11

Performance Measurements of CCR and MPI on Multicore Systems

Expanded from a Poster at Grid 2007 Austin TexasSeptember 21 2007

Xiaohong QiuResearch Computing UITS, Indiana University Bloomington IN

Geoffrey Fox, H. Yuan, Seung-Hee BaeCommunity Grids Laboratory, Indiana University Bloomington IN 47404

George Chrysanthakopoulos, Henrik Frystyk Nielsen

Microsoft Research, Redmond WA

Presented by Geoffrey Fox gcf@indiana.eduhttp://www.infomall.org

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/workflow 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

• http://msdn.microsoft.com/robotics/

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 UsedAMD4: HPxw9300 workstation, 2 AMD Opteron CPUs Processor 275 at 2.19GHz, 4 coresL2 Cache 4x1MB (summing both chips), Memory 4GB, XP Pro 64bit , Windows Server, Red HatC# Benchmark Computational unit: 1.388 µs

Intel4: Dell Precision PWS670, 2 Intel Xeon Paxville CPUs at 2.80GHz, 4 coresL2 Cache 4x2MB, Memory 4GB, XP Pro 64bitC# Benchmark Computational unit: 1.475 µs

Intel8a: Dell Precision PWS690, 2 Intel Xeon CPUs E5320 at 1.86GHz, 8 coresL2 Cache 4x4M, Memory 8GB, XP Pro 64bit C# Benchmark Computational unit: 1.696 µs

Intel8b: Dell Precision PWS690, 2 Intel Xeon CPUs E5355 at 2.66GHz, 8 coresL2 Cache 4x4M, Memory 4GB, Vista Ultimate 64bit, Fedora 7C# Benchmark Computational unit: 1.188 µs

Intel8c: Dell Precision PWS690, 2 Intel Xeon CPUs E5345 at 2.33GHz, 8 coresL2 Cache 4x4M, Memory 8GB, Red Hat 5.0, Fedora 7

Basic Performance of CCR

AMD4: 4 Core Number of Parallel Computations

(μs) 1 2 3 4 7 8

Spawned

Pipeline 1.76 4.52 4.4 4.84 1.42 8.54

Shift 4.48 4.62 4.8 0.84 8.94

Two Shifts 7.44 8.9 10.18 12.74 23.92

(MPI)

Pipeline 3.7 5.88 6.52 6.74 8.54 14.98

Shift 6.8 8.42 9.36 2.74 11.16

Exchange As Two Shifts

14.1 15.9 19.14 11.78 22.6

Exchange 10.32 15.5 16.3 11.3 21.38

CCR Overhead for a computation of 27.76 µs between messaging

Rendezvous

CCR Overhead for a computation of 29.5 µs between messaging

Rendezvous

Intel4: 4 Core Number of Parallel Computations

(μs) 1 2 3 4 7 8

Spawned

Pipeline 3.32 8.3 9.38 10.18 3.02 12.12

Shift 8.3 9.34 10.08 4.38 13.52

Two Shifts 17.64 19.32 21 28.74 44.02

MPI

Pipeline 9.36 12.08 13.02 13.58 16.68 25.68

Shift 12.56 13.7 14.4 4.72 15.94

Exchange AsTwo Shifts

23.76 27.48 30.64 22.14 36.16

Exchange 18.48 24.02 25.76 20 34.56

CCR Overhead for a computation of 23.76 µs between messaging

Rendezvous

Intel8b: 8 Core Number of Parallel Computations

(μs) 1 2 3 4 7 8

Spawned

Pipeline 1.58 2.44 3 2.94 4.5 5.06

Shift 2.42 3.2 3.38 5.26 5.14

Two Shifts 4.94 5.9 6.84 14.32 19.44

MPI

Pipeline 2.48 3.96 4.52 5.78 6.82 7.18

Shift 4.46 6.42 5.86 10.86 11.74

Exchange As Two Shifts

7.4 11.64 14.16 31.86 35.62

Exchange 6.94 11.22 13.3 18.78 20.16

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

0

5

10

15

20

25

30

0 2 4 6 8 10

AMD Exch

AMD Exch as 2 Shifts

AMD Shift

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

0

10

20

30

40

50

60

70

0 2 4 6 8 10

Intel Exch

Intel Exch as 2 Shifts

Intel Shift

Stages (millions)

Time Microseconds

Basic Performance of MPI for C and Java

MPI Exchange Latency in µs with 500,000 stages (20-30 µs computation between messaging)

Machine OS Runtime Grains Parallelism MPI Exchange Latency

Intel8c:gf12 Redhat MPJE Process 8 181

MPICH2 Process 8 40.0

MPICH2: Fast Process 8 39.3

Nemesis Process 8 4.21

Intel8c:gf20 Fedora MPJE Process 8 157

mpiJava Process 8 111

MPICH2 Process 8 64.2

Intel8b Vista MPJE Process 8 170

Fedora MPJE Process 8 142

Fedora mpiJava Process 8 100

Vista CCR Thread 8 20.2

AMD4 XP MPJE Process 4 185

Redhat MPJE Process 4 152

Redhat mpiJava Process 4 99.4

Redhat MPICH2 Process 4 39.3

XP CCR Thread 4 16.3

Intel4 XP CCR Thread 4 25.8

0 2 4 6 8 10

Stages (millions)

MPICH mpiJava MPJE MPI Shift Latency on AMD4

0

50

100

150

200

250

0 2000000 4000000 6000000 8000000 10000000 12000000

0

50

100

150

200

250

0 2000000 4000000 6000000 8000000 10000000 12000000

WindowsXP (MPJE)

RedHat (MPJE)

RedHat (mpiJava)

RedHat (MPICH2)

0 2 4 6 8 10

Stages (millions)

MPICH mpiJava MPJE MPI Exchange Latency on AMD4

0 2 4 6 8 10

Stages (millions)

MPICH Nemesis MPJE MPI Exchange Latency on Intel8c RedHat

Cache Line Interference

Cache Line Interference• Early implementations of our clustering algorithm

showed large fluctuations due to the cache line interference effect discussed here and on next slide in a simple case

• We have one thread on each core each calculating a sum of same complexity storing result in a common array A with different cores using different array locations

• 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 < 8 (64 bytes) with Windows– Note A is a double (8 bytes)– Less interference effect with Linux – especially Red Hat

Time µs versus Thread Array Separation (unit is 8 bytes)

1 4 8 1024 Machine

OS

Run Time Mean Std/

Mean Mean Std/

Mean Mean Std/

Mean Mean Std/

Mean Intel8b Vista C# CCR 8.03 .029 3.04 .059 0.884 .0051 0.884 .0069 Intel8b Vista C# Locks 13.0 .0095 3.08 .0028 0.883 .0043 0.883 .0036 Intel8b Vista C 13.4 .0047 1.69 .0026 0.66 .029 0.659 .0057 Intel8b Fedora C 1.50 .01 0.69 .21 0.307 .0045 0.307 .016 Intel8a XP CCR C# 10.6 .033 4.16 .041 1.27 .051 1.43 .049 Intel8a XP Locks C# 16.6 .016 4.31 .0067 1.27 .066 1.27 .054 Intel8a XP C 16.9 .0016 2.27 .0042 0.946 .056 0.946 .058 Intel8c Red Hat C 0.441 .0035 0.423 .0031 0.423 .0030 0.423 .032 AMD4 WinSrvr C# CCR 8.58 .0080 2.62 .081 0.839 .0031 0.838 .0031 AMD4 WinSrvr C# Locks 8.72 .0036 2.42 0.01 0.836 .0016 0.836 .0013 AMD4 WinSrvr C 5.65 .020 2.69 .0060 1.05 .0013 1.05 .0014 AMD4 XP C# CCR 8.05 0.010 2.84 0.077 0.84 0.040 0.840 0.022 AMD4 XP C# Locks 8.21 0.006 2.57 0.016 0.84 0.007 0.84 0.007 AMD4 XP C 6.10 0.026 2.95 0.017 1.05 0.019 1.05 0.017

Cache Line Interference

• Note measurements at a separation of 8 (and values between 8 and 1024 not shown) are essentially identical

• Measurements at 7 (not shown) are higher than that at 8 (except for Red Hat which shows essentially no enhancement at X<8)

• If effects due to co-location of thread variables in a 64 byte cache line, the array must be aligned with cache boundaries

– In early implementations we found poor X=8 performance expected in words of A split across cache lines

Clustering Problem

Deterministic Annealing • See K. Rose, "Deterministic Annealing for Clustering,

Compression, Classification, Regression, and Related Optimization Problems," Proceedings of the IEEE, vol. 80, pp. 2210-2239, 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

• For more details see – http://grids.ucs.indiana.edu/ptliupages/presentations/Grid

2007PosterSept19-07.ppt and

– http://grids.ucs.indiana.edu/ptliupages/presentations/PC2007/PC07BYOPA.ppt

Parallel MulticoreDeterministic Annealing Clustering

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0 0.5 1 1.5 2 2.5 3 3.5 4

Parallel Overheadon 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) due to threadruntime fluctuations

10 Clusters

20 Clusters

Parallel Multicore Deterministic Annealing Clustering

0.000

0.050

0.100

0.150

0.200

0.250

0 5 10 15 20 25 30 35

#cluster

over

head

“Constant1”

Increasing number of clusters decreases communication/memory bandwidth overheads

Parallel Overhead for large (2M points) Indiana Census clustering on 8 Threads Intel 8bThis fluctuating overhead due to 5-10% runtime fluctuations between threads

Scaled Speed up Tests• The full clustering algorithm involves different values of the

number of clusters NC as computation progresses• The amount of computation per data point is proportional to NC

and so overhead due to memory bandwidth (cache misses) declines as NC increases

• We did a set of tests on the clustering kernel with fixed NC

• Further we adopted the scaled speed-up approach looking at the performance as a function of number of parallel threads with constant number of data points assigned to each thread– This contrasts with fixed problem size scenario where the number of

data points per thread is inversely proportional to number of threads• We plot Run time for same workload per thread divided by

number of data points multiplied by number of clusters multiped by time at smallest data set (10,000 data points per thread)

• Expect this normalized run time to be independent of number of threads if not for parallel and memory bandwidth overheads– It will decrease as NC increases as number of computations per points

fetched from memory increases proportional to NC

Intel 8b C with 1 Cluster: Vista Scaled Run Time for Clustering Kernel

• Note the smallest dataset has highest overheads as we increase the number of threads– Not clear why this is

Number of Threads

Scaled Run Time

Intel 8b C with 80 Clusters: Vista Scaled Run Time for Clustering Kernel

• As we increase number of clusters, the effects at 10,000 data points decrease

Number of Threads

Scaled Run Time

Intel 8b C# with 1 Cluster: Vista Scaled Run Time for Clustering Kernel

• C# is similar to C with larger effects

Number of Threads

Scaled Run Time

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 Scaled Run Time for Clustering Kernel• C# is similar to C with larger effects

Number of Threads

Scaled Run Time

AMD4 C with 1 Cluster: XP Scaled Run Time for Clustering Kernel

• This is significantly more stable than Intel runs and shows little or no memory bandwidth effect

Number of Threads

Scaled Run Time

AMD4 C# with 1 Cluster: XP Scaled Run Time for Clustering Kernel

• This is significantly more stable than Intel C# 1 Cluster runs

Number of Threads

Scaled Run Time

AMD4 C# with 80 Clusters: XP Scaled Run Time for Clustering Kernel

• This is broadly similar to 80 Cluster Intel C# runs unlike one cluster case that was very different

Number of Threads

Scaled Run Time

AMD4 C# with 1 Cluster: Windows Server Scaled Run Time for Clustering Kernel

• This is significantly more stable than Intel C# runs

Number of Threads

Scaled Run Time

AMD4 C# with 80 Clusters: Windows Server Scaled Run Time for Clustering Kernel

• Curiously run time decreases a bit as number of threads increases in some AMD4 scenarios

Number of Threads

Scaled Run Time

Intel 8c C with 1 Cluster: Red Hat Scaled Run Time for Clustering Kernel

• Deviations from “perfect” scaled speed-up are much less for Red Hat than for Windows

Number of Threads

Scaled Run Time

Intel 8c C with 80 Clusters: Red Hat Scaled Run Time for Clustering Kernel

• Deviations from “perfect” scaled speed-up are much less for Red Hat

Number of Threads

Scaled 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

AMD4 with 1 Cluster: Windows Server Run Time Fluctuations for Clustering Kernel

• This is average of standard deviation of run time of the 8 threads between messaging synchronization points

• XP (not shown) is similar

Number of Threads

Standard Deviation/Run Time

Intel 8c 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

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

4242

0

50

100

150

200

250

300

350

1 10 100 1000 10000

Round trips

Av

era

ge

ru

n t

ime

(m

icro

se

co

nd

s)

Timing of HP Opteron Multicore as a function of number of simultaneous two-way service messages processed (November 2006 DSS Release)

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 improvedHere we see 10 increasing to 30 as algorithm progresses