Are New Languages Necessary for Manycore?

1
Are New Languages Necessary for Manycore?

• David I. August
• Department of Computer Science
• Princeton University

2
THIS is the Problem!
?
2004
SPEC CPU INTEGER PERFORMANCE
TIME
3
Why New Multicore Languages Will Fail

1. Money is earned by relieving customer pain
2. The Market
3. Legacy, Legacy, Legacy
4. Programmers adopt new programming models
5. Parallel programming is more difficult
6. Parallel programming models have longevity issues
7. Automatic Thread Extraction (ATE)

4
Automatic Thread Extraction
That isn't to say we are parallelizing arbitrary
C code, that's a fool's errand! Richard Lethin
Compiler cant determine a tree from a graph
Burton Smith
Compiler cant determine dependences without
type information. Even then Burton Smith
Decades of automatic parallelization work has
been a failure James Larus
All that icky pointer chasing code... Tim
Mattson
5
How To Get Parallelism For Multicore?

• Nine months ago, with an open mind
• A priori select ALL C programs from SPEC CINT
  2000
• Our objective function (in priority order)
• Extract meaningful parallelism
• Prefer automatic over manual
• Minimize impact to the programmer when manual

6
Our Results
Benchmark Threads at Peak Speedup LOCs Changed
164.gzip 32 29.91 26
175.vpr 15 3.59 1
176.gcc 16 5.06 17
181.mcf 32 2.84 0
186.crafty 32 25.18 9
197.parser 32 24.50 2
253.perlbmk 5 1.21 0
254.gap 10 1.94 1
255.vortex 32 4.92 0
256.bzip2 12 6.72 0
300.twolf 8 2.06 1
GEOMEAN 17 5.54
ARITHMEAN 20 9.81
M.L.O.P. 5 Generations 32 Cores 5.3x Speedup
7
Our Recipe

• Recent Compiler Technology
• Decoupled Software Pipelining (DSWP) MICRO 05
• Parallel-Stage DSWP (PS-DSWP)
• Speculative DSWP (Spec-DSWP) PACT 07
• Existing Technology Speculative DOALL, TLS
• Targeted Memory Profiling
• Procedure Boundary Elimination PLDI 06
• Hardware Support
• Compiler-Controlled Speculation
• Streaming Communication MICRO 06

8
Typical Example 197.parser
Threads run on multicore model with Itanium 2
cores.
Find English Sentences
Parse Sentences (95)
Emit Results
DSWP
PS-DSWP (Spec DOALL Middle Stage)
9
What We Learned

• A new way of thinking about dependences
• Go With the Flow
• TLP is easier to extract than ILP
• A holistic approach is better
• A limitation exists in the sequential model
• Determinism

10
Determinism A Double Edged Sword
while(ltcondgt) ltworkgt x Rand() ltworkgt
int Rand() state f2(state) return f1(state)
1
2
3
4
DOALL
1
2
3
4
SEQUENTIAL
56 LOCs in 11 programs 22 annotations Only 2
programs needed more Most common culprit Custom
Allocators
11
What about Manycore?

• Multicore
• New languages arent necessary
• Legacy code easily adjusted
• Manycore
• Implicitly Parallel Sequential Programming
• No optimization for sequential (custom
  allocators)
• Points of non-determinism specified
• Parallel algorithms in sequential codes
• Debuggability, Understandability, Sanity

12
The Answer Originates with ATE

• The Old Way PL folks would write languages,
  Architecture folks would make HW, andCompiler
  folks would dutifully connect the two.
• This will fail for Manycore
• Unduly burden the programmer
• Performance will suffer
• Theres a New Way

13
DO NOT POST ANYTHING AFTER THIS SLIDE
14
How Code Was Transformed
Benchmark LOC (All) LOC (Model) Model Techniques Compiler Techniques Applied
164.gzip 26 2 Y-Branch TLS Memory, DSWP
175.vpr 1 1 PURE Alias, Value, Control Spec, TLS Mem, DSWP
176.gcc 17 7 PURE Alias Control Spec, TLS MEM, DSWP
181.mcf 0 0 Alias, Silent Store, Control Spec, TLS Mem, DSWP, Nested
186.crafty 9 9 PURE TLS Mem, DSWP, Nested
197.parser 2 2 PURE TLS Mem, DSWP
253.perlbmk 0 0 Alias, Control, Value Spec, DSWP
254.gap 1 1 PURE TLS Memory, DSWP, Alias Spec
255.vortex 0 0 Alias Value Spec, TLS Mem, DSWP
256.bzip2 0 0 TLS Memory, DSWP
300.twolf 1 1 PURE Alias Control Spec, TLS Mem, DSWP
15
PURE
16
Y-Branch
17
SPEC 2006 403.gcc
Threads run on multicore model with Itanium 2
cores.