Microprocessor and Microcontroller Lecture.pptx
- 1. Microprocessors & Microcontrollers CSE 3815 Spring 25 Lecture 1
- 2. Microprocessor • Computer's Central Processing Unit (CPU) built on a single Integrated Circuit (IC) is called a microprocessor. • It is a programmable, multipurpose, clock -driven, register-based electronic device that reads binary instructions from a storage device called memory, accepts binary data as input and processes data according to those instructions and provides results as output. • The microprocessor contains millions of tiny components like transistors, registers, and diodes that work together. • Modern microprocessors integrate multiple cores, GPUs, NPUs, and memory controllers, enabling applications like AI, gaming, and IoT
- 3. Microprocessor Block Diagram A microprocessor consists of an ALU, control unit and register array. Where ALU performs arithmetic and logical operations on the data received from an input device or memory. Control unit controls the instructions and flow of data within the computer. And, register array consists of registers identified by letters like B, C, D, E, H, L, and accumulator.
- 4. Modern Day Microprocessor Block Diagram
- 5. How Does a Microprocessor Works? • A microprocessor accepts binary data as input, processes that data, and then provides output based on the instructions stored in the memory. The data is processed using the microprocessor's ALU (arithmetical and logical unit), control unit, and a register array. The register array processes the data via a number of registers that act as temporary fast access memory locations. The flow of instructions and data through the system is managed by the control unit.
- 7. RISC Processor • RISC stands for Reduced Instruction Set Computer. • It is designed to reduce the execution time by simplifying the instruction set of the computer. • Using RISC processors, each instruction requires only one clock cycle to execute results in uniform execution time. This reduces the efficiency as there are more lines of code, hence more RAM is needed to store the instructions. • The compiler also has to work more to convert high-level language instructions into machine code. • Example: Power PC: 601, 604, 615, 620, DEC Alpha: 210642, 211066, 21068, 21164, MIPS: TS (R10000) RISC Processor, PA-RISC: HP 7100LC, ARM Cortex- A/M and RISC-V
- 9. Characteristics of RISC • It consists of simple instructions. • It supports various data-type formats. • It utilizes simple addressing modes and fixed length instructions for pipelining. • It supports register to use in any context. • One cycle execution time. • “LOAD” and “STORE” instructions are used to access the memory location. • It consists of larger number of registers. • It consists of a smaller number of transistors.
- 10. Use cases of RISC Processors Domain Example Processors Devices Mobile & Tablets ARM Cortex-A, Apple M1/M2 (based on ARM) Smartphones (iPhone, Samsung), iPads, Android tablets Embedded Systems ARM Cortex-M, RISC-V cores Smartwatches, smart home devices, appliances, drones IoT Devices ESP32, RP2040 (RISC-like) Smart agriculture, wearables, home automation Data Centers ARM Neoverse, Amazon Graviton (ARM) Cloud servers, edge computing Laptops/PCs (Emerging) Apple M-series, ARM-based Windows MacBooks, Surface devices Academia/Research RISC-V Educational kits, open-source research tools
- 11. Advantages and Disadvantages Advantages Disadvantages 1. Each instruction is typically one cycle long and easier to decode. This makes execution faster and simplifies hardware design. 2. Due to uniform instruction length and simple operations, RISC is ideal for deep pipelining — increasing instruction throughput. 3. Simpler hardware logic means less power usage. 4. Less complex control logic and fewer transistors make RISC processors smaller, reducing cost. 5. Uniform and predictable instruction format simplifies the work of compilers in generating efficient machine code. 1. Complex tasks require multiple simple instructions, increasing instruction count and code size. 2. Since more instructions are needed, the total size of the executable can be larger, affecting memory usage. 3. Frequent instruction fetches can stress the memory system, especially without a good cache design. 4. Tasks like string manipulation or floating-point math may require many steps, which CISC can do in one instruction.
- 12. CISC Processor • CISC stands for Complex Instruction Set Computer. • It is designed to minimize the number of instructions per program, ignoring the number of cycles per instruction. The emphasis is on building complex instructions directly into the hardware. • The compiler has to do very little work to translate a high-level language into assembly level language/machine code because the length of the code is relatively short, so very little RAM is required to store the instructions. • Example-IBM 370/168, VAX 11/780, Intel 80486, Intel 8085 and 8086, Intel Core i3/i5/i7/i9, AMD Ryzen etc.
- 14. Characteristics of CISC • Variety of addressing modes. • Larger number of instructions. • Variable length of instruction formats. • Several cycles may be required to execute one instruction. • Instruction-decoding logic is complex. • One instruction is required to support multiple addressing modes.
- 15. Advantages and Disadvantages Advantages Disadvantages 1. Complex instructions can perform multiple low- level operations in one go (e.g., load + add + store). This reduces total instruction count. 2. Since each instruction does more, the compiled program tends to be smaller, which is beneficial when memory is limited. 3. High-level instructions reduce the need to write many low-level instructions manually. 4. Early compilers benefited from CISC's rich instruction set to generate efficient code without optimization. 5. CISC instructions can operate directly on memory without loading into registers first, simplifying programming. 1. Instruction decoder and control unit are more complex, requiring more silicon and power. 2. Complex instructions may take multiple cycles to execute, reducing speed per instruction. 3. Variable-length and multi-step instructions make pipelining difficult and less efficient compared to RISC. 4. Even simple tasks may involve fetching and decoding complex instructions, slowing down performance. 5. More logic and more cycles per instruction can result in higher power use — not ideal for mobile devices.
- 16. CISC vs RISC
- 17. Exercise • List the differences between RISC and CISC
- 18. Evolution of Intel Microprocessors Intel 4004: • The Intel 4004 was released by Intel Corporation in 1971 and was the first commercially available microprocessor. • The 4004 was a 4-bit CPU, designed for use in the Busicom 141-PF printing calculator • The chip, which is clocked at 740 KHz, employs a 10µm process silicon-gate, capable of executing 92,000 instructions per second. • The chip was capable of accessing 4KB of program memory and 640 bytes of RAM. • It employed a 10 µm process silicon-gate enhancement load pMOS technology • The 4004-instruction set consists of only 46 instructions: 41 were 8 bits wide and 5 were 16 bits wide.
- 19. Evolution of Intel Microprocessors Intel 8008: • Introduced in April 1972, The Intel 8008 was the world’s first 8-bit programmable microprocessor. • It featured 50 percent more transistors, eight times the clock speed and was capable of data/character manipulation where the 4004 could only handle arithmetic. • 8008, which was part of the MCS-8, operated at 500 kHZ, had 8-bit data words, and could address 16KB of memory. • 8008 has seven levels of call stack, seven registers, and 48 instructions. • 8008 has seven levels of call stack, seven registers, and 48 instructions.
- 20. Evolution of Intel Microprocessors Intel 8080: • It was released in April of 1974. It is a 8 bit processor • Maximum memory size on the Intel 8080 was increased from 16 KB to 64 KB. • The number of I/O ports was increased to 256. • It is made up using Intel’s N-channel silicon gate MOS process. • the 8080 is largely credited with starting the microcomputer industry. • 2 microseconds clock cycle time • Drawback was that it needed three power supplies.
- 21. Evolution of Intel Microprocessors Intel 8085: • Year of introduction 1975 • 8-bit microprocessor-upgraded • 64 KB main memory • 1.3 microseconds clock cycle time • 246 instructions • uses only one +5v power supply.
- 22. Evolution of Intel Microprocessors Intel 8086/8088: • Year of introduction 1978 for 8086 and 1979 for 8088 • 16-bit microprocessors • Data bus width of 8086 is 16 bit and 8 bit for 8088 • 1 MB main memory • 400 nanoseconds clock cycle time • 6 byte instruction cache for 8086 and 4 byte for 8088 • In 1981 IBM decided to use 8088 in its personal computer
- 23. Evolution of Intel Microprocessors Intel 80186: • Year of introduction 1982 • 16-bit microprocessor-upgraded • 1 MB main memory • Never used in the PC • But was ideal for systems that required a minimum of hardware • Contained special hardware like programmable counters, interrupt controller etc.
- 24. Evolution of Intel Microprocessors Intel 80286: • Year of introduction 1983 • 16-bit high performance microprocessor with memory management & protection • 16 MB main memory • Few additional instructions to handle extra 15 MB • Instruction execution time is as little as 250 ns • Concentrates on the features needed to implement Multitasking
- 25. Evolution of Intel Microprocessors Intel 80386: • Year of introduction 1986 • Intel’s first practical 32-bit microprocessor • 4 GB main memory • Improvements include page handling in virtual environment • Includes hardware circuitry for memory management and memory assignment • Memory paging and enhanced I/O permissions
- 26. Evolution of Intel Microprocessors Intel 80486: • Year of introduction 1989 • 32-bit high performance microprocessor • 4 GB main memory • Incorporates 80387-like floating point coprocessor and 8 K byte cache on one package • About half of the instructions executed in 1 clock instead of 2 on the 80386
- 27. Evolution of Intel Microprocessors Pentium: • Year of introduction 1993 • 32-bit microprocessor, 64-bit data bus and 32-bit address bus • 4 GB main memory • Double clocked 120 and 133MHz versions • Fastest version is the 233MHz, Dual integer processor • 16 KB L1 cache (split instruction and data: 8 KB each)
- 28. Evolution of Intel Microprocessors Pentium Pro: • Year of introduction 1995 • 32-bit microprocessor, formerly code-named P6 • 64 GB main memory, 64-bit data bus and 36-bit address bus • 16 KB L1 cache (split instruction/data: 8 KB each) • Intel launched this processor for the server market • 256 KB L2 cache • Uses three execution engines
- 29. Evolution of Intel Microprocessors Pentium II: • Year of introduction 1997 • 32-bit microprocessor, 64-bit data bus and 36-bit address bus • 64 GB main memory • 32 KB split instruction/data L1 caches (16 KB each) • Module integrated 512KB L2 cache (133MHz) • A version of P2 called Xeon; specifically designed for high-end applications
- 30. Evolution of Intel Microprocessors Pentium III: • Year of introduction 1999 • 32-bit microprocessor, 64-bit data bus and 36-bit address bus • 64 GB main memory • Dual Independent Bus (simultaneous L2 and system memory access) • On-chip 256 KB L2 cache • P3 was available in clock frequencies of up to 1 GHz.
- 31. Evolution of Intel Microprocessors Pentium IV: • Year of introduction 2002 • 32-bit microprocessor, 64-bit data bus and 36-bit address bus • 64 GB main memory • 1.4 to 1.9 GHz and the latest at 3.20 GHz and 3.46GHz (Hyper-Threading) • 1MB/512KB/256KB L2 cache • Specialized for streaming video, game and DVD applications
- 32. New Generation Intel Microprocessors
- 33. Core and Generations Core: • core is a small CPU or processor built into a big CPU or CPU socket • It can independently perform or process all computational tasks. Generations: • The generation of the processor is the first number after i9, i7, i5, or i3. Here are some examples: Intel® Core™ Processor i7- 13700K Processor is 13th generation • Instead of changing the processor altogether Intel prefers to make small changes in their processors to cope up with other up-to-date processors in the market, so they name it as gen 2, gen 3
- 34. Core A processor core is a fundamental unit within a CPU (Central Processing Unit) that executes instructions and performs calculations. It's essentially the "brain" of the computer, and the number of cores a CPU has determines how many tasks it can handle simultaneously. Modern CPUs can have multiple cores, enabling them to execute multiple instructions at the same time, which is crucial for multitasking and running demanding applications. Originally, processors had only one core — they could run only one task (thread) at a time. A multicore processor has two or more independent cores on a single chip. Each core can run its own thread or process in parallel. If a program is single-threaded, it will only use one core no matter how many you have.
- 35. Core Multicore Advantages: • Multicore CPUs can run multiple tasks at once, without switching context as much. • Modern software (e.g., video editors, compilers, 3D games) are written to use multiple threads. More cores = more threads can run at once = faster execution of such applications. • Instead of ramping up one core to a high frequency (which uses more power), you can split the work across multiple cores at lower frequencies, saving power and reducing heat.
- 36. Thread A thread is the smallest unit of execution in a program — basically, it’s a sequence of instructions that the CPU can execute. One program → one or more processes → each process → one or more threads Single-threaded program: Only one task is being executed at a time. Example: A simple calculator app that only processes input when you click. Multi-threaded program: Has multiple threads doing tasks in parallel or semi-parallel. Example: A web browser might have one thread for the UI, one for downloading a page, one for rendering the page, one for playing media.
- 37. Thread • A CPU core can execute one thread at a time (per hardware thread unit). • So, more cores → more threads can run simultaneously. • Operating systems use scheduling to manage how threads are assigned to cores. Hyper-threading / SMT (Simultaneous Multithreading) In Intel CPUs (Hyper-Threading) and AMD (SMT), each physical core can handle 2 threads by interleaving execution. So a 4-core / 8-thread CPU has 4 physical cores, but can run 8 threads simultaneously.
- 38. Why You Can Run More Than Two Programs on a Dual-Core Processor • Modern operating systems (like Windows, Linux, macOS) use a method called time-slicing or context switching. • They quickly switch between tasks, so even if you have only 2 cores, your CPU can give the illusion of running dozens of programs at once. • Your web browser, music player, antivirus, etc., often sit idle while waiting for your input or for data. • So even a small number of cores can keep up with many programs because not all are active at the same time.
- 39. Why You Can Run More Than Two Programs on a Dual-Core Processor • Programs often create multiple threads internally (e.g., one for UI, one for downloading, one for processing). • A dual-core CPU can manage multiple threads by switching between them intelligently. What Does a Dual-Core Limit? It does not limit how many programs can run, It does limit how many things can run truly in parallel, at full performance.
- 40. Clock Speed Clock speed, also called clock frequency, is the rate at which a CPU executes instructions. It's measured in gigahertz (GHz) — which means billions of cycles per second. 3.0 GHz = 3,000,000,000 cycles per second Each cycle allows the processor to perform basic operations like arithmetic, data movement, and decision-making. Higher clock speed means the CPU can process more instructions per second; programs and tasks can finish faster. clock speed is a measure of how fast each core of a CPU works.
- 41. Clock Speed While clock speed is important, it's not the only factor that determines CPU performance. Two CPUs at the same GHz can perform very differently, depending on: Factor Impact CPU architecture Newer designs do more per cycle (better instructions, shorter pipelines) Core count More cores = more tasks in parallel (multitasking, video rendering, etc.) Cache size More cache = faster access to frequently used data Thermal & power limits CPUs may reduce clock under high heat (thermal throttling) Instruction per cycle (IPC) Some CPUs can do more in a single cycle
- 42. Cache Cache is a small, ultra-fast memory inside or very close to the CPU. It stores frequently used data and instructions so the CPU can access them much faster than from RAM. Cache Level Size (approx.) Speed Location Purpose L1 Cache 16KB–128KB Fastest Inside each core Stores the most frequently accessed data and instructions L2 Cache 256KB–1MB Very fast Per core Slightly larger and slower than L1, stores less-used data L3 Cache 4MB–64MB Slower (but still faster than RAM) Shared among all cores Stores common data across cores
- 43. Turbo Boost • Intel Turbo Boost is a performance-enhancing feature that allows a CPU to automatically run faster than its base clock speed when needed — as long as the system stays within safe power, temperature, and current limits. • Every Intel processor has a base clock speed (e.g., 2.5 GHz) and a maximum Turbo Boost frequency (e.g., 4.2 GHz). • When you run a demanding task — like gaming, video editing, or large calculations — Turbo Boost kicks in and increases the core frequency to improve performance. • It only happens if the CPU is not too hot and has power headroom.
- 44. Comparison Between i3, i5, i7 and i9 processors Feature / Aspect Intel Core i3 Intel Core i5 Intel Core i7 Intel Core i9 Target Use Budget-friendly, basic tasks Mainstream, general use High performance, gaming, productivity Enthusiast-level, heavy workloads, gaming, content creation Core Count Usually 2–4 cores Usually 4–6 cores Usually 6–8 cores Usually 8–16 cores (varies by generation) Threads 4 threads (with Hyper-Threading) 4–12 threads (with Hyper-Threading) 8–16 threads (with Hyper-Threading) 16+ threads (with Hyper-Threading) Clock Speed Base to moderate turbo speeds Moderate to high turbo speeds High turbo speeds Very high turbo speeds Cache Size Smaller (3–6 MB) Medium (6–12 MB) Larger (12–16 MB) Largest (16–24+ MB) Hyper-Threading Usually enabled Often enabled (varies) Usually enabled Always enabled Integrated Graphics Basic integrated graphics Better integrated graphics Better integrated graphics Some models have advanced graphics or none (discrete GPU expected) Typical Price Range Lowest Mid-range Higher-end Premium / highest-end Use Cases Office apps, web browsing, light multitasking Gaming, multimedia, moderate multitasking Heavy gaming, video editing, software development Extreme gaming, 3D rendering, heavy multitasking, professional workloads
- 45. Intel’s Latest Processors 14th Gen Intel Core Processors (Raptor Lake Refresh): • Core i9, i7, and i5: These processors continue the performance hybrid architecture of previous generations, offering a balance of performance and efficiency for gaming, content creation, and productivity. • Desktop Processors: These are designed to deliver an immersive experience for gaming and other demanding tasks. • Raptor Lake Refresh: This is the last generation to use the old "Core i" branding scheme.
- 46. Intel’s Latest Processors Intel Core Ultra Processors (Meteor Lake): Intel Core Ultra processors are designed for both laptops and desktops, with different suffixes indicating their intended use (e.g., H for high-performance laptops, U for ultra-mobile laptops, K, F, KF, or T for desktops). • New Branding: These processors introduce a new "Core 3/5/7" branding scheme, replacing the "Core i" branding. • Mobile and Desktop: Intel Core Ultra processors are available in both mobile and desktop versions. • SKU Number: The SKU number (1 or 2) indicates the generation. • Suffixes: Mobile processors have suffixes like H or U, while desktop processors have suffixes like K, F, KF, or T. • AI Acceleration: These processors include AI acceleration capabilities.
- 47. AMD Processors AMD (Advanced Micro Devices) is a major semiconductor company known for designing and manufacturing computer processors. Key Features of AMD Processors • Zen Microarchitecture: Provides a good balance between performance and power efficiency. • Multi-core and multi-threading capabilities: AMD offers more cores and threads in many product tiers compared to Intel counterparts at similar prices. • Chiplet design: Uses multiple smaller chiplets interconnected to scale performance and manufacturing flexibility (used in Ryzen and EPYC). • Unlocked processors: Most Ryzen CPUs support overclocking, allowing users to boost performance. • Support for fast memory: DDR4 and DDR5 support, depending on generation. • Strong performance in multi-threaded applications: Such as video editing, 3D modeling, scientific simulations. • Competitive pricing and value: AMD often provides better cost-to-performance ratios.
- 48. AMD Processor Families 1. AMD Ryzen (Mainstream Desktop and Laptop CPUs) • Launched in 2017, Ryzen is AMD's flagship CPU brand for desktops, laptops, and all-purpose computing. • Based on the Zen microarchitecture (Zen, Zen+, Zen 2, Zen 3, Zen 4), Ryzen processors emphasize a balance of strong multi-core performance, energy efficiency, and competitive pricing. • Ryzen CPUs come in different tiers to suit various users: • Ryzen 3: Entry-level, budget-friendly • Ryzen 5: Mid-range, balanced performance • Ryzen 7: High-performance, multitasking and gaming • Ryzen 9: Enthusiast level, for heavy workloads and content creation • Ryzen also has “G” variants with integrated graphics.
- 49. AMD Processor Families 2. AMD Threadripper (High-End Desktop Workstations) • Designed for professionals and enthusiasts needing extreme multi- core performance. • Features many cores (up to 64+), massive cache, and high memory bandwidth. • Ideal for 3D rendering, video production, scientific computing.
- 50. AMD Processor Families 3. AMD EPYC (Server and Data Center CPUs) • Enterprise-grade processors optimized for servers and data centers. • High core counts, large cache sizes, advanced security features, and excellent multi-threading capabilities. 4. AMD Athlon and A-Series (Budget and Legacy CPUs) • Earlier budget and entry-level CPUs before Ryzen era. • Still used in some low-cost or legacy systems.
- 51. AI Accelerators AI accelerators are specialized processors that work alongside or inside the main CPU, to handle AI-specific computations faster and more efficiently. Component Role CPU General-purpose processor that handles all types of instructions. AI Accelerator Specialized hardware focused on executing AI operations (like matrix multiplications, convolutions, etc.) much faster. Integration Can be separate chips, on-chip modules, or on the same die as CPU (like in Apple M1/M2 or ARM SoCs).
- 52. AI Accelerators In a typical AI system (like a laptop, phone, or server): • The CPU acts like the manager: loading the AI model, setting up the data, assigning the task. • The AI accelerator (GPU/NPU/TPU/etc.) acts like the worker: executing the heavy matrix and neural network computations. • The result is sent back to the CPU. This offloading frees the CPU to do other tasks and makes the AI operation much faster and more power-efficient.