Analysis and Performance Results of a Molecular Modeling Application on Merrimac

1
Analysis and Performance Results of a Molecular
Modeling Application on Merrimac

• Mattan Erez, Jung Ho Ahn, Ankit Garg, William J.
  Dally, Eric Darve (Stanford Univ.)
• Presented by Jiahua He

2
Content

• Background
• Motivation
• Merrimac Architecture
• Application StreamMD
• Performance Evaluation
• Conclusions and Discussions

3
Parallel Architectures

• Flynn taxonomy
• SISD (sequential machine), SIMD, MIMD, MISD (no
  commercial system)
• SIMD
• Processor-array machine
• Single processor vector machine
• MIMD
• PVP, SMP, DSM, MPP, Cluster

4
Processor-Array Machine

• Control processor issues instructions
• All processors in processor array operate the
  instructions in lock-step
• Distributed memory
• Need permutation if data not aligned

5
Vector Machine

• A processor can do element-wise operations on
  entire vectors with a single instruction
• Dominated the high performance computer market
  for about 15 years
• Overtaken by MPP in 90s
• Re-emerges in recent years (Earth Simulator and
  Cray X1)

6
MPP and Cluster

• Distributed memory
• Each processor/node has its own private memory
• Nodes may be SMPs
• MIMD
• Nodes execute different instructions
  asynchronously
• Nodes communicate and synchronize by
  interconnection network

7
Earth Simulator

• Vector machine re-emerges
• Rmax 36 GFLOPS gt Rmax sum of top 10
• Vector machines focused on powerful processors
• MPP or Cluster focused on large-scale
  clustering
• Trend merge the above two

8
Content

• Background
• Motivation
• Merrimac Architecture
• Application StreamMD
• Performance Evaluation
• Conclusions and Discussions

9
Modern VLSI Technology

• Arithmetic is cheap
• 100s of GFLOPS/chip today
• TFLOPS in 2010
• Bandwidth is expensive
• General purpose processor architectures have not
  adapted to this change

10
Stream Processor

• One control unit and 100s of FPUs
• 90nm fabrication process 64-bit, 0.5mm2, 50pJ
• Deep register hierarchy with high local bandwidth
• Match bandwidth demands and tech. limits
• Stream sequence of data objects
• Expose large amounts of data parallelism
• Keep 100s of FPUs per processor busy
• Hide long latencies of memory operations

11
Stream Processor (contd)

• Expose multiple levels of locality
• Short term producer-consumer locality (LRF)
• Long term producer-consumer locality (SRF)
• Cannot be exploited by caches no reuse, no
  spatial locality
• Scalable
• 128GFLOPS processor
• 16 node 2TFLOPS single board workstation
• 16,384 node 2PFLOPS supercomputer with 16
  cabinets

12
Content

• Background
• Motivation
• Merrimac Architecture
• Application StreamMD
• Performance Evaluation
• Conclusions and Discussions

13
Merrimac Processor

• Scalar core (1)
• Perform control code and issue stream
  instructions
• Arithmetic clusters (16)
• 64-bit multiply-accumulate (MADD) FPUs (4)
• Execute the same VLIW instruction
• Local register file (LRF) per FPU (192 words)
• Short term producer-consumer locality in a kernel
• Stream register file (SRF) per cluster (8K words)
• Long term producer-consumer locality across
  kernels
• Staging area for memory data transfer to hide
  latencies

14
Architecture of Merrimac
15
Stream Programming Model

• Cast the computation as a collection of streams
  passing through a series of computational
  kernels.
• Data parallelism
• Across stream elements
• Task parallelism
• Across kernels

16
Memory System

• A stream memory instruction transfers entire
  stream
• Address generator (2)
• 8 single-word addresses every cycle
• Stride access or gathers/scatters pattern
• Cache (128K words, 64GB/s)
• Directly interface with external DRAM and network
• External DRAM (2GB, 38.4GB/s)
• Single-word remote memory access
• Scatter-add operation

17
Interconnection Network (Fat Tree)
18
Content

• Background
• Motivation
• Merrimac Architecture
• Application StreamMD
• Performance Evaluation
• Conclusions and Discussions

19
Molecular Dynamics

• Explore kinetic and thermodynamic properties of
  molecular system by simulating atomic models
• water ?? water molecule
• protein ?? water molecules
• GROMAC fastest MD code available
• Cut-off distance approximation
• Neighbor list (neighbors within rc)

20
StreamMD

• Single kernel non-bonded interaction between all
  atom pairs of a molecule and one of its neighbor
• Pseudo code
• c_positions gather(positions, i_central)
• n_positions gather(positions, i_neighbor)
• partial_forces compute_force(c_positions,
  n_positions)
• forces scatter_add(partial_forces, i_forces)

21
Latency Tolerance

• Pipeline the requests
• To amortize long initial latency
• By issuing a memory op of long stream
• Hide memory ops with computations
• Concurrently executing memory ops and kernel
  computations
• Strip-mining
• Large data set ? smaller strips
• Outer loop (done manually)

22
Parallelism

• 4 variants to exploit parallelism
• Also implemented on Pentium 4 for comparison

23
Expanded Variant

• Simplest version
• Fully expand the interaction list
• For each cluster per iteration
• Read 2 interacting molecules
• Produce 2 partial forces

24
Fixed Variant

• Fixed-length neighbor list of length L
• For each cluster
• Read a central molecule once every L iteration
• Read a neighbor molecule each iteration
• Partial forces of central molecule are reduced in
  cluster
• Repeat central molecule
  in i_central
• Add dummy_neighbor
  in i_neighbor if needed

25
Variable Variant

• Variable-length neighbor list
• Process inputs and produce outputs at a different
  rate for each cluster
• Merrimacs inter-cluster communication
• Conditional streams mechanism
• Indexable SRF
• Instructions to read new central position and
  write partial forces are issued on every
  iteration but with a condition
• Slight overhead of unexecuted instructions

26
Duplicated Variant

• Fixed-length neighbor list
• Duplicate all interaction calculations
• Reduce complete force for central molecule within
  cluster
• No partial force for neighbor molecule is written
  out

27
Locality

• Only short term producer consumer locality within
  a single kernel
• Computing partial forces
• Internal reduction of forces within a cluster
• Computation/bandwidth trade-off
• Extra computation for interactions with dummy
  molecules fixed variant
• Extreme case duplicated variant
• Need more sophisticated schemes (discuss later)

28
Content

• Background
• Motivation
• Merrimac Architecture
• Application StreamMD
• Performance Evaluation
• Conclusions and Discussions

29
Experiment Setup

• Single-node experiments
• 900 water-molecule system
• Cycle-accurate simulator of Merrimac
• 4 variants of StreamMD
• Pentium 4 version
• Latest version of GROMACS
• Fully hand optimized
• Single precision SSE

30
Latency Tolerance

• Snippet of the execution of duplicated variant
• Left column
• Kernel computations
• Right column
• Memory operations
• Perfect overlap of memory and computation

31
Locality

• Arithmetic intensities
• fixed and variable depend on data set
• Small diff ? compiler efficiently utilize
  register hierarchy
• Reference percentages
• Nearly all to LRFs
• Small diff ? use SRF just as staging area for
  memory

32
Performance

• variable outperforms expanded by 84, fixed
  by 26, duplicated by 119, and Pentium 4 by
  a factor of 13.2
• 38.8 GFLOPS is 50 of the optimal solution GFLOPS

33
Automatic Optimizations

• Communication scheduling
• SRF decouples memory from computation
• Loop unrolling and software pipelining
• Improve execution rate by 83
• Stream scheduling
• SRF is software managed
• Capture long term producer
  consumer locality by
  intelligent eviction

34
Computation/bandwidth Trade-off

• Blocking technique
• Group molecules into cubic clusters of size r3
• Pave the rc3 (cut-off radius) sphere with cubic
  clusters
• Memory bandwidth requirement scales as O(r-3)
• Extra computation between rc and rc2sqrt(3)r
• Minimum occurs at about 3 molecules per cluster
  (1.43)

35
Content

• Background
• Motivation
• Merrimac Architecture
• Application StreamMD
• Performance Evaluation
• Conclusions and Discussions

36
Conclusions

• Reviewed the architecture and organization of
  Merrimac
• Presented app StreamMD, implemented 4 variants
  and evaluated their performance
• Compared Merrimacs suitability for molecular
  dynamic app against a conventional Pentium 4
  processor

37
Special Applications?

• Merrimac is tuned for scientific applications
• Programming model
• A collection of streams pass through a series of
  computational kernels
• Need large data level parallelism to utilize the
  FPUs
• Task parallelism just can be exploit across nodes
  because of SIMD

38
Easy to Program?

• Effective automatic compilation
• Communication scheduling and stream scheduling
  (shown earlier)
• Highly optimized code for conventional processors
  is often written in assembly
• Performance of different StreamMD variants vary
  only by 2 fold (shown earlier)

39
Compare with Supercomputer?

• Only comparing with Pentium 4 seems not
  convincing
• MDGRAPE-3 of Protein Explorer can achieve 165
  GFLOPS out of 200 GFLOPS (peak)
• But it is special purpose design
• How about vector machines?
• Lack of standard benchmarks

40
Thanks! And questions?