EE Department, Technion, Israel November 2004 Chip Multi Processor VLSI Architectures Seminar - 048879 Evgeny Bolotin 1 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
Outline Intro Single Chip Multiprocessor (IEEE Computer 97) IBM Power 5 (IEEE Micro 2004) Energy Efficiency (CMP vs. SMT), ICS 2004 Intel Network Processor-IXP 2800 Niagara – Sun, October 2004 Summary 2 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
When are two heads better than one? 3 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
CMP Motivation • How to utilize available silicon? § Speculation (aggressive superscalar) § Simultaneous Multithreading (SMT, Hyperthreading) § Several processors on a single chip • What is a CMP (Chip Multi. Processor)? § Several processors (several masters) § Both shared and distributed memory architectures § Both homogenous and heterogeneous processor types • Why? § Wire Delays § Diminishing of Uniprocessors § Very long design and verification times for modern processors 4 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
A Reminder: SMT (Simultaneous Multi Threading) SMT CMP • Simple Cores • Several Logical processors • Moderate amount of parallelism • Copy of State for each • Threads are running concurrently on • Mul. Threads are running concurrently different cores • Better utilization and Latency Tolerance • Pool of execution units (Wide machine) 5 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
A Reminder: SMT (Simultaneous Multi Threading) SMT vs. CMP 6 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
A Single Chip Multiprocessor L. Hammond at al. (Stanford), IEEE Computer 97 Superscalar (SS) SMT • For Same area (a billion tr. DRAM area) Superscalar and SMT: Very Complex • Wide • Advanced Branch prediction • Register Renaming • OOO Instruction Issue • Non-Blocking data caches 7 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004 CMP
SS and SMT vs. CMP CPU Cores: Three main hardware design problems (of SS and SMT): • Area increases quadratically with core complexity • Number of Registers O(Instruction window size) • Register ports - O(Issue width) üCMP solves this problem (~ linear Area to Issue width) • Longer Cycle Times • Long Wires, many MUXes and crossbars • Large buffers, queues and register files Clustering (decreases ILP) or Deep Pipelining (Branch mispredication penalties) üCMP allows small cycle time (with little effort) üSmall and fast üRelies on software to schedule - Poor ILP • Complex Design and Verification 8 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
SS and SMT vs. CMP Memory: • 12 issue SS or SMT require multiport data cache (4 -6 ports) • 2 X 128 Kbyte (2 cycle latency) üCMP 16 X 16 Kbyte (single cycle latency), but secondary cache is slower (multiport) üShared memory: write through caches SMT 9 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004 CMP
Performance comparison • Compress: (Integer apps) Low ILP and no TLP • Mpeg-2: (MMedia apps) High ILP and TLP and moderate memory requirement (parallelized by hand) + SMT utilizes core resources better + But CMP has 16 issue slots instead of 12 • Tomcatv: (FP applications) Large loop-level parallelism and large memory bandwidth (TLP by compiler) + CMP has large memory bandwidth on primary cache - SMT fundamental problem: unified and slow cache • Multiprogram: Integer multiprogramming workload, all computation-intensive (Low ILP, High PLP) 10 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
A Single Chip Multiprocessor L. Hammond at al. (Stanford), IEEE Computer 97 • TLP and PLP become widespread in future applications • Various Multimedia applications • Compilers and OS Ø Favours CMP: • Better performance with simple hardware • Higher clock rates, better memory bandwidth • Shorter pipelines SMT: has better utilizations but CMP has more resources (no wideissue logic) Although CMP bad for no TLP and ILP (compress), SMT and SS not much better 11 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
IBM Power 5 chip: a dual-core multithreaded processor Ron Kalla at al. ; Micro, IEEE , Volume: 24 , Issue: 2 , Mar-Apr 2004 Power 4 system (2001) • 174 Milion trans. , 2 cores • L 1: 64 K I$, 128 KB d$, 128 B lines • Shared L 2 (1. 5 MB) and L 3 (32 MB) • 5 way each (8 issue/5 retire) • 100 Instruction window • 15 stage pipe 12 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004 Power 5 system(2004) • 130 nm Enhancements: • Two-way SMT (additional complexity is unjustified…diminishing + cache trashing ) • L 3 closer to L 2 (less traffic, system scales better to 64 from 32) • Shared L 2 (1. 875 MB) and L 3 (36 MB) • Memory controller on chip • Less Sys. Chips, reduced Mem. latency
IBM Power 5 chip • 8 metals • 389 mm^2 • L 3 directory • on chip MC 13 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
IBM Power 5 chip Processor Core: • SMT and Single Thread (ST) operation modes • 8 way fetch and translation (shared resource) • Branch prediction (shared) • 3 BHT (shared) • Return stack (separate) • 120 GPRs and 120 FPRs (dynamically shared in SMT and all used in ST) • Shared issue queues and XUs 14 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
IBM Power 5 chip Enhanced SMT: • Dynamic resource balancing (i. e monitors L 2$ misses and throttles the thread) • Adjustable thread priority 8 levels- affects decode cycles (by software: idle loop, real-time apps. etc. ) 15 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
IBM Power 5 chip Dynamic Power Management: • Extensive Dynamic clock gating (on cycle basis) • Dual-Vt for reduced leakage • Low power mode (x 32 slower instruction dispatch, see figure) ST operation: • All physical resources to active thread 16 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
Idea: More Flexible CMP (? ) Poor Flexibility Bad for legacy code and low TLP 17 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
More Flexible CMP (? ) Programmable SMTs SMT 18 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
The Energy Efficiency of CMP vs. SMT for Multimedia Workloads, Ruchira Sasanka at al, from University of Illinois and Intel Arch. Research Lab , ICS’ 04 France, For Same Performance : Compare Energy Efficiency of CMP and SMT for MM apps. • MM applications become important and have high TLP • What about energy? • SMT might appear more energy efficient WRONG! • since it utilizes better the hardware resources Compare for many performance points – different core architectures and frequency/voltage 19 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
The Energy Efficiency of CMP vs. SMT for Multimedia Workloads Design Space: • 2 -thread and 4 -thread systems checked • Based on out-of-order superscalar (MIPS R 10000) • Fetch/decode from vary from 2 to 8 • other parameters, i. e. window size, number of executions units change accordingly • Frequency vary from 600 MHz to 1. 6 GHZ (and scale voltage accordingly) 20 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
The Energy Efficiency of CMP vs. SMT for Multimedia Workloads Core and Memory parameters: • In SMT: • Threads share most of the resources • Separate BPT, return address stack and 32 Integer and 32 FP registers • ICOUNT policy to prioritize Instr. Fetch (prefer “unstuck” thread) • Caches are modeled and the size is chosen to achieve 99% hit-rate for I$ and 98% for D$ • CMP: 16 K for I$ and 8 K for D$ • SMT was given same amount of cache per thread 21 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
The Energy Efficiency of CMP vs. SMT for Multimedia Workloads Workload: • 8 representing single threaded Multi-Media apps (video and speech codecs) • Given benchmark can be parallelized – each thread process different frame • Consider two- and four- threaded benchmarks • Run same number of frames on both systems • Compare using Energy per Instruction (EPI) and Time per Instruction (TPI) Simulation Environment: • Modified RSIM simulator-cycle level simulator, models branch and address speculation and contentions. • Use enhanced Wattch for dynamic power modeling • Model Bus power • Static power is modeled by Hot. Leakage model (2% of dynamic at 0. 13 um ) • Assume 90% of clock gating 22 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
The Energy Efficiency of CMP vs. SMT for Multimedia Workloads Metrics: Vary freq. 23 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
The Energy Efficiency of CMP vs. SMT for Multimedia Workloads Results: two-thread workloads 24 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
The Energy Efficiency of CMP vs. SMT for Multimedia Workloads Results: four-thread workloads 25 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
The Energy Efficiency of CMP vs. SMT for Multimedia Workloads Summary of simulation results: • CMP always gives least EPI (they assumed enough threads …) • For four-threads workload CMP is significantly better • Difference increases with increased performance • For fixed system: Overall best architecture can be picked up (by minimum deviation from the “best” curve) • Interesting graph: 26 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
The Energy Efficiency of CMP vs. SMT for Multimedia Workloads IDEAS: ? ? ? • Hybrid much better than SMT and comes very close to CMP • best of all worlds? • Maybe adaptive Architecture and DVS? • Adaptive fetch width? • DVS much easier for CMP? • High and medium performance regions? Heterogeneous ( ask Zvika) 27 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
Intel Network Processor (IXP 2800) • Xscale processor (700 MHz) • 16 Microengines (1. 4 GHz) • Power dissipation ~30 Watts • Package area : 38 x 38 mm^2 • Vdd = 1. 3 V 28 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
Intel Network Processor (IXP 2800) Microengine: • 6 stage pipeline • Huge register set • Multiple thread (8) • Memories • Hardware accelerators. 29 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
Sun Roadmap Blades: Niagara: • 8 Ultra Sparc II x 4 threads • They hope for x 15 performance 30 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
27 October 2004 (EETimes): Sun Microsystems has manufactured first silicon of its next-generation Niagara processor, which isn't due to ship until 2006. The advanced chip contains eight 64 -bit Ultra. Sparc cores and will power new systems, which Sun plans to position as "throughput computing" engines capable of handing network-intensive tasks. "We are now with a working chip, " David Yen, Sun's executive vice president for scalable systems, tells VARBusiness. "The 1. 0 design is running in the lab. It's running the Solaris 9 operating system on top of it, with 32 application threads on top of Solaris. " The eight-core Niagara will dissipate only 60 W of power, according to Yen. That's a fraction of the 100 W or so consumed by today's dual-core Ultra. Sparc IV and is also likely beneath the power figure expected from dual-core processors due out of Intel and AMD in 2005. Niagara will be fabricated in an advanced 90 -nm process. It also boasts a host of on-chip features, which make its design highly integrated. The initial version will include an on-board Ethernet controller and a built-in memory controller. Subsequent versions, according to Yen, "will have 10 -Gigabit Ethernet and even cryptologic [capability] built on the chip. " Sun says the year-long interval until the chip comes to market will be used to bring Sun's partners up to speed. 31 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
Conclusions • CMP reduces hardware/power overhead • SMT can yield better single-thread (at high cost) • CMP can improve application performance if the compiler can extract thread-level parallelism • What is the most effective use of on-chip real estate? § Depends on the workload § Depends on compiler technology • Hybrid/Reconfigurable? • Heterogeneous? 32 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
References: • VLSI Architectures Lecture (Uri Weiser) • Hyper-Threading Technology Architecture and Microarchitecture, Intel • A Single Chip Multiprocessor , L. Hammond at al. (Stanford), IEEE Computer 97 • IBM Power 5 chip: a dual-core multithreaded processor, Kalla, R. ; Balaram Sinharoy; Tendler, J. M. ; Micro, IEEE , Volume: 24 , Issue: 2 , Mar-Apr 2004 Pages: 40 – 47 • The Energy Efficiency of CMP vs. SMT for Multimedia Workloads, Ruchira Sasanka at al (University of Illinois and Intel Arch. Research Lab ), ICS’ 04 • Network processors, Intel Technology Journal. • “Sun weaves multi-media future”, K. Krewel Microprocessor report, april 2003 33 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
Energy: CMP vs SMT 34 Evgeny Bolotin – VLSI Arch Seminar, Nov 2004
Скачать презентацию EE Department Technion Israel November 2004 Chip Multi
a05f225196af48836218573efe3757a2.ppt