The Intel 8008
    Introduced in April 1972, the Intel 8008 was the world's first 8-bit programmable microprocessor and only the second microprocessor from Intel.
  45. [45]
    The Beginning of a Legend: The 8086 - Explore Intel's history
    Future generations of x86 chips would build on the 8086's architecture rather than discarding it, allowing users to upgrade their devices without needing to ...
  46. [46]
    The Motorola 68000: A 32-Bit Brain in a 16-Bit Body - All About Circuits
    Sep 26, 2025 · When the Motorola 68000 landed in 1979, it represented a leap in how CPUs could be designed. Internally, it used a full 32-bit architecture, ...
  47. [47]
    Microprocessors: the engines of the digital age - PubMed Central - NIH
    Mar 15, 2017 · The microprocessor—a computer central processing unit integrated onto a single microchip—has come to dominate computing across all of its scales ...
  48. [48]
    Moore's Law - Intel
    Observing these trends, Moore published a paper entitled "Cramming More Components onto Integrated Circuits" in the April 19, 1965 issue of the trade journal ...
  49. [49]
    Fetch, decode, execute (repeat!) – Clayton Cafiero
    Sep 9, 2025 · Once execution is complete, the cycle begins again: the CPU fetches the next instruction, decodes it, executes it, and so forth. This process ...
  50. [50]
    5.6. The Processor's Execution of Program Instructions
    Because it takes one clock cycle to complete one stage of CPU instruction execution, a processor with a four-stage instruction execution sequence (Fetch, Decode ...
  51. [51]
    Chapter 6 Central Processing Unit - Robert G. Plantz
    In this chapter we move on to consider a programmer's view of the Central Processing Unit (CPU) and how it interacts with memory.
  52. [52]
    [PDF] Computer Architecture 2/26/01 Lecture #9 16.070 - MIT
    Feb 26, 2001 · ➢ "Points" to memory location of next instruction to be executed. ➢ After Fetch operation, program counter is incremented to point to the next.
  53. [53]
    15. Inside a Modern CPU - University of Iowa
    The instruction fetch stage begins the exeuction of each instruction by fetching it. This may involve a complete memory cycle, or at least, a an I-cache cycle.Missing: phase | Show results with:phase
  54. [54]
    [PDF] Fetch Directed Instruction Prefetching - UCSD CSE
    Instruction prefetching has been proposed as a mechanism to help reduce instruction cache misses, which in turn can help increase instruction supply to the ...
  55. [55]
    [PDF] Instruction Prefetch Strategies in a Pipelined Processor
    Jan 13, 1983 · Instructions in prior stages of the pipeline prior can be aborted at any time without changing the processor's state. The control pointt is also ...
  56. [56]
    [PDF] Reducing Error Correction Latency for On-Chip Memories
    One simple approach to account for the additional delay in the instruction-fetch and data-load stages due to error correction without stalling the processor ...
  57. [57]
    [PDF] Reducing Cache Power with Low-Cost, Multi-bit Error-Correcting ...
    Jun 23, 2010 · Another promising approach to increase DRAM refresh times is the use of error-correcting codes (ECC) to dynamically identify and repair bits ...
  58. [58]
    5.2. The von Neumann Architecture - Dive Into Systems
    In the fetch phase, the control unit reads the instruction at the memory address stored in the PC (1234). It sends the address on the address bus, and a READ ...
  59. [59]
    [PDF] Instruction Codes - Systems I: Computer Organization and Architecture
    The instructions of a program are carried out by a process called the instruction cycle. • The instruction cycle consists of these phases: – Fetch an ...
  60. [60]
    [PDF] LECTURE 7 Pipelining - FSU Computer Science
    ID stage: decodes the instruction, reads source registers from register file, sign- extends the immediate value, calculates the branch target address and ...
  61. [61]
    [PDF] Improving CISC Instruction Decoding Performance Using a Fill Unit
    As described in the introduction, we propose to use a ll unit to collect decoded microoperations and then store them into the lines of a decoded instruction ...Missing: phase | Show results with:phase
  62. [62]
    [PDF] Context-Sensitive Decoding: On-Demand Microcode Customization ...
    May 8, 2019 · This paper proposes context-sensitive decoding, a technique that enables customization of the micro-op translation based on the current ...
  63. [63]
    12. Handling Data Hazards - UMD Computer Science
    By identifying the hazard in the ID stage, we can insert a bubble into the pipeline by changing the EX, MEM, and WB control fields of the ID/EX pipeline ...
  64. [64]
    Lecture 5: Out-of-order Execution
    Instructions are fetched, decoded, executed, etc. The decode stage is where we find out about structural, data, and control hazards.
  65. [65]
    [PDF] Processor Pipeline
    Spring 2015 :: CSE 502 – Computer Architecture. Pipeline: Data Hazard. • Only RAW in our case. • How to detect? – Compare read register specifiers for newer ...
  66. [66]
    1.4 Instruction Cycles - Engineering LibreTexts
    Aug 20, 2021 · Execute Stage: The control unit of the CPU passes the decoded information as a sequence of control signals to the relevant functional units of ...
  67. [67]
    5.6. The Processor's Execution of Program Instructions
    After fetching the instruction, the CPU decodes the instruction bits stored in the IR register into four parts: the high-order bits of an instruction encode ...
  68. [68]
    Primary Instruction Cycles - GeeksforGeeks
    Oct 22, 2025 · The Instruction Cycle is the basic operational process of a computer's CPU, referring to the sequence of steps it follows to fetch, decode, and ...
  69. [69]
    [PDF] 7- Execution of the instruction
    Computer Architecture ………………………… … Lecture No.6,7. 39. 7- Execution of the ... • Common status flags: – Carry (C). – Positive result (P). – Zero result (Z).
  70. [70]
    CPU: Central Processing Unit | AP CSP (article) - Khan Academy
    The CPU can process those instructions easily, thanks to a control unit that knows how to interpret program instructions and an Arithmetic Logic Unit (ALU) ...
  71. [71]
    None
    ### Summary of Hardwired and Microprogrammed Control
  72. [72]
    The processor - Isaac Computer Science
    The control unit organises the execution of instructions, including managing the other components in the processor. The control unit is responsible for the ...
  73. [73]
    [PDF] Topic 1 Evolution of ILP in Microprocessors - Rice University
    What is ILP? – Processor and Compiler design techniques that speed up execution by causing individual machine operations to execute in parallel.
  74. [74]
    Organization of Computer Systems: Processor & Datapath - UF CISE
    Instruction Fetch and Decode, Data Fetch. Each instruction execution first fetches the instruction, decodes it, and computes both the sequential PC and ...Missing: phase | Show results with:phase
  75. [75]
    [PDF] Constructing a Basic Arithmetic Logic Unit
    These four operations—add, subtract, AND, OR—are found in the ALU of almost every computer, and the operations of most MIPS instructions can be performed by ...
  76. [76]
    Research and design of reconfigurable 64-bit ALU - IEEE Xplore
    Research and design of reconfigurable 64-bit ALU. Abstract: The logic design of ALU, an important constituent part of CPU, is described in this paper. The ...
  77. [77]
    [PDF] Lecture 8:
    An N-bit carry-lookahead adder is generally much faster than a ripple-carry ... ALU with Status Flags: Carry. C = 1 if: Cout of Adder is 1. AND. ALU is ...
  78. [78]
    Registers in AArch64 - general-purpose registers - Arm Developer
    Most A64 instructions operate on registers. The architecture provides 31 general purpose registers. Each register can be used as a 64-bit X register (X0.<|separator|>
  79. [79]
    Intel® 64 and IA-32 Architectures Software Developer Manuals
    Oct 29, 2025 · These manuals describe the architecture and programming environment of the Intel® 64 and IA-32 architectures.Missing: general | Show results with:general
  80. [80]
    Processor core registers summary - Arm Developer
    R0-R12 are general-purpose registers for data operations. MSP (R13). The Stack Pointer (SP) is register R13. In Thread mode, the CONTROL register indicates ...
  81. [81]
    Addressing modes - Arm Developer
    The addressing modes are described as follows: Offset. The offset is added to or subtracted from the base register to form the memory address.
  82. [82]
    Loads and stores - addressing - Arm Developer
    There are several addressing modes that define how the address is formed. Base register - The simplest form of addressing is a single register.
  83. [83]
    A compile-time managed multi-level register file hierarchy
    This work redesigns the register file system of a modern throughput processor with a combined hardware and software solution that reduces register file ...
  84. [84]
  85. [85]
    [PDF] Virtual Memory: Protection and Address Translation - cs.Princeton
    Generic Address Translation. ◇ Memory Management Unit. (MMU) translates virtual address into physical address for each load and store. ◇ Combination of ...
  86. [86]
    [PDF] Virtual Memory - Computer Systems: A Programmer's Perspective
    Dedicated hardware on the CPU chip called the memory management unit (MMU) translates virtual addresses on the fly, using a look-up table stored in main memory ...
  87. [87]
    PDP-11 Memory Management - Computer History Wiki
    Jul 27, 2024 · PDP-11s which provide memory management use a standard PDP-11 Memory Management architecture. When memory management is enabled, ...Simplified subset · Registers · Control · PARs and PDRs
  88. [88]
    [PDF] Performance Analysis of the Memory Management Unit under Scale ...
    Thus, a page walk requires between one and four memory operations to perform an address translation based on the contents of the MMU cache. III. METHODOLOGY.
  89. [89]
  90. [90]
    [PDF] Translation Pass-Through for Near-Native Paging Performance in VMs
    Nested paging, however, introduces over- heads during address translation: it virtualizes guest physical addresses by combining guest page tables with an ...
  91. [91]
    The locality principle | Communications of the ACM
    Locality of reference is a fundamental principle of computing with many applications. Here is its story.Missing: CPU | Show results with:CPU
  92. [92]
    The case for SRAM main memory
    The level closest to the CPU is the first level or L1 cache, with numbers increasing as the cache is closer to the main memory (L2, L3, etc.).
  93. [93]
    Memory Performance in a Nutshell - Intel
    Jun 6, 2016 · L1 cache, 32 KB, 1 nanosecond ; L2 cache, 256 KB, 4 nanoseconds ; L3 cache, 8 MB or more, 10x slower than L2 ; MCDRAM, 2x slower than L3 ...Missing: hierarchy SRAM
  94. [94]
    AMD Ryzen™ 9 9900X3D Gaming and Content Creation Processor
    L1 Cache: 960 KB. L2 Cache: 12 MB. L3 Cache: 128 MB. Default TDP: 120W. Processor Technology for CPU Cores: TSMC 4nm FinFET. Processor Technology for I/O Die ...Missing: hierarchy sizes
  95. [95]
    [PDF] 356477-Optimization-Reference-Manual-V2-002.pdf - Intel
    Alternative to Prevent AGU and Execution Unit Dependency . ... The address generation unit (AGU) may be used directly in three situations ...
  96. [96]
    Impact of Illinois on Parallel Computing Advances - I2PC
    Illinois Cache Coherence Protocol (MESI): The Illinois protocol, developed by Janak Patel in 1983, became the IEEE MESI standard and is used today by ...
  97. [97]
    A Survey of Recent Prefetching Techniques for Processor Caches
    If no match occurs, then a miss is identified. Once a miss is identified in L1 cache, a miss request can be issued to L2 cache to prefetch data in the L1 cache.
  98. [98]
    [1706.10156] Noise-induced Synchronization of Crystal Oscillators
    Apr 11, 2017 · Our study presents a guideline for synchronizing clocks of multiple CPU systems, distributed sensor networks, and other engineering devices.Missing: source:
  99. [99]
  100. [100]
  101. [101]
  102. [102]
    [PDF] Voltage Regulator Module (VRM) and Enterprise Voltage ... - Intel
    Intel® Xeon™ processor with 800 MHz system bus and Low Voltage Intel® Xeon™ processor with 800 MHz system bus. The intent of this document is to define the ...
  103. [103]
    Performance optimal processor throttling under thermal constraints
    Sep 30, 2007 · We derive analytically, the performance optimal throttling curve for a processor under thermal constraints for a given task sequence.
  104. [104]
    [PDF] 3rd-gen-core-lga1155-socket-guide.pdf - Intel
    Thermal Design Power (TDP) should be used for processor thermal solution design targets. TDP is not the maximum power that the processor can dissipate. TDP ...
  105. [105]
    [PDF] Instruction Level Parallelism (ILP)
    Instructions are issued to the pipeline in-order but executed and completed out-of-order. out-of-order execution leading to the possibility of out-of-order.
  106. [106]
    Computer Architecture - 6th Edition - Elsevier Shop
    In stock Free deliveryA Quantitative Approach · 6th Edition - November 23, 2017 · Newer edition is available · Imprint: Morgan Kaufmann · Authors: John L. Hennessy, David A. Patterson.
  107. [107]
    The Pentium: An Architectural History of the World's Most Famous ...
    Jul 11, 2004 · While its superscalar design (Intel's first) certainly improved on the performance of its predecessor, the 486, the main thing that the Pentium ...
  108. [108]
    [PDF] An Efficient Algorithm for Exploiting Multiple Arithmetic Units
    The common data bus improves performance by efficiently utilizing the execution units without requiring specially optimized code.
  109. [109]
    [PDF] Alternative Implementations of Two-Level Adaptive Branch Prediction
    In Two-Level Adap- tive Branch Prediction, the 2-bit saturating up-down counter keeps track of the history of a certain history pattern. The counter is ...
  110. [110]
    Itanium — A System Implementor's Tale - USENIX
    Mar 2, 2005 · Itanium is a fairly new and rather unusual architecture. Its defining feature is explicitly-parallel instruction-set computing (EPIC).
  111. [111]
    [PDF] Validity of the Single Processor Approach to Achieving Large Scale ...
    Amdahl. TECHNICAL LITERATURE. This article was the first publica- tion by Gene Amdahl on what became known as Amdahl's Law. Interestingly, it has no equations.
  112. [112]
    [PDF] Intel Technology Journal
    Feb 14, 2002 · The two hardware contexts are exposed to the user as two symmetric multiprocessing logical processors. The on- chip cache hierarchy has the ...
  113. [113]
    July 13, 2005 - PRESS RELEASE - 8-K: Current report filing - AMD
    AMD shipped true Dual-Core technology with the AMD Athlon™ 64 X2 Dual-Core processors for desktop PCs and Dual-Core AMD Opteron™ processors for servers and ...Missing: multi- | Show results with:multi-
  114. [114]
    Pioneering Chiplet Technology and Design for the AMD EPYC ...
    This paper details the technology challenges that motivated AMD to use chiplets, the technical solutions we developed for our products, and how we expanded the ...Missing: scalability | Show results with:scalability<|separator|>
  115. [115]
    Intel® Instruction Set Extensions Technology
    Explains Instruction Set Extensions include SSE Streaming SIMD Extensions technologies, including SSE2, SSE3, SSE4, and AVX (Advanced Vector Extensions).Missing: x86 | Show results with:x86
  116. [116]
    Neon - Arm Developer
    Arm Neon technology is an advanced Single Instruction Multiple Data (SIMD) architecture extension for the A-profile and R-profile processors.
  117. [117]
    Scalability Techniques For Practical Synchronization Primitives
    Jan 1, 2015 · Locks are a way of allowing multiple threads to execute concurrently, providing safe and correct execution context through mutual exclusion. To ...
  118. [118]
    An Overview of Non-Uniform Memory Access
    Sep 1, 2013 · An overview of non-uniform memory access. NUMA becomes more common because memory controllers get close to execution units on microprocessors.
  119. [119]
    Parallel Approaches in Deep Learning: Use Parallel Computing
    This approach facilitates the simultaneous processing of multiple data sets through a single instruction on a multi-core processor. However, if a GPU is ...
  120. [120]
    [PDF] Asynchronous vs. Synchronous Microporcessors - Auburn University
    There are five main advantages to asynchronous circuits: low power, low Electromagnetic Interference (EMI), high speed, high tolerance to some types of errors, ...Missing: adaptive | Show results with:adaptive
  121. [121]
  122. [122]
    [PDF] A Case for Asynchronous Computer Architecture
    The architecture has some of the benefits of an out-of-order processor without the complex instruction issue logic present in modern clocked proces- sors such ...
  123. [123]
    [PDF] An Introduction to Reconfigurable Computing
    Reconfigurable computing performs computations in hardware for increased performance, while retaining software flexibility, bridging the gap between hardware ...
  124. [124]
    [PDF] TrueNorth: Design and Tool Flow of a 65 mW 1 Million Neuron ...
    In this paper, we present the design of the TrueNorth chip and the novel asynchronous–synchronous design tool flow. In Section II, we review related ...
  125. [125]
    [PDF] Architecture and RISC-V ISA Extension Supporting Asynchronous ...
    We propose an architecture that supports explicitly manipu- lating massively parallel asynchronous memory accesses and its. RISC-V ISA extension ...
  126. [126]
    [PDF] pdp11-40.pdf - PDOS-MIT
    The machine operates in two modes; Kernel and User. When the machine is in Kernel mode a program has complete control of the machine; when in User mode the ...
  127. [127]
    [PDF] Intel® 64 and IA-32 Architectures Software Developer's Manual
    NOTE: The Intel® 64 and IA-32 Architectures Software Developer's Manual consists of nine volumes: Basic Architecture, Order Number 253665; Instruction Set ...
  128. [128]
    Exception levels - Learn the architecture - AArch64 Exception Model
    For example, the lowest level of privilege is referred to as EL0. As shown in Exception levels, there are four Exception levels: EL0, EL1, EL2 and EL3. Figure 1 ...
  129. [129]
    [PDF] Chapter 3 System calls, exceptions, and interrupts - Columbia CS
    Interrupts are similar to system calls, except devices generate them at any time. There is hardware on the motherboard to signal the CPU when a device needs ...
  130. [130]
    How to Find the Execute Disable Bit for Intel® Processors
    The Execute Disable Bit is a hardware-based security feature that can reduce exposure to viruses and malicious-code attacks.
  131. [131]
    Address Space Layout Randomization (ASLR) - Arm Developer
    Oct 24, 2015 · ASLR is a security feature which randomizes where various parts of a Linux application are loaded into memory.
  132. [132]
    [PDF] Intel® Virtualization Technology Specification for the IA-32 Intel ...
    1.1 ABOUT THIS DOCUMENT​​ This documents describes Intel® Virtualization Technology for IA-32 processors, referred to as VT-x. VT-x constitutes a set of virtual- ...
  133. [133]
    Intel® Software Guard Extensions (Intel® SGX)
    Intel® SGX protects data actively being used in the processor and memory by creating a trusted execution environment (TEE) called an enclave. Users can scale ...
  134. [134]
    What is million instructions per second (MIPS)? - TechTarget
    Mar 1, 2023 · MIPS measures a processor's speed, delivering a standard representing the number of instructions CPUs can process in 1 second.
  135. [135]
    What is floating-point operations per second (FLOPS)? - TechTarget
    Aug 22, 2023 · FLOPS is a measure of a computer's performance based on the number of floating-point arithmetic calculations that the processor can perform within a second.
  136. [136]
    SPEC CPU®2017 Overview / What's New?
    Sep 3, 2024 · SPEC CPU 2017 focuses on compute intensive performance, which means these benchmarks emphasize the performance of: Processor - The CPU chip(s).
  137. [137]
    Analysis of benchmark characteristics and benchmark performance ...
    This paper uses a machine-independent model to characterize machine and program execution, estimate execution times, and identify key operations in benchmarks.
  138. [138]
    For better or worse, benchmarks shape a field: technical perspective
    We then went through a sequence of performance metrics, with each being an improvement on its predecessor: average instruction time, millions of instructions ...Missing: FLOPS | Show results with:FLOPS
  139. [139]
    About Performance Counters - Win32 apps | Microsoft Learn
    Jul 14, 2025 · Counters are used to provide information as to how well the operating system or an application, service, or driver is performing.
  140. [140]
    perf: Linux profiling with performance counters
    Aug 10, 2024 · Performance counters are CPU hardware registers that count hardware events such as instructions executed, cache-misses suffered, or branches ...perf: Linux profiling with... · Introduction
  141. [141]
    CoreMark - CPU Benchmark - EEMBC
    EEMBC's CoreMark® is a benchmark that measures the performance of microcontrollers (MCUs) and central processing units (CPUs) used in embedded systems.Scores · FAQ · Download · CoreMark-PRO
  142. [142]
    Geekbench 6 - Cross-Platform Benchmark
    CPU Benchmark. Geekbench 6 measures your processor's single-core and multi-core power, for everything from checking your email to taking a picture to playing ...
  143. [143]
    SPEC CPU ® 2017 benchmark
    SPEC designed these suites to provide a comparative measure of compute-intensive performance across the widest practical range of hardware using workloads ...SPEC CPU2017 Results · Overview · Documentation · SPEC releases major new...
  144. [144]
    PassMark CPU Benchmarks - Single Thread Performance
    This chart comparing the single thread performance of CPUs is based on the average PerformanceTest benchmark results from millions of machines and is updated ...
  145. [145]
    Quantifying the impact of generational mobile CPU design trends on ...
    We analyze the evolution often cutting-edge mobile CPU designs released over the past seven years. Specifically, we report measured performance, power, energy ...
  146. [146]
    How to Overclock Your Unlocked Intel® Core™ Processor
    Simply put: BCLK x Multipliers = CPU Core Frequency. Example: 100 MHz (BCLK) ... It's important that you use an adequate cooling solution when attempting to ...Missing: techniques | Show results with:techniques
  147. [147]
    How to Overclock Your CPU: Get the Most GHz from Your Processor
    May 6, 2023 · The formula to determine the processor's frequency consists of multiplying the base clock (BCLK) by the CPU multiplier. For example, a processor ...
  148. [148]
    History of overclocking - The Silicon Underground
    Jan 11, 2017 · By the early 1990s, motherboard makers started putting jumpers on the boards for clock speed. This let them sell one board and a reseller could ...
  149. [149]
    The 486 CPU Era – The Birth of Overclocking. – Part 1
    Feb 21, 2021 · To extend the lifespan of the 486 platform, Intel in February 1995 releases a special version of the processor – Intel Pentium OverDrive with a ...
  150. [150]
    Prime95 - Stress Test Your CPU - Windows 10 Forums
    Aug 15, 2015 · It is now widely used as a CPU stress testing utility to gauge the stability of a CPU, especially when overclocking a system.
  151. [151]
    The right way to stress-test an overclocked PC - Hope This Helps
    Nov 25, 2012 · For proper Prime95 stability testing, you really should run for 24 hours with a priority setting of “10”. Also the settings you choose can ...
  152. [152]
    Overclocking vs Undervolting: Whats The Difference?
    Sep 21, 2023 · Advantages: More efficient thanks to reduced power consumption; Quieter; Less heat generated; Can achieve the same clock speeds. Disadvantages:.
  153. [153]
    Undervolting Your PC: What It Is and Why You Should Try It
    Jun 12, 2024 · 1. Improved thermal performance · 2. Increased battery life · 3. Enhanced stability and longevity · 4. Improved performance potential.
  154. [154]
    What Is Intel® Turbo Boost Technology?
    Intel Turbo Boost Technology is an energy-efficient solution to this imbalance: it lets the CPU run at its base clock speed when handling light workloads.
  155. [155]
    AMD Ryzen™ Technology: Precision Boost 2 Performance ...
    Precision Boost 2 is a performance-maximizing technology available in all AMD Ryzen™ and Ryzen™ Threadripper™ 2000 Series processors (or newer). This technology ...