Simultaneous Speculative Threading: A Novel Pipeline Architecture Implemented in Sun's ROCK Processor
|
|
- Georgina Casey
- 5 years ago
- Views:
Transcription
1 Simultaneous Speculative Threading: A Novel Pipeline Architecture Implemented in Sun's ROCK Processor Shailender Chaudhry, Robert Cypher, Magnus Ekman, Martin Karlsson, Anders Landin, Sherman Yip, Haakan Zeffer, Marc Tremblay SUN Microsystems, Inc. June 24, 2009
2 Outline Core Architectures for Chip Multi-Processors (CMPs) Simultaneous Speculative Threading (SST) Core Architecture SST Implementation SST Performance Related Work Conclusions 2
3 Core Architectures for CMPs Goal 1: Maximum Core Count for Throughput > Small core area > Low core power > Small caches Goal 2: Provide Good Per-Thread Performance Option 1: In-Order Cores > Small and power-efficient > Poor per-thread performance Option 2: Traditional Out-of-Order (OoO) Cores > Good per-thread performance > Large and power-inefficient Large issue windows Register renaming logic Reorder buffers (ROBs) Memory disambiguation buffers (MDBs) 3
4 SST Cores SST: HW Automatically Parallelizes a Single SW Thread > Load miss in L1 cache starts parallelization using 2 HW threads > Ahead thread Checkpoints state and executes speculatively Instructions independent of load miss are speculatively retired Load miss(es) and dependent instructions are deferred to behind thread > Behind thread Executes deferred instructions and re-defers them if necessary Memory-Level Parallelism (MLP) > Ahead thread executes past load miss and generates additional load misses Instruction-Level Parallelism (ILP) > Ahead and behind threads execute independent instructions from different points in program in parallel Efficient > Eliminates need for expensive OoO structures 4
5 SST Example Instructions in Program Order DQ 1 DQ 2 DQ 3 DQ 4 Chkpt 0 Chkpt 1 Chkpt 2 Chkpt 3 Thread Sync Nonspec Spec Episode 1 Spec Episode 2 Spec Episode 3 Spec Episode 4 Nonspec KEY: Nonspeculative Instruction Load Miss Instruction Dependent on Load Miss Speculative Instruction Independent of Load Misses 5
6 SST Pipeline with N Checkpoints Fetch DQ 1..N DQ 1..N Decode SST with N Checkpoints Adds: > N Deferred Queues (DQ 1..N) Issue + Bypass Execution Pipes Working Working Register Register Files Files > N-1 Architectural Register Files (for speculative checkpoints) > 2 Working Register Files (used by behind thread) > Not Available (NA) bit per Register (dependent on load miss) Arch Arch Register Register Files Files 1..N 1..N ROCK Implements SST with N = 2 Checkpoints 6
7 SST Overview Speculative Episodes Can Fail Due To: > Hardware resource limit exceeded (e.g., store queue, DQ) > Executed down wrong path > Memory ordering violation > Exception encountered If Ahead Thread Detects Failure > Ahead thread continues execution as HW scout thread If Behind Thread Detects Failure > All speculative checkpoints are discarded > Nonspeculative execution starts from committed checkpoint 7
8 SST Overview Speculative Episode i Succeeds > When all instructions from DQ i retire > Checkpoint i commits > Checkpoint i-1 is discarded Thread Sync > When all deferred instructions retire > Results passed from behind thread to ahead thread Based on register's NA bit > Ahead thread continues execution nonspeculatively > Behind thread is inactive until next load miss starts speculation 8
9 SST Implementation Working Register File Operation > All register reads use working registers > Failed speculation copies committed checkpoint to working register file 3 Working Register Files > 1 Written & Read by Ahead Thread > 2 Written & Read by Behind Thread How are Values Passed from Ahead Thread to Behind Thread? > Known operands are stored in DQ with deferred instructions 9
10 SST Implementation Why 2 Working Register Files for Behind Thread? > Because behind thread can re-defer instructions CODE EXAMPLE: instr1: writes r5 = 0 instr2: reads r5 = 0 instr3: writes r5 = instr4: reads r5 = 1 KEY: red = DQ i instruction blue = DQ i+1 instruction EXECUTION EXAMPLE: instr1: writes r5 =? (NA = 1) instr2: reads r5 =? (NA = 1) instr3: writes r5 = 1 (NA = 0) instr1: writes r5 = 0 (NA = 0) instr2: reads r5 = 0 (CORRECT!) instr4: reads r5 = 0 (WRONG!) 10
11 SST Implementation Architectural Register Files Implemented as Circular FIFO > 1 copy holds nonspeculative (committed checkpoint) value > N-1 copies hold speculative checkpoint values Implementation is Efficient > Taking a new checkpoint requires 0 cycles > Registers are single-ported > Registers are implemented in area-efficient SRAM 11
12 SST Implementation Preventing Read-After-Write (RAW) Register Hazards > NA bits prevent use of register value before it is known Preventing Write-After-Read (WAR) Register Hazards > Known operands are stored with deferred instructions Preventing Write-After-Write (WAW) Register Hazards > NA bits in architectural registers show if register value is known > Used to determine if behind thread writes to architectural register 12
13 SST Implementation Store Ordering > Ahead thread places stores in store queue in program order Store queue entries have NA bits > Stores not committed to memory until checkpoint commits Load Ordering > Speculative loads set "speculatively-read-bit" in L1 Data Cache > Invalidation or replacement of speculatively-read line fails speculation OR > CAM-based Memory Disambiguation Buffer (MDB) 13
14 SST Performance Model In-House Cycle-Accurate CMP Simulator Validated against ROCK Hardware FEATURE SST DEFAULT OoO DEFAULT # Cores 32 L1 I-Cache L1 D-Cache L2 Cache Store Queue Mem Lat 32KB/IFU, 4-way, 2-cycle 32KB/core, 4-way, 3-cycle 4MB/chip, 8-way, 25-cycle 64 entries/core (banked) 300 cycles Trap/Fail Lat 24 cycles Checkpoints 8 per core NA DQ Size 32 entries NA Issue Window 2 entries/pipe 32 entries Reorder Buffer NA 128 entries 14
15 SST Performance Results SST Execution Mode Breakdown 100% 80% 60% 40% 20% 0% OLTP JBB Scout Speculative NonSpeculative SAP COMMERCIAL gcc mcf hmmer SPEC INT gromacs cactusadm lbm SPEC FP 15
16 SST Performance Results SST Performance vs. Number of Checkpoints Normalized Performance CKPTS 4 CKPTS 8 CKPTS 16 CKPTS COMMERCIAL SPEC INT SPEC FP 16
17 SST Performance Results Comparison of STALL, Scout, SST & OoO Architectures Normalized Improvement over STALL 60% 50% 40% 30% 20% 10% 0% Scout SST OOO COMMERCIAL 8P SPEC INT 8P SPEC FP 8P COMMERCIAL 32P SPEC INT 32P SPEC FP 32P 17
18 SST Performance Results 60% Reliance on CAM-Based Structures Normalized Improvement over STALL 50% 40% 30% 20% 10% 0% COMMERCIAL 8P SPEC INT 8P SPEC FP 8P COMMERCIAL 32P SPEC INT 32P SPEC FP 32P SST SST NO RENAME NO MDB OOO OOO NO RENAME NO MDB 18
19 SST Performance Results Load Miss in OoO Core Without Register Renaming > Blocks instructions with RAW dependence on load > Blocks instructions with WAR or WAW dependence on load Load Miss in SST Core Without Register Renaming > Defers instructions with RAW dependence on load > Speculatively retires instructions with WAR or WAW dependence on load Load Miss in OoO Core Without MDB > Blocks store dependent on load > Blocked store blocks all younger loads Load Miss in SST Core Without MDB > Store with NA data placed in store queue > Loads with RAW-bypass from store with NA data are deferred > Store with NA address starts scout mode 19
20 Related Work Checkpoint-Based OoO Designs > Kilo-Instruction Processor (KIP) > Checkpoint Processing and Recovery (CPR) > Cherry Similarities with SST > Checkpoints reduce or eliminate expensive OoO structures Issue window Register renaming ROB MDB Differences from SST > SST achieves ILP via simultaneous execution from 2 points in program > SST scouts when resource limitations exceeded 20
21 Related Work Flea-Flicker > Uses ahead thread and behind thread > Behind thread executes load misses and dependents Similarity with SST > Use of ahead and behind threads Differences from SST > SST executes ahead and behind threads on same pipelines More area-efficient for CMPs SST design can be leveraged for supporting multiple SW threads on core > SST redefers DQ instructions that are cache misses Flea-flicker stalls behind thread on cache miss > SST supports multiple checkpoints 21
22 Related Work Continual Flow Pipelines > Uses checkpoints to allow speculation past cache misses Similarities with SST > Simple, nonblocking issue logic Differences from SST > SST executes from multiple locations in program in parallel > SST supports multiple checkpoints 22
23 Conclusions SST is an Efficient Core Architecture for CMPs > Eliminates need for expensive CAM-based structures > Eliminates need for ROB > Simplifies instruction issue logic SST Obtains Effective ILP and MLP > Matches or exceeds performance of larger OoO cores in many cases 23
24 Panel Dynamic Performance Tuning for Speculative Threads > Yangchun Luo, Venkatesan Packirisamy, Nikhil Mungre, Ankit Tarkas, Antonia Zhai, and Wei-Chung Hsu Boosting Single-Thread Performance in Multi-Core Systems through Fine-Grain Multi-Threading > Carlos Madriles, Pedro Lopez, Josep M. Codina, Enric Gibert, Fernando Latorre, Alejandro Martinez, Raul Martinez, and Antonio Gonzalez Simultaneous Speculative Threading: A Novel Pipeline Architecture Implemented in Sun's ROCK Processor > Shailender Chaudhry, Robert Cypher, Magnus Ekman, Martin Karlsson, Anders Landin, Sherman Yip, Haakan Zeffer, and Marc Tremblay 24
Lecture 16: Checkpointed Processors. Department of Electrical Engineering Stanford University
Lecture 16: Checkpointed Processors Department of Electrical Engineering Stanford University http://eeclass.stanford.edu/ee382a Lecture 18-1 Announcements Reading for today: class notes Your main focus:
More informationHandout 2 ILP: Part B
Handout 2 ILP: Part B Review from Last Time #1 Leverage Implicit Parallelism for Performance: Instruction Level Parallelism Loop unrolling by compiler to increase ILP Branch prediction to increase ILP
More informationLecture 9: More ILP. Today: limits of ILP, case studies, boosting ILP (Sections )
Lecture 9: More ILP Today: limits of ILP, case studies, boosting ILP (Sections 3.8-3.14) 1 ILP Limits The perfect processor: Infinite registers (no WAW or WAR hazards) Perfect branch direction and target
More informationLecture 11: Out-of-order Processors. Topics: more ooo design details, timing, load-store queue
Lecture 11: Out-of-order Processors Topics: more ooo design details, timing, load-store queue 1 Problem 0 Show the renamed version of the following code: Assume that you have 36 physical registers and
More information250P: Computer Systems Architecture. Lecture 9: Out-of-order execution (continued) Anton Burtsev February, 2019
250P: Computer Systems Architecture Lecture 9: Out-of-order execution (continued) Anton Burtsev February, 2019 The Alpha 21264 Out-of-Order Implementation Reorder Buffer (ROB) Branch prediction and instr
More informationLecture: Out-of-order Processors. Topics: out-of-order implementations with issue queue, register renaming, and reorder buffer, timing, LSQ
Lecture: Out-of-order Processors Topics: out-of-order implementations with issue queue, register renaming, and reorder buffer, timing, LSQ 1 An Out-of-Order Processor Implementation Reorder Buffer (ROB)
More informationDynamic Performance Tuning for Speculative Threads
Dynamic Performance Tuning for Speculative Threads Yangchun Luo, Venkatesan Packirisamy, Nikhil Mungre, Ankit Tarkas, Wei-Chung Hsu, and Antonia Zhai Dept. of Computer Science and Engineering Dept. of
More informationCS377P Programming for Performance Single Thread Performance Out-of-order Superscalar Pipelines
CS377P Programming for Performance Single Thread Performance Out-of-order Superscalar Pipelines Sreepathi Pai UTCS September 14, 2015 Outline 1 Introduction 2 Out-of-order Scheduling 3 The Intel Haswell
More informationMultiple Instruction Issue and Hardware Based Speculation
Multiple Instruction Issue and Hardware Based Speculation Soner Önder Michigan Technological University, Houghton MI www.cs.mtu.edu/~soner Hardware Based Speculation Exploiting more ILP requires that we
More informationReorder Buffer Implementation (Pentium Pro) Reorder Buffer Implementation (Pentium Pro)
Reorder Buffer Implementation (Pentium Pro) Hardware data structures retirement register file (RRF) (~ IBM 360/91 physical registers) physical register file that is the same size as the architectural registers
More informationLecture 9: Dynamic ILP. Topics: out-of-order processors (Sections )
Lecture 9: Dynamic ILP Topics: out-of-order processors (Sections 2.3-2.6) 1 An Out-of-Order Processor Implementation Reorder Buffer (ROB) Branch prediction and instr fetch R1 R1+R2 R2 R1+R3 BEQZ R2 R3
More informationCS425 Computer Systems Architecture
CS425 Computer Systems Architecture Fall 2017 Thread Level Parallelism (TLP) CS425 - Vassilis Papaefstathiou 1 Multiple Issue CPI = CPI IDEAL + Stalls STRUC + Stalls RAW + Stalls WAR + Stalls WAW + Stalls
More informationLIMITS OF ILP. B649 Parallel Architectures and Programming
LIMITS OF ILP B649 Parallel Architectures and Programming A Perfect Processor Register renaming infinite number of registers hence, avoids all WAW and WAR hazards Branch prediction perfect prediction Jump
More informationInstruction Level Parallelism (ILP)
1 / 26 Instruction Level Parallelism (ILP) ILP: The simultaneous execution of multiple instructions from a program. While pipelining is a form of ILP, the general application of ILP goes much further into
More informationLecture 8: Branch Prediction, Dynamic ILP. Topics: static speculation and branch prediction (Sections )
Lecture 8: Branch Prediction, Dynamic ILP Topics: static speculation and branch prediction (Sections 2.3-2.6) 1 Correlating Predictors Basic branch prediction: maintain a 2-bit saturating counter for each
More informationCMSC 411 Computer Systems Architecture Lecture 13 Instruction Level Parallelism 6 (Limits to ILP & Threading)
CMSC 411 Computer Systems Architecture Lecture 13 Instruction Level Parallelism 6 (Limits to ILP & Threading) Limits to ILP Conflicting studies of amount of ILP Benchmarks» vectorized Fortran FP vs. integer
More informationSuperscalar Organization
Superscalar Organization Nima Honarmand Instruction-Level Parallelism (ILP) Recall: Parallelism is the number of independent tasks available ILP is a measure of inter-dependencies between insns. Average
More informationDynamic Scheduling. CSE471 Susan Eggers 1
Dynamic Scheduling Why go out of style? expensive hardware for the time (actually, still is, relatively) register files grew so less register pressure early RISCs had lower CPIs Why come back? higher chip
More informationRECAP. B649 Parallel Architectures and Programming
RECAP B649 Parallel Architectures and Programming RECAP 2 Recap ILP Exploiting ILP Dynamic scheduling Thread-level Parallelism Memory Hierarchy Other topics through student presentations Virtual Machines
More informationSuper Scalar. Kalyan Basu March 21,
Super Scalar Kalyan Basu basu@cse.uta.edu March 21, 2007 1 Super scalar Pipelines A pipeline that can complete more than 1 instruction per cycle is called a super scalar pipeline. We know how to build
More informationEITF20: Computer Architecture Part3.2.1: Pipeline - 3
EITF20: Computer Architecture Part3.2.1: Pipeline - 3 Liang Liu liang.liu@eit.lth.se 1 Outline Reiteration Dynamic scheduling - Tomasulo Superscalar, VLIW Speculation ILP limitations What we have done
More informationChapter 4 The Processor 1. Chapter 4D. The Processor
Chapter 4 The Processor 1 Chapter 4D The Processor Chapter 4 The Processor 2 Instruction-Level Parallelism (ILP) Pipelining: executing multiple instructions in parallel To increase ILP Deeper pipeline
More informationece4750-t11-ooo-execution-notes.txt ========================================================================== ece4750-l12-ooo-execution-notes.txt ==========================================================================
More informationComputer Architecture Spring 2016
Computer Architecture Spring 2016 Final Review Shuai Wang Department of Computer Science and Technology Nanjing University Computer Architecture Computer architecture, like other architecture, is the art
More informationEN164: Design of Computing Systems Topic 06.b: Superscalar Processor Design
EN164: Design of Computing Systems Topic 06.b: Superscalar Processor Design Professor Sherief Reda http://scale.engin.brown.edu Electrical Sciences and Computer Engineering School of Engineering Brown
More informationAdvanced issues in pipelining
Advanced issues in pipelining 1 Outline Handling exceptions Supporting multi-cycle operations Pipeline evolution Examples of real pipelines 2 Handling exceptions 3 Exceptions In pipelined execution, one
More information15-740/ Computer Architecture Lecture 5: Precise Exceptions. Prof. Onur Mutlu Carnegie Mellon University
15-740/18-740 Computer Architecture Lecture 5: Precise Exceptions Prof. Onur Mutlu Carnegie Mellon University Last Time Performance Metrics Amdahl s Law Single-cycle, multi-cycle machines Pipelining Stalls
More informationOutline EEL 5764 Graduate Computer Architecture. Chapter 3 Limits to ILP and Simultaneous Multithreading. Overcoming Limits - What do we need??
Outline EEL 7 Graduate Computer Architecture Chapter 3 Limits to ILP and Simultaneous Multithreading! Limits to ILP! Thread Level Parallelism! Multithreading! Simultaneous Multithreading Ann Gordon-Ross
More informationCopyright 2012, Elsevier Inc. All rights reserved.
Computer Architecture A Quantitative Approach, Fifth Edition Chapter 3 Instruction-Level Parallelism and Its Exploitation 1 Branch Prediction Basic 2-bit predictor: For each branch: Predict taken or not
More informationLecture-13 (ROB and Multi-threading) CS422-Spring
Lecture-13 (ROB and Multi-threading) CS422-Spring 2018 Biswa@CSE-IITK Cycle 62 (Scoreboard) vs 57 in Tomasulo Instruction status: Read Exec Write Exec Write Instruction j k Issue Oper Comp Result Issue
More informationDual-Core Execution: Building A Highly Scalable Single-Thread Instruction Window
Dual-Core Execution: Building A Highly Scalable Single-Thread Instruction Window Huiyang Zhou School of Computer Science University of Central Florida New Challenges in Billion-Transistor Processor Era
More informationComputer Architecture 计算机体系结构. Lecture 4. Instruction-Level Parallelism II 第四讲 指令级并行 II. Chao Li, PhD. 李超博士
Computer Architecture 计算机体系结构 Lecture 4. Instruction-Level Parallelism II 第四讲 指令级并行 II Chao Li, PhD. 李超博士 SJTU-SE346, Spring 2018 Review Hazards (data/name/control) RAW, WAR, WAW hazards Different types
More informationEN164: Design of Computing Systems Lecture 24: Processor / ILP 5
EN164: Design of Computing Systems Lecture 24: Processor / ILP 5 Professor Sherief Reda http://scale.engin.brown.edu Electrical Sciences and Computer Engineering School of Engineering Brown University
More informationUG4 Honours project selection: Talk to Vijay or Boris if interested in computer architecture projects
Announcements UG4 Honours project selection: Talk to Vijay or Boris if interested in computer architecture projects Inf3 Computer Architecture - 2017-2018 1 Last time: Tomasulo s Algorithm Inf3 Computer
More informationOut of Order Processing
Out of Order Processing Manu Awasthi July 3 rd 2018 Computer Architecture Summer School 2018 Slide deck acknowledgements : Rajeev Balasubramonian (University of Utah), Computer Architecture: A Quantitative
More informationCISC 662 Graduate Computer Architecture Lecture 13 - Limits of ILP
CISC 662 Graduate Computer Architecture Lecture 13 - Limits of ILP Michela Taufer http://www.cis.udel.edu/~taufer/teaching/cis662f07 Powerpoint Lecture Notes from John Hennessy and David Patterson s: Computer
More informationExploitation of instruction level parallelism
Exploitation of instruction level parallelism Computer Architecture J. Daniel García Sánchez (coordinator) David Expósito Singh Francisco Javier García Blas ARCOS Group Computer Science and Engineering
More informationSupporting Speculative Multithreading on Simultaneous Multithreaded Processors
Supporting Speculative Multithreading on Simultaneous Multithreaded Processors Venkatesan Packirisamy, Shengyue Wang, Antonia Zhai, Wei-Chung Hsu, and Pen-Chung Yew Department of Computer Science, University
More informationCPI < 1? How? What if dynamic branch prediction is wrong? Multiple issue processors: Speculative Tomasulo Processor
1 CPI < 1? How? From Single-Issue to: AKS Scalar Processors Multiple issue processors: VLIW (Very Long Instruction Word) Superscalar processors No ISA Support Needed ISA Support Needed 2 What if dynamic
More informationOutline. Exploiting Program Parallelism. The Hydra Approach. Data Speculation Support for a Chip Multiprocessor (Hydra CMP) HYDRA
CS 258 Parallel Computer Architecture Data Speculation Support for a Chip Multiprocessor (Hydra CMP) Lance Hammond, Mark Willey and Kunle Olukotun Presented: May 7 th, 2008 Ankit Jain Outline The Hydra
More informationNOW Handout Page 1. Review from Last Time #1. CSE 820 Graduate Computer Architecture. Lec 8 Instruction Level Parallelism. Outline
CSE 820 Graduate Computer Architecture Lec 8 Instruction Level Parallelism Based on slides by David Patterson Review Last Time #1 Leverage Implicit Parallelism for Performance: Instruction Level Parallelism
More informationCPI IPC. 1 - One At Best 1 - One At best. Multiple issue processors: VLIW (Very Long Instruction Word) Speculative Tomasulo Processor
Single-Issue Processor (AKA Scalar Processor) CPI IPC 1 - One At Best 1 - One At best 1 From Single-Issue to: AKS Scalar Processors CPI < 1? How? Multiple issue processors: VLIW (Very Long Instruction
More informationCSE 820 Graduate Computer Architecture. week 6 Instruction Level Parallelism. Review from Last Time #1
CSE 820 Graduate Computer Architecture week 6 Instruction Level Parallelism Based on slides by David Patterson Review from Last Time #1 Leverage Implicit Parallelism for Performance: Instruction Level
More informationInstruction Level Parallelism
Instruction Level Parallelism Software View of Computer Architecture COMP2 Godfrey van der Linden 200-0-0 Introduction Definition of Instruction Level Parallelism(ILP) Pipelining Hazards & Solutions Dynamic
More informationECE 4750 Computer Architecture, Fall 2018 T12 Advanced Processors: Memory Disambiguation
ECE 4750 Computer Architecture, Fall 208 T2 Advanced Processors: Memory Disambiguation School of Electrical and Computer Engineering Cornell University revision: 208--09-4-58 Adding Memory Instructions
More informationChapter 06: Instruction Pipelining and Parallel Processing
Chapter 06: Instruction Pipelining and Parallel Processing Lesson 09: Superscalar Processors and Parallel Computer Systems Objective To understand parallel pipelines and multiple execution units Instruction
More informationReferences EE457. Out of Order (OoO) Execution. Instruction Scheduling (Re-ordering of instructions)
EE457 Out of Order (OoO) Execution Introduction to Dynamic Scheduling of Instructions (The Tomasulo Algorithm) By Gandhi Puvvada References EE557 Textbook Prof Dubois EE557 Classnotes Prof Annavaram s
More informationCOSC4201 Instruction Level Parallelism Dynamic Scheduling
COSC4201 Instruction Level Parallelism Dynamic Scheduling Prof. Mokhtar Aboelaze Parts of these slides are taken from Notes by Prof. David Patterson (UCB) Outline Data dependence and hazards Exposing parallelism
More informationSpring 2010 Prof. Hyesoon Kim. Thanks to Prof. Loh & Prof. Prvulovic
Spring 2010 Prof. Hyesoon Kim Thanks to Prof. Loh & Prof. Prvulovic C/C++ program Compiler Assembly Code (binary) Processor 0010101010101011110 Memory MAR MDR INPUT Processing Unit OUTPUT ALU TEMP PC Control
More informationData/Thread Level Speculation (TLS) in the Stanford Hydra Chip Multiprocessor (CMP)
Data/Thread Level Speculation (TLS) in the Stanford Hydra Chip Multiprocessor (CMP) A 4-core Chip Multiprocessor (CMP) based microarchitecture/compiler effort at Stanford that provides hardware/software
More informationMultithreaded Processors. Department of Electrical Engineering Stanford University
Lecture 12: Multithreaded Processors Department of Electrical Engineering Stanford University http://eeclass.stanford.edu/ee382a Lecture 12-1 The Big Picture Previous lectures: Core design for single-thread
More informationChapter 3 (CONT II) Instructor: Josep Torrellas CS433. Copyright J. Torrellas 1999,2001,2002,2007,
Chapter 3 (CONT II) Instructor: Josep Torrellas CS433 Copyright J. Torrellas 1999,2001,2002,2007, 2013 1 Hardware-Based Speculation (Section 3.6) In multiple issue processors, stalls due to branches would
More informationCPE 631 Lecture 10: Instruction Level Parallelism and Its Dynamic Exploitation
Lecture 10: Instruction Level Parallelism and Its Dynamic Exploitation Aleksandar Milenković, milenka@ece.uah.edu Electrical and Computer Engineering University of Alabama in Huntsville Outline Tomasulo
More informationLecture 16: Core Design. Today: basics of implementing a correct ooo core: register renaming, commit, LSQ, issue queue
Lecture 16: Core Design Today: basics of implementing a correct ooo core: register renaming, commit, LSQ, issue queue 1 The Alpha 21264 Out-of-Order Implementation Reorder Buffer (ROB) Branch prediction
More informationEN164: Design of Computing Systems Topic 08: Parallel Processor Design (introduction)
EN164: Design of Computing Systems Topic 08: Parallel Processor Design (introduction) Professor Sherief Reda http://scale.engin.brown.edu Electrical Sciences and Computer Engineering School of Engineering
More informationChapter 3 Instruction-Level Parallelism and its Exploitation (Part 1)
Chapter 3 Instruction-Level Parallelism and its Exploitation (Part 1) ILP vs. Parallel Computers Dynamic Scheduling (Section 3.4, 3.5) Dynamic Branch Prediction (Section 3.3) Hardware Speculation and Precise
More informationArchitectures for Instruction-Level Parallelism
Low Power VLSI System Design Lecture : Low Power Microprocessor Design Prof. R. Iris Bahar October 0, 07 The HW/SW Interface Seminar Series Jointly sponsored by Engineering and Computer Science Hardware-Software
More informationInstruction Level Parallelism
Instruction Level Parallelism The potential overlap among instruction execution is called Instruction Level Parallelism (ILP) since instructions can be executed in parallel. There are mainly two approaches
More informationPage 1. Recall from Pipelining Review. Lecture 16: Instruction Level Parallelism and Dynamic Execution #1: Ideas to Reduce Stalls
CS252 Graduate Computer Architecture Recall from Pipelining Review Lecture 16: Instruction Level Parallelism and Dynamic Execution #1: March 16, 2001 Prof. David A. Patterson Computer Science 252 Spring
More informationCS 152, Spring 2011 Section 8
CS 152, Spring 2011 Section 8 Christopher Celio University of California, Berkeley Agenda Grades Upcoming Quiz 3 What it covers OOO processors VLIW Branch Prediction Intel Core 2 Duo (Penryn) Vs. NVidia
More informationThis Set. Scheduling and Dynamic Execution Definitions From various parts of Chapter 4. Description of Three Dynamic Scheduling Methods
10-1 Dynamic Scheduling 10-1 This Set Scheduling and Dynamic Execution Definitions From various parts of Chapter 4. Description of Three Dynamic Scheduling Methods Not yet complete. (Material below may
More informationFour Steps of Speculative Tomasulo cycle 0
HW support for More ILP Hardware Speculative Execution Speculation: allow an instruction to issue that is dependent on branch, without any consequences (including exceptions) if branch is predicted incorrectly
More informationAnnouncements. ECE4750/CS4420 Computer Architecture L11: Speculative Execution I. Edward Suh Computer Systems Laboratory
ECE4750/CS4420 Computer Architecture L11: Speculative Execution I Edward Suh Computer Systems Laboratory suh@csl.cornell.edu Announcements Lab3 due today 2 1 Overview Branch penalties limit performance
More informationHardware Speculation Support
Hardware Speculation Support Conditional instructions Most common form is conditional move BNEZ R1, L ;if MOV R2, R3 ;then CMOVZ R2,R3, R1 L: ;else Other variants conditional loads and stores nullification
More informationEfficient Runahead Threads Tanausú Ramírez Alex Pajuelo Oliverio J. Santana Onur Mutlu Mateo Valero
Efficient Runahead Threads Tanausú Ramírez Alex Pajuelo Oliverio J. Santana Onur Mutlu Mateo Valero The Nineteenth International Conference on Parallel Architectures and Compilation Techniques (PACT) 11-15
More informationHardware-Based Speculation
Hardware-Based Speculation Execute instructions along predicted execution paths but only commit the results if prediction was correct Instruction commit: allowing an instruction to update the register
More informationCISC 662 Graduate Computer Architecture Lecture 13 - CPI < 1
CISC 662 Graduate Computer Architecture Lecture 13 - CPI < 1 Michela Taufer http://www.cis.udel.edu/~taufer/teaching/cis662f07 Powerpoint Lecture Notes from John Hennessy and David Patterson s: Computer
More informationInstruction Level Parallelism. Appendix C and Chapter 3, HP5e
Instruction Level Parallelism Appendix C and Chapter 3, HP5e Outline Pipelining, Hazards Branch prediction Static and Dynamic Scheduling Speculation Compiler techniques, VLIW Limits of ILP. Implementation
More informationE0-243: Computer Architecture
E0-243: Computer Architecture L1 ILP Processors RG:E0243:L1-ILP Processors 1 ILP Architectures Superscalar Architecture VLIW Architecture EPIC, Subword Parallelism, RG:E0243:L1-ILP Processors 2 Motivation
More informationData/Thread Level Speculation (TLS) in the Stanford Hydra Chip Multiprocessor (CMP)
Data/Thread Level Speculation (TLS) in the Stanford Hydra Chip Multiprocessor (CMP) Hydra ia a 4-core Chip Multiprocessor (CMP) based microarchitecture/compiler effort at Stanford that provides hardware/software
More informationDynamic Memory Dependence Predication
Dynamic Memory Dependence Predication Zhaoxiang Jin and Soner Önder ISCA-2018, Los Angeles Background 1. Store instructions do not update the cache until they are retired (too late). 2. Store queue is
More informationCPE 631 Lecture 11: Instruction Level Parallelism and Its Dynamic Exploitation
Lecture 11: Instruction Level Parallelism and Its Dynamic Exploitation Aleksandar Milenkovic, milenka@ece.uah.edu Electrical and Computer Engineering University of Alabama in Huntsville Outline Instruction
More informationHardware-Based Speculation
Hardware-Based Speculation Execute instructions along predicted execution paths but only commit the results if prediction was correct Instruction commit: allowing an instruction to update the register
More informationInstruction Level Parallelism
Instruction Level Parallelism Dynamic scheduling Scoreboard Technique Tomasulo Algorithm Speculation Reorder Buffer Superscalar Processors 1 Definition of ILP ILP=Potential overlap of execution among unrelated
More informationStatic & Dynamic Instruction Scheduling
CS3014: Concurrent Systems Static & Dynamic Instruction Scheduling Slides originally developed by Drew Hilton, Amir Roth, Milo Martin and Joe Devietti at University of Pennsylvania 1 Instruction Scheduling
More informationKaisen Lin and Michael Conley
Kaisen Lin and Michael Conley Simultaneous Multithreading Instructions from multiple threads run simultaneously on superscalar processor More instruction fetching and register state Commercialized! DEC
More informationCS252 Graduate Computer Architecture Lecture 6. Recall: Software Pipelining Example
CS252 Graduate Computer Architecture Lecture 6 Tomasulo, Implicit Register Renaming, Loop-Level Parallelism Extraction Explicit Register Renaming John Kubiatowicz Electrical Engineering and Computer Sciences
More informationCISC 662 Graduate Computer Architecture Lecture 13 - Limits of ILP
CISC 662 Graduate Computer Architecture Lecture 13 - Limits of ILP Michela Taufer http://www.cis.udel.edu/~taufer/teaching/cis662f07 Powerpoint Lecture Notes from John Hennessy and David Patterson s: Computer
More informationTDT 4260 lecture 7 spring semester 2015
1 TDT 4260 lecture 7 spring semester 2015 Lasse Natvig, The CARD group Dept. of computer & information science NTNU 2 Lecture overview Repetition Superscalar processor (out-of-order) Dependencies/forwarding
More informationEE382A Lecture 7: Dynamic Scheduling. Department of Electrical Engineering Stanford University
EE382A Lecture 7: Dynamic Scheduling Department of Electrical Engineering Stanford University http://eeclass.stanford.edu/ee382a Lecture 7-1 Announcements Project proposal due on Wed 10/14 2-3 pages submitted
More informationPortland State University ECE 587/687. The Microarchitecture of Superscalar Processors
Portland State University ECE 587/687 The Microarchitecture of Superscalar Processors Copyright by Alaa Alameldeen and Haitham Akkary 2011 Program Representation An application is written as a program,
More informationDYNAMIC AND SPECULATIVE INSTRUCTION SCHEDULING
DYNAMIC AND SPECULATIVE INSTRUCTION SCHEDULING Slides by: Pedro Tomás Additional reading: Computer Architecture: A Quantitative Approach, 5th edition, Chapter 3, John L. Hennessy and David A. Patterson,
More informationLimitations of Scalar Pipelines
Limitations of Scalar Pipelines Superscalar Organization Modern Processor Design: Fundamentals of Superscalar Processors Scalar upper bound on throughput IPC = 1 Inefficient unified pipeline
More informationLoad1 no Load2 no Add1 Y Sub Reg[F2] Reg[F6] Add2 Y Add Reg[F2] Add1 Add3 no Mult1 Y Mul Reg[F2] Reg[F4] Mult2 Y Div Reg[F6] Mult1
Instruction Issue Execute Write result L.D F6, 34(R2) L.D F2, 45(R3) MUL.D F0, F2, F4 SUB.D F8, F2, F6 DIV.D F10, F0, F6 ADD.D F6, F8, F2 Name Busy Op Vj Vk Qj Qk A Load1 no Load2 no Add1 Y Sub Reg[F2]
More informationPrecise Exceptions and Out-of-Order Execution. Samira Khan
Precise Exceptions and Out-of-Order Execution Samira Khan Multi-Cycle Execution Not all instructions take the same amount of time for execution Idea: Have multiple different functional units that take
More informationPortland State University ECE 587/687. Memory Ordering
Portland State University ECE 587/687 Memory Ordering Copyright by Alaa Alameldeen, Zeshan Chishti and Haitham Akkary 2018 Handling Memory Operations Review pipeline for out of order, superscalar processors
More informationCS152 Computer Architecture and Engineering SOLUTIONS Complex Pipelines, Out-of-Order Execution, and Speculation Problem Set #3 Due March 12
Assigned 2/28/2018 CS152 Computer Architecture and Engineering SOLUTIONS Complex Pipelines, Out-of-Order Execution, and Speculation Problem Set #3 Due March 12 http://inst.eecs.berkeley.edu/~cs152/sp18
More informationThis Set. Scheduling and Dynamic Execution Definitions From various parts of Chapter 4. Description of Two Dynamic Scheduling Methods
10 1 Dynamic Scheduling 10 1 This Set Scheduling and Dynamic Execution Definitions From various parts of Chapter 4. Description of Two Dynamic Scheduling Methods Not yet complete. (Material below may repeat
More informationHardware-based speculation (2.6) Multiple-issue plus static scheduling = VLIW (2.7) Multiple-issue, dynamic scheduling, and speculation (2.
Instruction-Level Parallelism and its Exploitation: PART 2 Hardware-based speculation (2.6) Multiple-issue plus static scheduling = VLIW (2.7) Multiple-issue, dynamic scheduling, and speculation (2.8)
More informationCS 252 Graduate Computer Architecture. Lecture 4: Instruction-Level Parallelism
CS 252 Graduate Computer Architecture Lecture 4: Instruction-Level Parallelism Krste Asanovic Electrical Engineering and Computer Sciences University of California, Berkeley http://wwweecsberkeleyedu/~krste
More informationSRAMs to Memory. Memory Hierarchy. Locality. Low Power VLSI System Design Lecture 10: Low Power Memory Design
SRAMs to Memory Low Power VLSI System Design Lecture 0: Low Power Memory Design Prof. R. Iris Bahar October, 07 Last lecture focused on the SRAM cell and the D or D memory architecture built from these
More informationEECC551 - Shaaban. 1 GHz? to???? GHz CPI > (?)
Evolution of Processor Performance So far we examined static & dynamic techniques to improve the performance of single-issue (scalar) pipelined CPU designs including: static & dynamic scheduling, static
More information5008: Computer Architecture
5008: Computer Architecture Chapter 2 Instruction-Level Parallelism and Its Exploitation CA Lecture05 - ILP (cwliu@twins.ee.nctu.edu.tw) 05-1 Review from Last Lecture Instruction Level Parallelism Leverage
More informationGetting CPI under 1: Outline
CMSC 411 Computer Systems Architecture Lecture 12 Instruction Level Parallelism 5 (Improving CPI) Getting CPI under 1: Outline More ILP VLIW branch target buffer return address predictor superscalar more
More informationINSTITUTO SUPERIOR TÉCNICO. Architectures for Embedded Computing
UNIVERSIDADE TÉCNICA DE LISBOA INSTITUTO SUPERIOR TÉCNICO Departamento de Engenharia Informática Architectures for Embedded Computing MEIC-A, MEIC-T, MERC Lecture Slides Version 3.0 - English Lecture 09
More informationPipelining. CSC Friday, November 6, 2015
Pipelining CSC 211.01 Friday, November 6, 2015 Performance Issues Longest delay determines clock period Critical path: load instruction Instruction memory register file ALU data memory register file Not
More informationComputer Architecture A Quantitative Approach, Fifth Edition. Chapter 3. Instruction-Level Parallelism and Its Exploitation
Computer Architecture A Quantitative Approach, Fifth Edition Chapter 3 Instruction-Level Parallelism and Its Exploitation Introduction Pipelining become universal technique in 1985 Overlaps execution of
More informationSimultaneous Multithreading Architecture
Simultaneous Multithreading Architecture Virendra Singh Indian Institute of Science Bangalore Lecture-32 SE-273: Processor Design For most apps, most execution units lie idle For an 8-way superscalar.
More informationRecall from Pipelining Review. Lecture 16: Instruction Level Parallelism and Dynamic Execution #1: Ideas to Reduce Stalls
CS252 Graduate Computer Architecture Recall from Pipelining Review Lecture 16: Instruction Level Parallelism and Dynamic Execution #1: March 16, 2001 Prof. David A. Patterson Computer Science 252 Spring
More informationData/Thread Level Speculation (TLS) in the Stanford Hydra Chip Multiprocessor (CMP)
Data/Thread Level Speculation (TLS) in the Stanford Hydra Chip Multiprocessor (CMP) Hydra is a 4-core Chip Multiprocessor (CMP) based microarchitecture/compiler effort at Stanford that provides hardware/software
More information