export_week 10 chapter 8 cpu and memory design enhancement and implementation

Question 1

; tab ,

▪ Current CPU Architecture Designs:

Answer 1

▪ Traditional modern architectures

▪ VLIW (Transmeta) – Very Long Instruction Word

▪ EPIC (Intel) – Explicitly Parallel Instruction Computer

Question 2

; tab ,

Current CPU Architectures:

Answer 2

* IBM Mainframe series * Intel x86 family * IBM POWER/PowerPC family * Sun SPARC family

Question 3

; tab ,

Problems with early CPU Architectures and solutions:

Answer 3

▪ Large number of specialized instructions were rarely used but added hardware complexity and slowed down other instructions ▪ Slow data memory accesses could be reduced by increasing the number of general purpose registers ▪ Using general registers to hold addresses could reduce the number of addressing modes and simplify architecture design

▪ Fixed-length, fixed format instruction words would allow instructions to be fetched and decoded independently and in parallel

Question 4

; tab ,

how VLIW Architecture?

Answer 4

▪ Transmeta Crusoe CPU ▪ 128-bit instruction bundle = molecule ▪ Four 32-bit atoms (atom = instruction) ▪ Parallel processing of 4 instructions ▪ 64 general purpose registers ▪ Code morphing layer ▪ Translates instructions written for other CPUs into molecules ▪ Instructions are not written directly for the Crusoe CPU

Question 5

; tab ,

EPIC Architecture

Answer 5

▪ 128-bit instruction bundle ▪ 3 41-bit instructions ▪ 5 bits to identify type of instructions in bundle * 128 64-bit general purpose registers * 128 82-bit floating point registers * Intel X86 instruction set included * Programmers and compilers follow guidelines to ensure parallel execution of instructions

Question 6

; tab ,

Fetch-Execute Cycle Timing Issues?

Answer 6

▪ Computer clock is used for timing purposes for each step of the instruction cycle ▪ GHz (gighertz) – billion steps per second ▪ Instructions can (and often) take more than one step ▪ Data word width can require multiple steps

Question 7

; tab ,

CPU Features and Enhancements

Answer 7

Separate Fetch/Execute Units Pipelining Multiple, Parallel Execution Units Scalar Processing Superscalar Processing Branch Instruction Processing

Question 8

; tab ,

what include fetch unit ?

Answer 8

▪ Instruction fetch unit ▪ Instruction decode unit Determine opcode Identify type of instruction and operands ▪ Several instructions are fetched in parallel and held in a buffer until decoded and executed ▪ IP – Instruction Pointer register holds instruction location of current being processed

Question 9

; tab ,

what is including in the Execute Unit?

Answer 9

▪ Receives instructions from the decode unit ▪ Appropriate execution unit services the instruction

Question 10

; tab ,

define Instruction Pipelining ?

Answer 10

▪ Assembly-line technique to allow overlapping between fetch-execute cycles of sequences of instructions

Question 11

; tab ,

define Scalar processing ?

Answer 11

Average instruction execution is approximately equal to the clock speed of the CPU

Question 12

; tab ,

Branch Problem Solutions

Answer 12

▪ Separate pipelines for both possibilities ▪ Probabilistic approach ▪ Requiring the following instruction to not be dependent on the branch ▪ Instruction Reordering (superscalar processing)

Question 13

; tab ,

Multiple, Parallel Execution Units what is it ?

Answer 13

▪ Different instructions have different numbers of steps in their cycle ▪ Differences in each step ▪ Each execution unit is optimized for one general type of instruction ▪ Multiple execution units permit simultaneous execution of several instructions

Question 14

; tab ,

talk about Superscalar Processing?

Answer 14

▪ Process more than one instruction per clock cycle

▪ Separate fetch and execute cycles as much as possible

▪ Buffers for fetch and decode phases

▪ Parallel execution units

Question 15

; tab ,

talk about Superscalar Issues

Answer 15

▪ Out-of-order processing – dependencies (hazards)

▪ Data dependencies

▪ Branch (flow) dependencies and speculative execution

▪ Parallel speculative execution or branch prediction

▪ Branch History Table

▪ Register access conflicts

▪ Rename or logical registers

Question 16

; tab ,

why need Memory Enhancementsand what are they ?

Answer 16

Memory is slow compared to CPU processing speeds!
▪ 2Ghz CPU = 1 cycle in ½ of a billionth of a second
▪ 70ns DRAM = 1 access in 70 millionth of a second
▪ Methods to improvement memory accesses:
▪ Wide Path Memory Access
• Retrieve multiple bytes instead of 1 byte at a time
▪ Memory Interleaving
• Partition memory into subsections, each with its own address register and data register
▪ Cache Memory

Question 17

; tab ,

Cache Memory

Answer 17

▪ Blocks: 8 or 16 bytes ▪ Tags: pointer to location in main memory ▪ Cache controller ▪ hardware that checks tags ▪ Cache Line ▪ Unit of transfer between storage and cache memory ▪ Hit Ratio: ratio of hits out of total requests ▪ Synchronizing cache and memory ▪ Write through ▪ Write back

Question 18

; tab ,

Step-by-Step Use of Cache1

Answer 18

Question 19

; tab ,

Performance Advantages of cache memory?

Answer 19

▪ Hit ratios of 90% common ▪ 50%+ improved execution speed ▪ Locality of reference is why caching works ▪ Most memory references confined to small region of memory at any given time ▪ Well-written program in small loop, procedure or function ▪ Data likely in array ▪ Variables stored togeth

Question 20

; tab ,

Why do the sizes of the caches have to be different?

Question 21

; tab ,

reasons for Multiprocessing?

Answer 21

▪ Reasons ▪ Increase the processing power of a system ▪ Parallel processing

Question 22

; tab ,

Multiprocessor system in Multiprocessing is ?

Answer 22

▪ Tightly coupled ▪ Multicore processors - when CPUs are on a single integrated circuit

Question 23

; tab ,

what is Multiprocessor Systems for ?

Answer 23

▪ Identical access to programs, data, shared memory, I/O, etc. ▪ Easily extends multi-tasking, and redundant program executio

Question 24

; tab ,

▪ Two ways to configure Multiprocessor Systems

Answer 24

▪ Two ways to configure ▪ Master-slave multiprocessing ▪ Symmetrical multiprocessing (SMP)

Question 25

; tab ,

Master-Slave Multiprocessing, Master CPU?

Answer 25

▪ Manages the system ▪ Controls all resources and scheduling ▪ Assigns tasks to slave CPUs

Question 26

; tab ,

aAdvantages of Master-Slave Multiprocessing?

Answer 26

Advantages ▪ Simplicity ▪ Protection of system and data

Question 27

; tab ,

▪ Disadvantages of Master-Slave Multiprocessing?

Answer 27

▪ Disadvantages ▪ Master CPU becomes a bottleneck ▪ Reliability issues – if master CPU fails entire system fails

Question 28

; tab ,

Symmetrical Multiprocessing

Answer 28

▪ Each CPU has equal access to resources ▪ Each CPU determines what to run using a standard algorithm

Question 29

; tab ,

▪ Disadvantages of Symmetrical Multiprocessing

Answer 29

▪ Resource conflicts – memory, i/o, etc. ▪ Complex implementation

Question 30

; tab ,

Advantages Symmetrical Multiprocessing ?

Answer 30

▪ High reliability ▪ Fault tolerant support is straightforward ▪ Balanced workload

Question 31

; tab ,

General Enhancements – Use RISC-based techniques

Answer 31

– Fewer instruction formats, fixed-length → faster decoding – More general purpose registers → fewer memory accesses

Question 32

; tab ,

Clock cycle and instruction cycle

Answer 32

– Most instructions take several clock cycles to execute: Fetch the new instruction [IF]. Decode the instruction [ID]. Execute the instruction [EX]. Access memory (if needed) [MEM]. Write back to the registers [WB]

Question 33

; tab ,

Each stage takes a clock cycle, so complete execution takes 5 cycles. Can we do better?

Answer 33

Waiting for all five stages of instruction execution to complete is like building something from start to finish

Question 34

; tab ,

Clock cycle and instruction cycle Or can the CPU overlap the execution of several instructions at once because they’re all similar?

Answer 34

https://s3.amazonaws.com/classconnection/655/flashcards/7082655/png/selection_221-14A61CABCA1789F393E.png

Question 35

; tab ,

– Five stages of instruction execution

Question 36

; tab ,

Five stages of instruction execution

Question 37

; tab ,

Clock cycle and instruction cycle

Question 38

; tab ,

– Problems with pipelining

Answer 38

Dependencies (register interlock)—if an instruction needs a result from the immediately preceding instruction, that result won’t be written back until WB, but the result is needed in EX.

Question 39

; tab ,

Problems with pipelining

Answer 39

– Branching—when the instruction being executed is a branch, we can’t know if the branch will be taken until after stage 3. But by that time, other instructions are “in flight.”

Question 40

; tab ,

Clock cycle and instruction cycle

Question 41

; tab ,

Superscalar Processing

Answer 41

RISC and pipelining lets each functional unit in a CPU be fully utilized all of the time. – But, what if there were multiple ALUs or multiple decoders? Then multiple instructions could be executed at once. – Prerequisite: Multiple instructions should be fetched at once via a large path to memory.

Question 42

; tab ,

Superscalar Processing D

Question 43

; tab ,

Superscalar Processing DD

Question 44

; tab ,

– Problems with superscalar processing

Answer 44

– Same general categories as with pipelining: dependencies and branches – Except now forwards, stalls, or canceling may need to be between several functional units! – CPUs become very complex again, yet it is common to have 2 to 4 separate pipelines per core in modern processors.

Question 45

; tab ,

Did you know 1

Answer 45

RISC-based CPUs offer general performance enhancements due to simplified formats and single-clock cycle execution.

Question 46

; tab ,

Pipelining allows…..

Answer 46

multiple instructions to be in various stages of execution at once.

Question 47

; tab ,

Superscalar processing duplicates……

Answer 47

pipelines in a single core to have multiple instructions executing simultaneously.

Question 48

; tab ,

Data dependencies and branches are……?

Answer 48

hazards to both pipelining and superscalar architectures.

Question 49

; tab ,

Recall CPU Pipelining

Question 50

; tab ,

Three complementary approaches memory ?

Answer 50

All three are used simultaneously in the system design

Question 51

; tab ,

Wide path memory access

Question 52

; tab ,

Wide path memory access 1

Question 53

; tab ,

Wide path memory access 2

Question 54

; tab ,

Wide path memory access

Answer 54

https://s3.amazonaws.com/classconnection/655/flashcards/7082655/png/selection_234-14A61E290734D580E81.png

Question 55

; tab ,

Memory interleaving

Answer 55

https://s3.amazonaws.com/classconnection/655/flashcards/7082655/png/selection_235-14A61E356D54EB3628B.png

Question 56

; tab ,

Memory interleaving D

Answer 56

https://s3.amazonaws.com/classconnection/655/flashcards/7082655/png/imagesgm2qx-14A61E3B32D3AF72413.png

Question 57

; tab ,

Cache Memory

Answer 57

Use a small amount of expensive SRAM as a buffer against the large amount of DRAM

Question 58

; tab ,

Cache Memory

Answer 58

https://s3.amazonaws.com/classconnection/655/flashcards/7082655/png/imagekfh3qx-14A61E51C745DDD5B10.png

Question 59

; tab ,

Cache Memory

Answer 59

https://s3.amazonaws.com/classconnection/655/flashcards/7082655/png/selection_237-14A61E5ABA96F16AC1E.png

Question 60

; tab ,

cache entries consists of ?

Answer 60

https://s3.amazonaws.com/classconnection/655/flashcards/7082655/png/selection_238-14A61E64D2943E88C7F.png

Question 61

; tab ,

cache replacement algorithm ?

Answer 61

https://s3.amazonaws.com/classconnection/655/flashcards/7082655/png/selection_239-14A61E6D6156AC4FB57.png

Question 62

; tab ,

cache memory what should happen on memory write ?

Answer 62

Cache coherency gets particularly tricky with multiple cores and multiple levels of cache.

Question 63

; tab ,

DID you know

Answer 63

https://s3.amazonaws.com/classconnection/655/flashcards/7082655/png/selection_241-14A61E86E6900A46784.png