Multiprocessors and Thread-Level Parallelism
description
Transcript of Multiprocessors and Thread-Level Parallelism
ENGS 116 Lecture 15 1
Multiprocessors and Thread-Level Parallelism
Vincent BerkNovember 12th , 2008
Reading for Friday: Sections 4.1-4.3
Reading for Monday: Sections 4.4 – 4.9
ENGS 116 Lecture 15 2
Parallel Computers• Definition: “A parallel computer is a collection of processing
elements that cooperate and communicate to solve large problems fast.”
Almasi and Gottlieb, Highly Parallel Computing, 1989• Questions about parallel computers:
– How large a collection?– How powerful are processing elements?– How do they cooperate and communicate?– How is data transmitted? – What type of interconnection?– What are HW and SW primitives for programmer?– Does it translate into performance?
ENGS 116 Lecture 15 3
Parallel Processors “Religion”• The dream of computer architects since 1960: replicate processors to
add performance vs. design a faster processor• Led to innovative organization tied to particular programming models
since “uniprocessors can’t keep going”– e.g., uniprocessors must stop getting faster due to limit of speed
of light: 1972, … , 1989, ... , 2008– Borders religious fervor: you must believe!– Fervor damped some when 1990s companies went out of
business: Thinking Machines, Kendall Square, ...• Argument instead is the “pull” of opportunity of scalable
performance, not the “push” of uniprocessor performance plateau• Recent advancement: Sun Niagara “T1 Coolthreads” => 8 in-order
simple execution cores. Switch thread on stalls...
ENGS 116 Lecture 15 4
Opportunities: Scientific Computing• Nearly Unlimited Demand (Grand Challenge):
App Perf (GFLOPS) Memory (GB)48 hour weather 0.1 0.172 hour weather 3 1Pharmaceutical design 100 10Global Change, Genome 1000 1000
(Figure 1-2, page 25, of Culler, Singh, Gupta [CSG97])
• Successes in some real industries: – Petroleum: reservoir modeling– Automotive: crash simulation, drag analysis, engine– Aeronautics: airflow analysis, engine, structural mechanics– Pharmaceuticals: molecular modeling– Entertainment: full length movies (“Finding Nemo”)
ENGS 116 Lecture 15 5
Flynn’s Taxonomy • SISD (Single Instruction Single Data)
– Uniprocessors• MISD (Multiple Instruction Single Data)
– ???• SIMD (Single Instruction Multiple Data)
– Examples: Illiac-IV, CM-2• Simple programming model• Low overhead• Flexibility• All custom integrated circuits
• MIMD (Multiple Instruction Multiple Data)– Examples: Sun Enterprise 10000, Cray T3D, SGI Origin
• Flexible• Use off-the-shelf microprocessors
ENGS 116 Lecture 15 6
Flyn
n’s
clas
sific
atio
n of
com
pute
r sys
tem
s
a) S
ISD
Com
pute
rStre
amM
emor
yM
odul
ePr
oces
sor
Uni
tD
ata
Con
trol
Uni
t
Inst
ruct
ion
Stre
am
Inst
ruct
ion
Stre
am
Mem
ory
Mod
ule
Mem
ory
Mod
ule
Mem
ory
Mod
ule
Mem
ory
Mod
ule
Proc
esso
rU
nit
Proc
esso
rU
nit
Proc
esso
rU
nit
Proc
esso
rU
nit
Stre
am 1
Dat
a
Stre
am 2
Dat
a
Stre
am 3
Dat
a
Stre
am 4
Dat
a
Inst
ruct
ion
Stre
amC
ontro
lU
nit
b) S
IMD
Com
pute
r
Inst
ruct
ion
Stre
am
Inst
ruct
ion
Stre
am
Inst
ruct
ion
Stre
am 3
Inst
ruct
ion
Stre
am 4
Mem
ory
Mod
ule
Mem
ory
Mod
ule
Mem
ory
Mod
ule
Proc
esso
rU
nit
Proc
esso
rU
nit
Proc
esso
rU
nit
Stre
am 2
Dat
a
Stre
am 3
Dat
a
Stre
am 4
Dat
a
Stre
am 2
Inst
ruct
ion
Stre
am 3
Inst
ruct
ion
Stre
am 4
Inst
ruct
ion
Mem
ory
Mod
ule
Proc
esso
rU
nit
Stre
am 1
Dat
aSt
ream
1In
stru
ctio
nC
ontro
lU
nit
Con
trol
Uni
t
Con
trol
Uni
t
Con
trol
Uni
t
c) M
IMD
Com
pute
rIn
stru
ctio
n St
ream
1In
stru
ctio
n St
ream
2
ENGS 116 Lecture 15 7
Parallel Framework
• Programming Model:– Multiprogramming: lots of jobs, no communication– Shared address space: communicate via memory– Message passing: send and receive messages– Data parallel: several agents operate on several data sets
simultaneously and then exchange information globally and simultaneously (shared or message passing)
• Communication Abstraction:– Shared address space: e.g., load, store, atomic swap– Message passing: e.g., send, receive library calls– Debate over this topic (ease of programming, scaling)
ENGS 116 Lecture 15 8
Diagrams of Shared and Distributed Memory Architectures
P P P P
Network
MM M
Network
P
MM
PP
MDistributed Memory
systemShared Memory System
ENGS 116 Lecture 15 9
Shared Address/Memory Multiprocessor Model
• Communicate via Load and Store– Oldest and most popular model
• Based on timesharing: processes on multiple processors vs. sharing single processor
• Process: a virtual address space and ≥ 1 thread of control– Multiple processes can overlap (share), but ALL threads
share a process address space• Writes to shared address space by one thread are visible to reads
of other threads– Usual model: share code, private stack, some shared heap,
some private heap (shared page table!)
ENGS 116 Lecture 15 10
Example: Small-Scale MP Designs
• Memory: centralized with uniform memory access (“UMA”) and bus interconnect, I/O
• Examples: Sun Enterprise, SGI Challenge, Intel SystemPro
Processor Processor Processor Processor
One or more levels
of cache
One or more levels
of cache
One or more levels
of cache
One or more levels
of cache
Main memory I/O System
ENGS 116 Lecture 15 11
SMP Interconnect
• Processors connected to memory AND to I/O
• Bus based: all memory locations equal access time so SMP = “Symmetric MP”
– Sharing limited BW as we add processors, I/O
• Crossbar: expensive to expand
• Multistage network: less expensive to expand than crossbar with more BW
• “Dance Hall” designs: all processors on the left, all memories on the right
ENGS 116 Lecture 15 12
Large-Scale MP Designs• Memory: distributed with nonuniform memory access (“NUMA”) and
scalable interconnect (distributed memory)
Interconnection Network Low Latency High Reliability
Processor+ cache
Memory I/O
Processor+ cache
Memory I/O
Processor+ cache
Memory I/O
Processor+ cache
Memory I/O
1 cycle
40 cycles
Processor+ cache
Memory I/O
Processor+ cache
Memory I/O
Processor+ cache
Memory I/O
Processor+ cache
Memory I/O
100 cycles
ENGS 116 Lecture 15 13
Shared Address Model Summary
• Each processor can name every physical location in the machine
• Each process can name all data it shares with other processes
• Data transfer via load and store
• Data size: byte, word, ... or cache blocks
• Uses virtual memory to map virtual to local or remote physical
• Memory hierarchy model applies: now communication moves data to local processor’s cache (as load moves data from memory to cache)
– Latency, BW, scalability of communication?
ENGS 116 Lecture 15 14
Message Passing Model
• Whole computers (CPU, memory, I/O devices) communicate as explicit I/O operations
– Essentially NUMA but integrated at I/O devices vs. memory system• Send specifies local buffer + receiving process on remote computer• Receive specifies sending process on remote computer + local buffer
to place data– Usually send includes process tag and receive has rule on tag:
match one, match any– Synchronization: when send completes, when buffer free, when request
accepted, receive waits for send• Send + receive => memory-memory copy, where each supplies local
address AND does pairwise synchronization!
ENGS 116 Lecture 15 15
Logical View of a Message Passing System
= processor
= process
= receiver’s buffer or queue= message
Port 2
Port 1
Processor 2Processor 1Communications Network
ENGS 116 Lecture 15 16
Message Passing Model
• Send + receive => memory-memory copy, synchronization on OS even on 1 processor
• History of message passing:– Network topology important because could only send to immediate
neighbor– Typically synchronous, blocking send & receive– Later DMA with non-blocking sends, DMA for receive into buffer until
processor does receive, and then data is transferred to local memory– Later SW libraries to allow arbitrary communication
• Example: IBM SP-2, RS6000 workstations in racks– Network Interface Card has Intel 960– 8x8 Crossbar switch as communication building block– 40 MB/sec per link
ENGS 116 Lecture 15 17
Message Passing Model
• Special type of Message Passing: Pipe• Pipe is a point-to-point connection• Data is pushed in one way by first process, and• Received, in-order, by the second process.• Data can stay in pipe indefinitely• Asynchronous, simplex communication• Advantages:
– Simple implementation– Works with Send-Receive calls– Good way of implementing producer consumers systems.– No latency, since communication is 1-way
ENGS 116 Lecture 15 18
Communication Models• Shared Memory
– Processors communicate with shared address space– Easy on small-scale machines– Advantages:
• Model of choice for uniprocessors, small-scale MPs• Ease of programming• Lower latency• Easier to use hardware controlled caching
• Message passing– Processors have private memories, communicate via messages– Advantages:
• Less hardware, easier to design• Focuses attention on costly non-local operations
• Can support either SW model on either HW base
ENGS 116 Lecture 15 19
How Do We Develop a Parallel Program?
• Convert a sequential program to a parallel program– Partition the program into tasks and combine tasks into processes– Coordinate data accesses, communication, and synchronization– Map the processes to processors
• Create a parallel program from scratch
ENGS 116 Lecture 15 20
Review: Parallel Framework
• Layers:– Programming Model:
• Multiprogramming: lots of jobs, no communication• Shared address space: communicate via memory• Message passing: send and receive messages• Data parallel: several agents operate on several data sets simultaneously and
then exchange information globally and simultaneously (shared or message passing)
– Communication Abstraction:• Shared address space: e.g., load, store, atomic swap• Message passing: e.g., send, receive library calls• Debate over this topic (ease of programming, scaling)
Programming ModelCommunication AbstractionInterconnection SW/OS Interconnection HW
ENGS 116 Lecture 15 21
Fundamental Issues
• Fundamental issues in parallel machines
1) Naming
2) Synchronization
3) Latency and Bandwidth
ENGS 116 Lecture 15 22
Fundamental Issue #1: Naming
• Naming:– what data is shared– how it is addressed– what operations can access data– how processes refer to each other
• Choice of naming affects code produced by a compiler; via load where just remember address or keep track of processor number and local virtual address for message passing
• Choice of naming affects replication of data; via load in cache memory hierarchy or via SW replication and consistency
ENGS 116 Lecture 15 23
Fundamental Issue #1: Naming
• Global physical address space: any processor can generate, address and access it in a single operation
• Global virtual address space: if the address space of each process can be configured to contain all shared data of the parallel program
• Segmented shared address space: locations are named < process number, address > uniformly for all processes of the parallel program
• Independent local physical addresses
ENGS 116 Lecture 15 24
Fundamental Issue #2: Synchronization
• To cooperate, processes must coordinate
• Message passing is implicit coordination with transmission or arrival of data
• Shared address system needs additional operations to explicitly coordinate: e.g., write a flag, awaken a thread, interrupt a processor
ENGS 116 Lecture 15 25
Fundamental Issue #3: Latency and Bandwidth
• Bandwidth– Need high bandwidth in communication– Cannot scale, but stay close– Match limits in network, memory, and processor– Overhead to communicate is a problem in many machines
• Latency– Affects performance, since processor may have to wait– Affects ease of programming, since requires more thought to
overlap communication and computation• Latency Hiding
– How can a mechanism help hide latency?– Examples: overlap message send with computation, prefetch
data, switch to other tasks
ENGS 116 Lecture 15 26
Small-Scale — Shared Memory
• Caches serve to:– Increase bandwidth
versus bus/memory– Reduce latency of
access– Valuable for both
private data and shared data
• What about cache coherence?
Processor Processor Processor Processor
One or more levels
of cache
One or more levels
of cache
One or more levels
of cache
One or more levels
of cache
Main memory I/O System
ENGS 116 Lecture 15 27
The Problem Cache Coherency
CPU
I/O
B
100
200
A
Memory
Cache
A’
B’ 200
100
CPU
I/OOutput Agives 100
Cache
A’
B’ 200
550
B
100
200
A
Memory
CPU
I/OInput
400 to B
Cache
200
A’
B’
100
B
100
400
A
Memory
(a) Cache andmemory coherent:A’ = A & B’ = B
(b) Cache andmemory incoherent:A’ ≠ A (A stale)
(c) Cache andmemory incoherent:B’ ≠ B (B’ stale)
ENGS 116 Lecture 15 28
What Does Coherency Mean?• Informally:
– “Any read must return the most recent write”– Too strict and too difficult to implement
• Better:– “Any write must eventually be seen by a read”– All writes are seen in proper order (“serialization”)
• Two rules to ensure this:– “If P writes x and P1 reads it, P’s write will be seen by P1 if
the read and write are sufficiently far apart”– Writes to a single location are serialized: seen in one order
• Latest write will be seen• Otherwise could see writes in illogical order
(could see older value after a newer value)
ENGS 116 Lecture 15 29
The cache-coherence problem for a single memory location (X), read and written by two processors (A and B).This example assumes a write-through cache.
Cache contents Cache contents Memory contents forTime Event for CPU A for CPU B location X
0 11 CPU A reads X 1 12 CPU B reads X 1 1 13 CPU A stores 0
into X0 1 0
ENGS 116 Lecture 15 30
Potential HW Coherency Solutions• Snooping Solution (Snoopy Bus):
– Send all requests for data to all processors– Processors snoop to see if they have a copy and respond
accordingly – Requires broadcast, since caching information is at processors– Works well with bus (natural broadcast medium)– Dominates for small-scale machines (most of the market)
• Directory-Based Schemes– Keep track of what is being shared in one centralized place– Distributed memory => distributed directory for scalability
(avoids bottlenecks)– Send point-to-point requests to processors via network– Scales better than Snooping– Actually existed BEFORE Snooping-based schemes
ENGS 116 Lecture 15 31
Basic Snoopy Protocols• Write Invalidate Protocol:
– Multiple readers, single writer– Write to shared data: an invalidate is sent to all caches which
snoop and invalidate any copies– Read Miss:
• Write-through: memory is always up-to-date• Write-back: snoop in caches to find most recent copy
• Write Broadcast Protocol (typically write through):– Write to shared data: broadcast on bus, processors snoop, and
update any copies– Read miss: memory is always up-to-date
ENGS 116 Lecture 15 32
An example of an invalidation protocol working on a snooping bus for a single cache block (X) with write-back caches.
Contents of Contents of Contents of memory Processor activity Bus activity CPU A's cache CPU B's cache location X
0CPU A reads X Cache miss for X 0 0CPU B reads X Cache miss for X 0 0 0CPU A writes a 1 to X Invalidation for X 1 0CPU B reads X Cache miss for X 1 1 1
ENGS 116 Lecture 15 33
An example of a write update or broadcast protocol working on a snooping bus for a single cache block (X) with write-back caches.
Contents of Contents of Contents of memoryProcessor activity Bus activity CPU A's cache CPU B's cache location X
0CPU A reads X Cache miss for X 0 0CPU B reads X Cache miss for X 0 0 0CPU A writes a 1 to X Write broadcast of X 1 1 1CPU B reads X 1 1 1
ENGS 116 Lecture 15 34
Basic Snoopy Protocols
• Write Invalidate versus Broadcast:– Invalidate requires one transaction per write-run (writing to one
location multiple times by the same CPU)– Invalidate uses spatial locality: one transaction per block– Broadcast has lower latency between write and read