De gemiddelde GPU bevat tegenwoordig meer PK's dan de CPU. Naar aanleiding hiervan komen er steeds meer mogelijkheden om computationele problemen te verplaatsen van de CPU naar de GPU. Deze presentatie zal een inleiding zijn hoe je dit in Java kunt doen met behulp van Jogamp JoCL. Aan de hand van enkele simpele problemen wordt aangetoond wanneer een GPU beter ingezet kan worden dan een CPU en vice versa. Dit is ook een van de speerpunten in Java 9 (Project Sumatra) wat o.a. JoCL als inspiratie gebruikt.

1. Java GPU Computing
Maarten Steur & Arjan Lamers

2. ● Overzicht OpenCL
● Simpel voorbeeld
● Casus
● Tips & tricks
● Vragen

3. Waarom GPU Computing

4. Afkortingen
● CPU, GPU, APU
● Khronos: OpenCL, OpenGL
● Nvidia: CUDA
● JogAmp JOCL, JavaCL, JOCL

5. GPU vergeleken met CPU
● Veel simpele cores
● Veel high bandwidth geheugen
● Intel core i7 GeForce GT 650M
8 cores 384 cores
180 Gflops 650 Gflops

6. Programmeer model
● Definieer stream (flow)
● Run in parallel

7. Gebruik
● Algorithme:
– Hoge Concurrency
– Partitioneerbaar
● Maar:
– Extra latency door on- en offloaden op
de GPU
– Extra complexiteit

8. Componenten

9. Componenten

10. Voorbeeld (MacBook Pro)
Platform name: Apple
Platform profile: FULL_PROFILE
Platform spec version: OpenCL 1.2
Platform vendor: Apple
Device 16925696 HD Graphics 4000
Driver:1.2(Aug 17 2014 20:29:07)
Max work group size:512
Global mem size: 1073741824
Local mem size: 65536
Max clock freq: 1200
Max compute units: 16
Device 16918272 GeForce GT 650M
Driver:8.26.28 310.40.55b01
Max work group size:1024
Global mem size: 1073741824
Local mem size: 49152
Max clock freq: 900
Max compute units: 2
Device 4294967295 Intel(R) Core(TM) i7-3720QM CPU @
2.60GHz
Driver:1.1
Max work group size:1024
Global mem size: 17179869184
Local mem size: 32768
Max clock freq: 2600
Max compute units: 8

11. Work & Memory

12. Application / Kernel
● Schrijf .cl files in C variant
● Kernels zijn de 'publieke' functies
● Java Bytecode
– Aparapi (OpenCL)
– RootBeer (CUDA)

13. Disclaimer

14. Parallel sort
kernel void sort(global const float* in, global float* out, int size) {
int i = get_global_id(0); // current thread
float id = in[i];
int pos = 0;
for (int j=0;j<size;j++)
{
float jd = in[j];
// in[j] < in[i] ?
bool smaller = (jx < ix) || (jx == ix && j < i);
pos += (smaller)?1:0;
}
out[pos] = id;
}

15. Java GPU Computing
CLContext globalContext = CLContext.create();
CLDevice device = globalContext.getMaxFlopsDevice(Type.GPU);
CLContext context = CLContext.create(device);
CLCommandQueue queue = device.createCommandQueue();
CLProgram program =
context.createProgram(
First8GpuComputing.class.getResourceAsStream("MyTask.cl")
).build();
Je kunt ook builden voor specifieke devices: build(device)

16. Java GPU Computing
CLBuffer<FloatBuffer> inBuffer = context.createFloatBuffer(
input.length , READ_ONLY);
CLBuffer<FloatBuffer> outBuffer = context.createFloatBuffer(
input.length, WRITE_ONLY);
mapToBuffer(inBuffer.getBuffer(), workLoad);

17. Java GPU Computing
CLBuffer<FloatBuffer> inBuffer = context.createFloatBuffer(
input.length , READ_ONLY);
CLBuffer<FloatBuffer> outBuffer = context.createFloatBuffer(
input.length, WRITE_ONLY);
mapToBuffer(inBuffer.getBuffer(), workLoad);
CLKernel kernel = program.createCLKernel("MyTask");
kernel.putArgs(inBuffer, outBuffer).putArg(workLoad.length);

18. Java GPU Computing
CLBuffer<FloatBuffer> inBuffer = context.createFloatBuffer(
input.length , READ_ONLY);
CLBuffer<FloatBuffer> outBuffer = context.createFloatBuffer(
input.length, WRITE_ONLY);
mapToBuffer(inBuffer.getBuffer(), workLoad);
CLKernel kernel = program.createCLKernel("MyTask");
kernel.putArgs(inBuffer, outBuffer).putArg(workLoad.length);
queue.putWriteBuffer(inBuffer, false)
.put1DRangeKernel(kernel, 0, globalWorkSize, localWorkSize)
.putReadBuffer(outBuffer, true);
FloatBuffer output = outBuffer.getBuffer();

19. Praktijkcasus

20. Praktijk casus
● Rekeninstrument ter ondersteuning van
de Programmatische Aanpak Stikstof.
● http://www.aerius.nl

21. Praktijk casus

22. Praktijk casus

23. Tips & tricks
● CL beheer
– getResourceAsStream()?
– Java constanten → #define
– Locale? Oops!

24. Tips & tricks
● Unit testen
– Aparte test kernels
– Test cases in batches
kernel void testDifficultCalculation(const int testCount,
global const double* distance, global double* results) {
const int testId = get_global_id(0);
if (testId < testCount) {
results[testId] = difficultCalculation(distance[testId]);
}
}

25. Direct memory management
● -XX:MaxDirectMemorySize=??M
● ByteBuffer.allocateDirect(int capacity)
– Max 2GB per buffer
● Garbage collection te laat
– Getriggered door heap collection
– Handmatig vrijgeven
– ((sun.nio.ch.DirectBuffer)
myBuffer).cleaner().clean();
● VisualVM plugin voor direct buffers

26. GPU vs CPU
● GPU's checken minder dan CPU's
– Div by zero
– Out of bounds checks
– Test eerst op CPU

27. Portabiliteit
● OpenCL is portable, de performance
niet
– Memory sizes verschillen
– Memory latencies verschillen
– Work group sizes verschillen
– Compute devices verschillen
– OpenCL implementatie verschillen
● Develop dus voor de productie
hardware

28. Ten slotte
● Float vs Double
– Dubbele precisie
– Halve performance
– Double support optioneel

29. Conclusie

30. Conclusie
● Wanneer te gebruiken?
– Als performance echt nodig is
– Als probleem hoge concurrency heeft
– Als probleem partitioneerbaar is

31. Vragen?
Setting up OpenCL test on Intel(R) Core(TM) i7-3720QM CPU @ 2.60GHz
Warming up OpenCL test
[thread 32003 also had an error][thread 33027 also had an error]
##
A fatal error has been detected by the Java Runtime Environment:
##
SIGSEGV[thread 32515 also had an error]
(0xb)[thread 32771 also had an error]
[thread 32259 also had an error]
at pc=0x00000001250ded70, pid=99851, tid=29475
##
JRE version: Java(TM) SE Runtime Environment (8.0_20-b26) (build 1.8.0_20-b26)
# Java VM: Java HotSpot(TM) 64-Bit Server VM (25.20-b23 mixed mode bsd-amd64 compressed oops)
# Problematic frame:
# [thread 17415 also had an error]
C [cl_kernels+0x1d70] sort_wrapper+0x1b0
##
Failed to write core dump. Core dumps have been disabled. To enable core dumping, try "ulimit -c unlimited" before starting Java again
##
An error report file with more information is saved as:
# /Users/arjanl/Documents/opencl/workspace/opencl-test/jogamp/hs_err_pid99851.log
[thread 31763 also had an error]
##
If you would like to submit a bug report, please visit:
# http://bugreport.sun.com/bugreport/crash.jsp
#