RISC vs CISC - Is it Still a Thing?

Summary

In this video, Gary Sims delves into the ongoing debate between Complex Instruction Set Computing (CISC) and Reduced Instruction Set Computing (RISC), exploring their historical context and impact on modern processors. He explains the origins of both architectures, focusing on the simplicity of RISC in contrast to the complexity of CISC, and discusses how RISC processors, like ARM and RISC-V, dominate smartphones and tablets, while CISC remains prevalent in desktops with x86. Gary also highlights how processor design has evolved, showing how modern CPUs, including x86 processors, now incorporate RISC-like elements through micro-operations to improve performance and efficiency.

Takeaways

• 😀 RISC (Reduced Instruction Set Computing) and CISC (Complex Instruction Set Computing) are two contrasting processor design philosophies, with RISC focusing on simpler instructions and CISC on more complex ones.
• 😀 RISC aims to execute one instruction per cycle, making it faster and more efficient for certain tasks compared to CISC.
• 😀 The main benefit of RISC processors is their simpler design, allowing for cheaper and quicker CPU manufacturing.
• 😀 In RISC processors, instructions are of fixed length, simplifying decoding and execution, while CISC processors use variable-length instructions.
• 😀 Unlike CISC, RISC does not allow operations directly on memory; instead, multiple instructions are required to fetch, modify, and store data in memory.
• 😀 Compiler optimization plays a significant role in maximizing the performance of RISC processors, as it generates efficient instruction sequences.
• 😀 Pipeline execution is crucial in RISC design, where instructions are processed in stages for greater efficiency and speed.
• 😀 A delayed jump is a feature of RISC processors where the CPU executes a no-operation instruction after a jump to prevent pipeline issues.
• 😀 Although RISC processors became dominant in mobile devices, the x86 architecture (CISC) still dominates in desktop PCs and servers.
• 😀 Modern processors, including those used in desktops, smartphones, and AI applications, have embraced RISC principles through micro-operations, effectively blending CISC and RISC concepts.

Q & A

What is the difference between CISC and RISC architectures?

CISC (Complex Instruction Set Computing) uses complex instructions that can perform multiple tasks in one cycle, while RISC (Reduced Instruction Set Computing) simplifies instructions to execute one per cycle, aiming for faster and more efficient processing.

Why did RISC architecture emerge in the 1980s?

RISC emerged in the 1980s to address the growing complexity of processors. It proposed simplifying instructions to reduce the design complexity of CPUs and to make them more efficient, faster, and cheaper to manufacture.

How does the number of transistors impact processor performance?

The number of transistors affects a processor's performance by allowing more complex designs, better optimization, and more functions, such as more stages in a pipeline, better branch prediction, and more efficient execution. Modern processors have billions of transistors, which enhance performance compared to earlier processors with only a million transistors.

What is the significance of fixed instruction sizes in RISC?

RISC uses fixed instruction sizes, making it easier to decode instructions and process them in a pipeline. This uniformity reduces the complexity of instruction fetching and decoding, resulting in faster execution.

Why do RISC processors require more instructions compared to CISC for certain operations?

RISC processors do not perform operations directly on memory. For example, to add one to a number stored in memory, RISC requires separate instructions to fetch the data, perform the addition, and store the result. In contrast, CISC can combine these operations into a single instruction.

What is a delayed jump in RISC processors?

A delayed jump is a technique used in RISC processors to optimize pipeline execution. When a jump instruction is executed, the processor continues to execute the next instruction that was already queued in the pipeline before the jump, avoiding a pipeline stall.

How do modern processors integrate both RISC and CISC techniques?

Modern processors, such as those from Intel, combine elements of both RISC and CISC. For instance, CISC processors like x86 use complex instructions that are broken down into simpler RISC-like operations (micro-operations) for processing, allowing for more efficient execution.

Why is RISC preferred in mobile devices?

RISC is preferred in mobile devices due to its efficiency and lower power consumption. RISC processors, like ARM, use simpler instructions that require fewer transistors and generate less heat, which is ideal for battery-powered devices.

How has branch prediction evolved in modern processors?

Branch prediction has evolved to become a standard feature in modern processors. It allows the processor to predict which direction a branch instruction will take, ensuring that instructions can continue flowing smoothly down the pipeline without delays.

What role does the compiler play in the performance of RISC processors?

The compiler plays a critical role in optimizing programs for RISC processors. Since RISC processors rely on simple instructions, the compiler must efficiently translate high-level code into a sequence of instructions that maximizes the use of the processor's capabilities.