CS136, Advanced Architecture

1. CS136, Advanced Architecture Limits to ILP Simultaneous Multithreading

2. Outline • Limits to ILP (another perspective) • Thread Level Parallelism • Multithreading • Simultaneous Multithreading • Power 4 vs. Power 5 • Head to Head: VLIW vs. Superscalar vs. SMT • Commentary • Conclusion CS136

3. Limits to ILP • Conflicting studies of amount • Benchmarks (vectorized Fortran FP vs. integer C programs) • Hardware sophistication • Compiler sophistication • How much ILP is available using existing mechanisms with increasing HW budgets? • Do we need to invent new HW/SW mechanisms to keep on processor performance curve? • Intel MMX, SSE (Streaming SIMD Extensions): 64 bit ints • Intel SSE2: 128 bit, including 2 64-bit Fl. Pt. per clock • Motorola AltiVec: 128 bit ints and FPs • Supersparc Multimedia ops, etc. CS136

4. Overcoming Limits • Advances in compiler technology + significantly new and different hardware techniques may be able to overcome limitations assumed in studies • However, unlikely such advances when coupledwith realistic hardware will overcome these limits in near future CS136

5. Limits to ILP Initial HW Model here; MIPS compilers. Assumptions for ideal/perfect machine to start: 1. Register renaming – infinite virtual registers => all register WAW & WAR hazards are avoided 2. Branch prediction – perfect; no mispredictions 3. Jump prediction – all jumps perfectly predicted (returns, case statements)2 & 3  no control dependencies; perfect speculation & an unbounded buffer of instructions available 4. Memory-address alias analysis – addresses known & a load can be moved before a store provided addresses not equal; 1&4 eliminates all but RAW Also: perfect caches; 1 cycle latency for all instructions (FP *,/); unlimited instructions issued/clock cycle CS136

6. Limits to ILP HW Model comparison CS136

7. Upper Limit to ILP: Ideal Machine(Figure 3.1) FP: 75 - 150 Integer: 18 - 60 Instructions Per Clock CS136

8. Limits to ILP HW Model comparison CS136

9. More Realistic HW: Window ImpactFigure 3.2 Change from Infinite window 2048, 512, 128, 32 FP: 9 - 150 Integer: 8 - 63 IPC CS136

10. Limits to ILP HW Model comparison CS136

11. More Realistic HW: Branch ImpactFigure 3.3 FP: 15 - 45 Change from Infinite window to 2048, and maximum issue of 64 instructions per clock cycle Integer: 6 - 12 IPC CS136 Perfect Tournament BHT (512) Profile No prediction

12. Misprediction Rates CS136

13. Limits to ILP HW Model comparison CS136

14. More Realistic HW: Renaming Register Impact (N int + N fp) Figure 3.5 FP: 11 - 45 Change to 2048 instr window, 64 instr issue, 8K 2 level Prediction IPC Integer: 5 - 15 Infinite 256 128 64 32 None CS136

15. Limits to ILP HW Model comparison CS136

16. More Realistic HW: Memory Address Alias ImpactFigure 3.6 Change 2048 instr window, 64 instr issue, 8K 2 level Prediction, 256 renaming registers FP: 4 - 45 (Fortran, no heap) IPC Integer: 4 - 9 Perfect Global/Stack perf;heap conflicts Compiler Inspection None CS136

17. Limits to ILP HW Model comparison CS136

18. Perfect disambiguation (HW), 1K Selective Prediction, 16 entry return, 64 registers, issue as many as window Realistic HW: Window Impact(Figure 3.7) FP: 8 - 45 IPC Integer: 6 - 12 Infinite 256 128 64 32 16 8 4 CS136

19. Outline • Limits to ILP (another perspective) • Thread Level Parallelism • Multithreading • Simultaneous Multithreading • Power 4 vs. Power 5 • Head to Head: VLIW vs. Superscalar vs. SMT • Commentary • Conclusion CS136

20. How to Exceed ILP Limitsof This Study? • These are not laws of physics • Just practical limits for today • Could be overcome via research • Compiler and ISA advances could change results • WAR and WAW hazards through memory: eliminated WAW and WAR hazards through register renaming, but not in memory usage • Can get conflicts via allocation of stack frames • Because called procedure reuses memory addresses of previous stack frames CS136

21. HW v. SW to increase ILP • Memory disambiguation: HW best • Speculation: • HW best when dynamic branch prediction better than compile-time prediction • Exceptions easier for HW • HW doesn’t need bookkeeping code or compensation code • Very complicated to get right in SW • Scheduling: SW can look ahead to schedule better • Compiler independence: HW does not require new compiler to run well CS136

22. Performance Beyond Single-Thread ILP • Much higher natural parallelism in some applications • Database or scientific codes • Explicit thread-levelor data-level parallelism • Thread: has own instructions and data • May be part of parallel program or independent program • Each thread has all state (instructions, data, PC, register state, and so on) needed to execute • Data-level parallelism: Perform identical operations on lots of data CS136

23. Thread Level Parallelism (TLP) • ILP exploits implicit parallel operations within loop or straight-line code segment • TLP explicitly represented by multiple threads of execution that are inherently parallel • Goal: Use multiple instruction streams to improve • Throughput of computers that run many programs • Execution time of multi-threaded programs • TLP could be more cost-effective to exploit than ILP CS136

24. Do Both ILP and TLP? • TLP and ILP exploit two different kinds of parallel structure in a program • Could a processor oriented to ILP still exploit TLP? • Functional units are often idle in data path designed for ILP because of either stalls or dependencies in the code • Could TLP be used as source of independent instructions that might keep the processor busy during stalls? • Could TLP be used to employ functional units that would otherwise lie idle when insufficient ILP exists? CS136

25. New Approach:Multithreaded Execution • Multithreading: multiple threads share functional units of 1 processor via overlapping • Processor must duplicate independent state of each thread • Separate copy of register file, PC • Separate page table if different process • Memory sharing via virtual memory mechanisms • Already supports multiple processes • HW for fast thread switch • Must be much faster than full process switch (which is 100s to 1000s of clocks) • When to switch? • Alternate instruction per thread (fine grain)—round robin • When thread is stalled (coarse grain) • E.g., cache miss CS136

26. Fine-Grained Multithreading • Switches between threads on each instruction, interleaving execution of multiple threads • Usually done round-robin, skipping stalled threads • CPU must be able to switch threads every clock • Advantage: can hide both short and long stalls • Instructions from other threads always available to execute • Easy to insert on short stalls • Disadvantage: slows individual threads • Thread ready to execute without stalls will be delayed by instructions from other threads • Used on Sun’s Niagara (will see later) CS136

27. Course-Grained Multithreading • Switches threads only on costly stalls • E.g., L2 cache misses • Advantages • Relieves need to have very fast thread switching • Doesn’t slow thread • Other threads only issue instructions when main one would stall (for long time) anyway • Disadvantage: pipeline startup costs make it hard to hide throughput losses from shorter stalls • Pipeline must be emptied or frozen on stall, since CPU issues instructions from only one thread • New thread must fill pipe before instructions can complete • Thus, better for reducing penalty of high-cost stalls where pipeline refill << stall time • Used in IBM AS/400 CS136

28. Simultaneous Multithreading (SMT) • Simultaneous multithreading (SMT): insight that dynamically scheduled processor already has many HW mechanisms to support multithreading • Large set of virtual registers that can be used to hold register sets for independent threads • Register renaming provides unique register identifiers • Instructions from multiple threads can be mixed in data path • Without confusing sources and destinations across threads! • Out-of-order completion allows the threads to execute out of order, and get better utilization of the HW • Just add per-thread renaming table and keep separate PCs • Independent commitment can be supported via separate reorder buffer for each thread Source: Micrprocessor Report, December 6, 1999 “Compaq Chooses SMT for Alpha” CS136

29. Simultaneous Multithreading ... One thread, 8 units Two threads, 8 units Cycle M M FX FX FP FP BR CC Cycle M M FX FX FP FP BR CC M = Load/Store, FX = Fixed Point, FP = Floating Point, BR = Branch, CC = Condition Codes CS136

30. Multithreaded Categories SMT Fine-Gr. Coarse-Gr. Superscalar Multiprocessing Time (processor cycle) Thread 1 Thread 3 Thread 5 Thread 2 Thread 4 Idle slot CS136

31. Design Challenges in SMT • What is impact on single-thread performance? • Preferred-thread approach • Sacrifices neither throughput nor single-thread performance? • Nope: processor will sacrifice some throughput when preferred thread stalls • Larger register file needed to hold multiple contexts • Must not affect clock cycle, especially in: • Instruction issue—more candidate instructions to consider • Instruction completion—hard to choose which to commit • Must ensure that cache and TLB conflicts caused by SMT don’t degrade performance CS136

32. Digression: Covert Channels • Imagine you’re spy with account on Knuth • Want to communicate a secret to Geoff • Secret is reasonably small • FBI is watching your account and your e-mail • Solution: process spawning • Once a second, Geoff spawns process • Records own PID, waits 10 ms, forks & records child PID • Once a second, you send one bit of information • If bit is zero, you do nothing • If bit is one, you spawn processes as fast as possible • If Geoff sees big PID gap, he records “1”, else “0” • Many variations on this basic idea CS136

33. Covert-Channel Attacks on Crypto • Most (not all) crypto code behaves differently on “1” bit in key vs. “0” bit • Runs longer or shorter • Uses more or less power • Accesses different memory • Etc. • Usually called “information leakage” • Has been successfully used in lab to crack strong crypto • Even recovering some bits from key makes brute-force attack practical for getting remainder • Some modern implementations try to fight by doing wasted work on shorter path of “if”, etc. CS136

34. SMT Attack on SSH • On SMT machine, lower-priority thread’s execution rate depends on higher-priority one’s instructions • More stalls in top thread mean more speed in bottom one • Stalls vary depending on what crypto code is doing • Operates at very low level • Thus much harder to defend against • Successful attack on ssh keys has been demonstrated in lab • Best known defense: don’t do SMT • Careful coding of crypto could probably also work • Note that this also applies to things like cache and TLB • Lots of ways to leak information unintentionally! CS136

35. Power 4 Single-threaded predecessor to Power 5. 8 execution units in out-of-order engine; each can issue instruction each cycle. CS136

36. Power 4 2 commits (architected register sets) Power 5 2 fetch (PC),2 initial decodes CS136

37. Power 5 data flow ... Why only 2 threads? With 4, one of the shared resources (physical registers, cache, memory bandwidth) would be prone to bottleneck CS136

38. Power 5 thread performance ... Relative priority of each thread controllable in hardware. For balanced operation, both threads run slower than if they “owned” the machine. CS136

39. Changes in Power 5 to support SMT • Increased associativity of L1 instruction cache and instruction address translation buffers • Added per-thread load and store queues • Increased size of L2 (1.92 vs. 1.44 MB) and L3 caches • Added separate instruction prefetch and buffering per thread • Increased virtual registers from 152 to 240 • Increased size of several issue queues • Power5 core is about 24% larger than Power4 because of SMT support CS136

40. Initial Performance of SMT • Pentium 4 Extreme SMT yields 1.01 speedup for SPECint_rate benchmark; 1.07 for SPECfp_rate • Pentium 4 is dual-threaded SMT • SPECRate requires each benchmark to be run against vendor-selected number of copies of same benchmark • Pairing each of 26 SPEC benchmarks with every other on Pentium 4 (262 runs) gives speedups from 0.90 to 1.58; average was 1.20 • 8-processor Power 5 server 1.23 faster for SPECint_rate w/ SMT, 1.16 faster for SPECfp_rate • Power 5 running 2 copies of each app had speedup between 0.89 and 1.41 • Most gained some • Floating-point apps had most cache conflicts and least gains CS136

41. Head-to-Head ILP Competition CS136

42. Performance on SPECint2000 CS136

43. Performance on SPECfp2000 CS136

44. Normalized Performance: Efficiency CS136

45. No Silver Bullet for ILP • No obvious overall leader in performance • AMD Athlon leads on SPECInt performance, followed by the Pentium 4, Itanium 2, and Power5 • Itanium 2 and Power5 clearly dominate Athlon and Pentium 4 on SPECFP • Itanium 2 is most inefficient processor both for floating-point and integer code for all but one efficiency measure (SPECFP/Watt) • Athlon and Pentium 4 both use transistors and area efficiently • IBM Power5 is most effective user of energy on SPECFP, essentially tied on SPECINT CS136

46. Limits to ILP • Doubling issue rates above today’s 3-6 instructions per clock probably requires processor to: • Issue 3-4 data-memory accesses per cycle, • Resolve 2-3 branches per cycle, • Rename and access over 20 registers per cycle, and • Fetch 12-24 instructions per cycle. • Complexity of implementing these capabilities is likely to mean sacrifices in maximum clock rate • E.g, widest-issue processor is Itanium 2 • It also has slowest clock rate • Despite consuming the most power! CS136

47. Limits to ILP (cont’d) • Most ways to increase performance also boost power consumption • Key question is energy efficiency: does a method increase power consumption faster than it boosts performance? • Multiple-issue techniques are energy inefficient: • Incurs logic overhead that grows faster than issue rate • Growing gap between peak issue rates and sustained performance • Number of transistors switching = f(peak issue rate); performance = f(sustained rate); growing gap between peak and sustained performance  Increasing energy per unit of performance CS136

48. Commentary • Itanium is not significant breakthrough in scaling ILP or in avoiding problems of complexity and power consumption • Instead of pursuing more ILP, architects turning to TLP using single-chip multiprocessors • In 2000, IBM announced Power4, 1st commercial single-chip, general-purpose multiprocessor: has two Power3 processors and integrated L2 cache • Sun Microsystems, AMD, and Intel have also switched focus from aggressive uniprocessors to single-chip multiprocessors • Right balance of ILP and TLP is unclear today • Maybe desktops (mostly single-threaded?) need different design than servers (can do lots of TLP) CS136

49. And in conclusion … • Limits to ILP (power efficiency, compilers, dependencies …) seem to limit to 3 to 6 issue for practical options • Explicitly parallel (data-level parallelism or thread-level parallelism) is next step to performance • Coarse-grained vs. fine-grained multithreading • Only on big stall vs. every clock cycle • Simultaneous multithreading is fine-grained multithreading based on OOO superscalar microarchitecture • Instead of replicating registers, reuse rename registers • Itanium/EPIC/VLIW is not a breakthrough in ILP • Balance of ILP and TLP decided in marketplace CS136