1 Performance of a Multi-Paradigm Messaging Runtime on Multicore Systems Poster at Grid 2007 Omni Austin Downtown Hotel Austin Texas September Xiaohong Qiu Research Computing UITS, Indiana University Bloomington IN Geoffrey Fox, H. Yuan, Seung-Hee Bae Community Grids Laboratory, Indiana University Bloomington IN George Chrysanthakopoulos, Henrik Frystyk Nielsen Microsoft Research, Redmond WA Presented by Geoffrey Fox
2 Motivation Exploring possible applications for tomorrow’s multicore chips (especially clients) with 64 or more cores (about 5 years) One plausible set of applications is data-mining of Internet and local sensors Developing Library of efficient data-mining algorithms –Clustering (GIS, Cheminformatics) and Hidden Markov Methods (Speech Recognition) Choose algorithms that can be parallelized well
3 Approach Need 3 forms of parallelism –MPI Style –Dynamic threads as in pruned search –Coarse Grain functional parallelism Do not use an integrated language approach as in Darpa HPCS Rather use “mash-ups” or “workflow” to link together modules in optimized parallel libraries Use Microsoft CCR/DSS where DSS is mash-up model built from CCR and CCR supports MPI or Dynamic threads
4 Microsoft CCR Supports exchange of messages between threads using named ports FromHandler: Spawn threads without reading ports Receive: Each handler reads one item from a single port MultipleItemReceive: Each handler reads a prescribed number of items of a given type from a given port. Note items in a port can be general structures but all must have same type. MultiplePortReceive: Each handler reads a one item of a given type from multiple ports. JoinedReceive: Each handler reads one item from each of two ports. The items can be of different type. Choice: Execute a choice of two or more port-handler pairings Interleave: Consists of a set of arbiters (port -- handler pairs) of 3 types that are Concurrent, Exclusive or Teardown (called at end for clean up). Concurrent arbiters are run concurrently but exclusive handlers are
Preliminary Results Parallel Deterministic Annealing Clustering in C# with speed-up of 7 on Intel 2 quadcore systems Analysis of performance of Java, C, C# in MPI and dynamic threading with XP, Vista, Windows Server, Fedora, Redhat on Intel/AMD systems Study of cache effects coming with MPI thread-based parallelism Study of execution time fluctuations in Windows (limiting speed-up to 7 not 8!)
Machines Used AMD4: HPxw9300 workstation, 2 AMD Opteron CPUs Processor 275 at 2.19GHz, 4 cores L2 Cache 4x1MB (summing both chips), Memory 4GB, XP Pro 64bit, Windows Server, Red Hat C# Benchmark Computational unit: µs Intel4: Dell Precision PWS670, 2 Intel Xeon Paxville CPUs at 2.80GHz, 4 cores L2 Cache 4x2MB, Memory 4GB, XP Pro 64bit C# Benchmark Computational unit: µs Intel8a: Dell Precision PWS690, 2 Intel Xeon CPUs E5320 at 1.86GHz, 8 cores L2 Cache 4x4M, Memory 8GB, XP Pro 64bit C# Benchmark Computational unit: µs Intel8b: Dell Precision PWS690, 2 Intel Xeon CPUs E5355 at 2.66GHz, 8 cores L2 Cache 4x4M, Memory 4GB, Vista Ultimate 64bit, Fedora 7 C# Benchmark Computational unit: µs Intel8c: Dell Precision PWS690, 2 Intel Xeon CPUs E5345 at 2.33GHz, 8 cores L2 Cache 4x4M, Memory 8GB, Red Hat 5.0, Fedora 7
AMD4: 4 CoreNumber of Parallel Computations (μs) Spawned Pipeline Shift Two Shifts (MPI) Pipeline Shift Exchange As Two Shifts Exchange CCR Overhead for a computation of µs between messaging Rendez vous
CCR Overhead for a computation of 29.5 µs between messaging Rendez vous Intel4: 4 CoreNumber of Parallel Computations (μs) Spawned Pipeline Shift Two Shifts MPI Pipeline Shift Exchange As Two Shifts Exchange
CCR Overhead for a computation of µs between messaging Rendez vous Intel8b: 8 CoreNumber of Parallel Computations (μs) Spawned Pipeline Shift Two Shifts MPI Pipeline Shift Exchange As Two Shifts Exchange
MPI Exchange Latency in µs with 500,000 stages (20-30 µs computation between messaging) MachineOSRuntimeGrainsParallelismMPI Exchange Latency Intel8c:gf12RedhatMPJEProcess8181 MPICH2Process840.0 MPICH2: FastProcess839.3 NemesisProcess84.21 Intel8c:gf20FedoraMPJEProcess8157 mpiJavaProcess8111 MPICH2Process864.2 Intel8bVistaMPJEProcess8170 FedoraMPJEProcess8142 FedorampiJavaProcess8100 VistaCCRThread820.2 AMD4XPMPJEProcess4185 RedhatMPJEProcess4152 RedhatmpiJavaProcess499.4 RedhatMPICH2Process439.3 XPCCRThread416.3 Intel4XPCCRThread425.8
Overhead (latency) of AMD4 PC with 4 execution threads on MPI style Rendezvous Messaging for Shift and Exchange implemented either as two shifts or as custom CCR pattern Stages (millions) Time Microseconds
Overhead (latency) of Intel8b PC with 8 execution threads on MPI style Rendezvous Messaging for Shift and Exchange implemented either as two shifts or as custom CCR pattern Stages (millions) Time Microseconds
MPICH mpiJava MPJE MPI Exchange Latency on AMD Stages (millions)
One thread on each core Thread i stores sum in A(i) is separation 1 – no variable access interference but cache line interference Thread i stores sum in A(X*i) is separation X Serious degradation if X < 64 bytes (8 words) and Vista or XP A is a double (8 bytes) Cache Line Interference
Deterministic Annealing See K. Rose, "Deterministic Annealing for Clustering, Compression, Classification, Regression, and Related Optimization Problems," Proceedings of the IEEE, vol. 80, pp , November 1998 Parallelization is similar to ordinary K-Means as we are calculating global sums which are decomposed into local averages and then summed over components calculated in each processor Many similar data mining algorithms (such as annealing for E-M expectation maximization) which have high parallel efficiency and avoid local minima
Clustering by Deterministic Annealing Use Physics Analogy for Clustering
Deterministically find cluster centers y j using “mean field approximation” – could use slower Monte Carlo
Annealing avoids local minima
Parallel Multicore Deterministic Annealing Clustering Parallel Overhead on 8 Threads Intel 8b Speedup = 8/(1+Overhead) 10000/(Grain Size n = points per core) Overhead = Constant1 + Constant2/n Constant1 = 0.05 to 0.1 (Client Windows) 10 Clusters 20 Clusters
Parallel Multicore Deterministic Annealing Clustering “Constant1” Increasing number of clusters decreases communication/memory bandwidth overheads Parallel Overhead for large (2M points) Indiana Census clustering on 8 Threads Intel 8b
Intel 8b C# with 1 Cluster: Vista Scaled Run Time for Clustering Kernel Run time for same workload per thread normalized by number of data points Expect Run Time independent of Number of threads if not for parallel and memory bandwidth overheads Work per data point proportional to number of clusters Number of Threads Run Time Secs
Intel 8b C# with 80 Clusters: Vista Scaled Run Time for Clustering Kernel Work per data point proportional to number of clusters so memory bandwidth and parallel overheads decrease as # clusters increase Number of Threads Run Time Secs
Intel 8c C with 80 Clusters: Redhat Run Time Fluctuations for Clustering Kernel This is average of standard deviation of run time of the 8 threads between messaging synchronization points Number of Threads Standard Deviation/Run Time
Intel 8c C with 80 Clusters: Redhat Scaled Run Time for Clustering Kernel Work per data point proportional to number of clusters so memory bandwidth and parallel overheads decrease as # clusters increase Number of Threads Run Time Secs
Intel 8b C# with 1 Cluster: Vista Run Time Fluctuations for Clustering Kernel This is average of standard deviation of run time of the 8 threads between messaging synchronization points Number of Threads Standard Deviation/Run Time
Intel 8b C# with 80 Clusters: Vista Run Time Fluctuations for Clustering Kernel This is average of standard deviation of run time of the 8 threads between messaging synchronization points Number of Threads Standard Deviation/Run Time
DSS Section We view system as a collection of services – in this case –One to supply data –One to run parallel clustering –One to visualize results – in this by spawning a Google maps browser –Note we are clustering Indiana census data DSS is convenient as built on CCR
PC07Intro 30 Timing of HP Opteron Multicore as a function of number of simultaneous two- way service messages processed (November 2006 DSS Release) CGL Measurements of Axis 2 shows about 500 microseconds – DSS is 10 times better DSS Service Measurements
Clustering algorithm annealing by decreasing distance scale and gradually finds more clusters as resolution improved Here we see increasing to 30 as algorithm progresses