fix bablestream benchmark #2420

mehmetyusufoglu · 2024-11-07T13:28:29Z

Some of the 5 kernels of Babelstream-benchmark were not connected to each other, with this change if one of them changed somehow and fails; error is cached in the last result. ( Since we don't check after each kernel run, this is needed to make sure all kernels are connected.)
Using arrays in a different order in calling different kernels might affect the performance (although not observed) due to caching, therefore using same arrays for the same kernels (as in the original babelstream of UoB) in the kernel call sequence is also done by above change.
An optional kernel is added, NStream. This can be run separately alone.
One of the 5 kernels of babelstream, the triad kernel, was optionally being run alone in the original code by UoB. This option is also added.

This PR is an extension of previous PR: #2299

New parameters and kernel calls with specific arrays in the kernel call sequence to avoid cache usage differences:

A = 0.1 B= 0.2 C= 0.0 scalar = 0.4
C = A // copy
B = scalar * C // mult
C = A + B // add
A = B + scalar * C // triad
Missing optional kernel NStream is added
Dot kernel is only run by multiple-threaded accs. Since original babelstream uses fixed 1024 blocksize which is the shared memory size per block as well. (search "#define TBSIZE 1024" in This Cuda code)

RESULTS

TEST_RUN : ./babelstream --array-size=33554432 --number-runs=100

./babelstream --array-size=33554432 --number-runs=100
Array size set to: 33554432
Number of runs provided: 100
Randomness seeded to: 3184604301
Kernels: Init, Copy, Mul, Add, Triad Kernels (and Dot kernel, if acc is multi-thread per block.)


AcceleratorType:AccGpuCudaRt<1,unsigned int>
NumberOfRuns:100
Precision:single
DataSize(items):33554432
DeviceName:NVIDIA RTX A500 Laptop GPU
WorkDivInit :{gridBlockExtent: (32768), blockThreadExtent: (1024), threadElemExtent: (1)}
WorkDivCopy :{gridBlockExtent: (32768), blockThreadExtent: (1024), threadElemExtent: (1)}
WorkDivMult :{gridBlockExtent: (32768), blockThreadExtent: (1024), threadElemExtent: (1)}
WorkDivAdd  :{gridBlockExtent: (32768), blockThreadExtent: (1024), threadElemExtent: (1)}
WorkDivTriad:{gridBlockExtent: (32768), blockThreadExtent: (1024), threadElemExtent: (1)}
WorkDivDot  :{gridBlockExtent: (256), blockThreadExtent: (1024), threadElemExtent: (1)}
AccToHost Memcpy Time(sec):0.223933
Kernels         Bandwidths(GB/s) MinTime(s) MaxTime(s) AvgTime(s) DataUsage(MB) 
 AddKernel       91.371          0.0044068 0.0044819 0.0044661 402.65 
 CopyKernel      90.075          0.0029801 0.0030822 0.0030193 268.44 
 DotKernel       92.759          0.0028939 0.0029579 0.0029319 268.44 
 InitKernel      92.418          0.0043569 0.0043569 0.0043569 402.65 
 MultKernel      90.276          0.0029735 0.0030676 0.003011 268.44 
 TriadKernel     90.763          0.0044363 0.0044944 0.0044705 402.65 

Kernels: Init, Copy, Mul, Add, Triad Kernels (and Dot kernel, if acc is multi-thread per block.)


AcceleratorType:AccGpuCudaRt<1,unsigned int>
NumberOfRuns:100
Precision:double
DataSize(items):33554432
DeviceName:NVIDIA RTX A500 Laptop GPU
WorkDivInit :{gridBlockExtent: (32768), blockThreadExtent: (1024), threadElemExtent: (1)}
WorkDivCopy :{gridBlockExtent: (32768), blockThreadExtent: (1024), threadElemExtent: (1)}
WorkDivMult :{gridBlockExtent: (32768), blockThreadExtent: (1024), threadElemExtent: (1)}
WorkDivAdd  :{gridBlockExtent: (32768), blockThreadExtent: (1024), threadElemExtent: (1)}
WorkDivTriad:{gridBlockExtent: (32768), blockThreadExtent: (1024), threadElemExtent: (1)}
WorkDivDot  :{gridBlockExtent: (256), blockThreadExtent: (1024), threadElemExtent: (1)}
AccToHost Memcpy Time(sec):0.570856
Kernels         Bandwidths(GB/s) MinTime(s) MaxTime(s) AvgTime(s) DataUsage(MB) 
 AddKernel       90.665          0.0088822 0.0089985 0.0089376 805.31 
 CopyKernel      89.087          0.0060264 0.0061119 0.0060773 536.87 
 DotKernel       93.055          0.0057694 0.0058486 0.0058113 536.87 
 InitKernel      84.437          0.0095374 0.0095374 0.0095374 805.31 
 MultKernel      89.35           0.0060086 0.0060852 0.0060568 536.87 
 TriadKernel     90.222          0.0089258 0.0090338 0.0089565 805.31 

===============================================================================
All tests passed (8 assertions in 2 test cases)

psychocoderHPC · 2024-11-12T09:16:24Z

@mehmetyusufoglu Can you please check if CPU is working too.

benchmarks/babelstream/src/babelStreamMainTest.cpp

psychocoderHPC · 2024-11-13T08:51:33Z

benchmarks/babelstream/src/babelStreamMainTest.cpp


+            DataType const* sumPtr = std::data(bufHostSumPerBlock);
+            float const result = std::reduce(sumPtr, sumPtr + gridBlockExtent, 0.0f);


This and the memcpy has be be part of the measurement, because that's our fault that we not execute the full reduction on device.
To have a fair comparison the allocation of the result must be part of measureKernelExec too but the cuda upstream implementation is cheating here too so I assume this is fine to be allocated outside of measureKernelExec

ok, I saw these are also part of measurement. Thanks.

done, thanks. [reduce is taken into measure]

psychocoderHPC · 2024-11-13T09:17:51Z

benchmarks/babelstream/src/babelStreamCommon.hpp


    // Block thread extent for DotKernel test work division parameters.
    [[maybe_unused]] constexpr auto blockThreadExtentMain = 1024;
+    [[maybe_unused]] constexpr auto dotGridBlockExtent = 256;


feel free to add a extent for CPUs too to support CPU dot execution required for the verification.

additional test case is added for blocksize = 1 for Dot Kernel. Hence Dot kernel is tested for CPU backend and for blocksize =1.

chillenzer

Hi, thanks for your update. I've mostly checked for compliance with upstream babelstream as that was apparently a major point of discussion. But I want to first applaud you for that nice auto [i] = getIdx...; trick. Very nice indeed!

It's mostly small things I've found some of which we might actively decide to do differently, e.g., not measuring some timings of the infrastructure-ish calls. There was one section that GH didn't allow me to comment on, so I'll put that in here:
Technically speaking, the dot kernel does a slightly different thing by accumulating the threadSums in registers and only storing them once into shared memory. Our version vs. upstream version. Likely to be optimised away by the compiler because both versions leave full flexibility concerning memory ordering here but we can't be sure I believe.

Also, just for my information: Why is tbSum a reference in that same kernel? It very much looks like it must be dangling but if this compiles and runs correctly it apparently isn't?

chillenzer · 2024-11-13T08:56:59Z

benchmarks/babelstream/src/babelStreamMainTest.cpp

-        [&]()
-        { alpaka::exec<Acc>(queue, workDivTriad, TriadKernel(), bufAccInputAPtr, bufAccInputBPtr, bufAccOutputCPtr); },
-        "TriadKernel");
+    if(kernelsToBeExecuted == KernelsToRun::All || kernelsToBeExecuted == KernelsToRun::Triad)


Suggested change

if(kernelsToBeExecuted == KernelsToRun::All || kernelsToBeExecuted == KernelsToRun::Triad)

else if(kernelsToBeExecuted == KernelsToRun::Triad)

following https://github.com/UoB-HPC/BabelStream/blob/2f00dfb7f8b7cfe8c53d20d5c770bccbf8673440/src/main.cpp#L532

There is a code repetition in the original code, both run_triad and run_all calls the triad kernel. Here, i call the same code piece for both cases.

alpaka/benchmarks/babelstream/src/main.cpp

Line 108 in 7a8b205

std::vector<std::vector<double>> run_all(Stream<T>* stream, T& sum)

benchmarks/babelstream/src/babelStreamMainTest.cpp

chillenzer · 2024-11-18T12:35:07Z

benchmarks/babelstream/src/babelStreamMainTest.cpp

+    alpaka::exec<Acc>(
+        queue,
+        workDivInit,
+        InitKernel(),
+        bufAccInputAPtr,
+        bufAccInputBPtr,
+        bufAccOutputCPtr,
+        static_cast<DataType>(initA),
+        static_cast<DataType>(initB));


In the original, this call is already timed.

It is not one of babelstream kernels but ok i am implementing.

Yes, I agree that it makes moderate sense at best but it might be interesting information and it brings us closer to upstream. Your announced change is not yet found in the PR.

ok, implemented. Thanks.

AcceleratorType:AccGpuCudaRt<1,unsigned int> NumberOfRuns:2 Precision:double DataSize(items):1048576 DeviceName:NVIDIA RTX A500 Laptop GPU WorkDivInit :{gridBlockExtent: (1024), blockThreadExtent: (1024), threadElemExtent: (1)} WorkDivCopy :{gridBlockExtent: (1024), blockThreadExtent: (1024), threadElemExtent: (1)} WorkDivMult :{gridBlockExtent: (1024), blockThreadExtent: (1024), threadElemExtent: (1)} WorkDivAdd :{gridBlockExtent: (1024), blockThreadExtent: (1024), threadElemExtent: (1)} WorkDivTriad:{gridBlockExtent: (1024), blockThreadExtent: (1024), threadElemExtent: (1)} WorkDivDot :{gridBlockExtent: (256), blockThreadExtent: (1024), threadElemExtent: (1)} Kernels Bandwidths(GB/s) MinTime(s) MaxTime(s) AvgTime(s) DataUsage(MB) AddKernel 87.201 0.0002886 0.0002886 0.0002886 25.166 CopyKernel 79.096 0.00021211 0.00021211 0.00021211 16.777 DotKernel 74.74 0.00022447 0.00022447 0.00022447 16.777 InitKernel 87.865 0.00028641 0.00028641 0.00028641 25.166 MultKernel 85.107 0.00019713 0.00019713 0.00019713 16.777 TriadKernel 87.046 0.00028911 0.00028911 0.00028911 25.166

benchmarks/babelstream/src/babelStreamMainTest.cpp

chillenzer · 2024-11-18T13:05:19Z

benchmarks/babelstream/src/babelStreamMainTest.cpp

    alpaka::memcpy(queue, bufHostOutputC, bufAccOutputC, arraySize);
    alpaka::memcpy(queue, bufHostOutputB, bufAccInputB, arraySize);
    alpaka::memcpy(queue, bufHostOutputA, bufAccInputA, arraySize);


These get timed int the original version.

Read time (actually copy time from Acc to Host for 3 arrays ) has been added to the output display as AccToHost Memcpy Time(sec). Thanks.

AcceleratorType:AccGpuCudaRt<1,unsigned int> NumberOfRuns:100 Precision:double DataSize(items):33554432 DeviceName:NVIDIA RTX A500 Laptop GPU WorkDivInit :{gridBlockExtent: (32768), blockThreadExtent: (1024), threadElemExtent: (1)} WorkDivCopy :{gridBlockExtent: (32768), blockThreadExtent: (1024), threadElemExtent: (1)} WorkDivMult :{gridBlockExtent: (32768), blockThreadExtent: (1024), threadElemExtent: (1)} WorkDivAdd :{gridBlockExtent: (32768), blockThreadExtent: (1024), threadElemExtent: (1)} WorkDivTriad:{gridBlockExtent: (32768), blockThreadExtent: (1024), threadElemExtent: (1)} WorkDivDot :{gridBlockExtent: (256), blockThreadExtent: (1024), threadElemExtent: (1)} AccToHost Memcpy Time(sec):0.570856 Kernels Bandwidths(GB/s) MinTime(s) MaxTime(s) AvgTime(s) DataUsage(MB) AddKernel 90.665 0.0088822 0.0089985 0.0089376 805.31 CopyKernel 89.087 0.0060264 0.0061119 0.0060773 536.87 DotKernel 93.055 0.0057694 0.0058486 0.0058113 536.87 InitKernel 84.437 0.0095374 0.0095374 0.0095374 805.31 MultKernel 89.35 0.0060086 0.0060852 0.0060568 536.87 TriadKernel 90.222 0.0089258 0.0090338 0.0089565 805.31

benchmarks/babelstream/src/babelStreamMainTest.cpp

mehmetyusufoglu · 2024-11-18T16:56:08Z

Hi, thanks for your update. I've mostly checked for compliance with upstream babelstream as that was apparently a major point of discussion. But I want to first applaud you for that nice auto [i] = getIdx...; trick. Very nice indeed!

It's mostly small things I've found some of which we might actively decide to do differently, e.g., not measuring some timings of the infrastructure-ish calls. There was one section that GH didn't allow me to comment on, so I'll put that in here: Technically speaking, the dot kernel does a slightly different thing by accumulating the threadSums in registers and only storing them once into shared memory. Our version vs. upstream version. Likely to be optimised away by the compiler because both versions leave full flexibility concerning memory ordering here but we can't be sure I believe.

Also, just for my information: Why is tbSum a reference in that same kernel? It very much looks like it must be dangling but if this compiles and runs correctly it apparently isn't?

tbSum is reference because the function return type is -> T& and returns a dereferenced value return *data;

mehmetyusufoglu · 2024-11-18T18:59:16Z

Hi, thanks for your update. I've mostly checked for compliance with upstream babelstream as that was apparently a major point of discussion. But I want to first applaud you for that nice auto [i] = getIdx...; trick. Very nice indeed!

It's mostly small things I've found some of which we might actively decide to do differently, e.g., not measuring some timings of the infrastructure-ish calls. There was one section that GH didn't allow me to comment on, so I'll put that in here: Technically speaking, the dot kernel does a slightly different thing by accumulating the threadSums in registers and only storing them once into shared memory. Our version vs. upstream version. Likely to be optimised away by the compiler because both versions leave full flexibility concerning memory ordering here but we can't be sure I believe.

Yes, this was the choice at the first implementation at our repo, i used directly like in the cuda implementation now. Checking the performance.

chillenzer · 2024-11-19T09:28:37Z

tbSum is reference because the function return type is -> T& and returns a dereferenced value return *data;

Thanks for the explanation! That makes sense.

Concerning the reduce implementation, I had an offline discussion with @psychocoderHPC: The concept to benchmark here is any implementation of a reduction based on alpaka. In that sense, we are not required to follow the reference implementation precisely. Not hammering on shared memory with every thread is probably a worthwhile change.

mehmetyusufoglu · 2024-11-19T18:04:25Z

@mehmetyusufoglu Can you please check if CPU is working too.

Randomness seeded to: 2905169299
Kernels: Init, Copy, Mul, Add, Triad Kernels (and Dot kernel, if acc is multi-thread per block.)

AcceleratorType:AccCpuSerial<1,unsigned int>
NumberOfRuns:2
Precision:single
DataSize(items):1048576
DeviceName:13th Gen Intel(R) Core(TM) i7-1360P
WorkDivInit :{gridBlockExtent: (1048576), blockThreadExtent: (1), threadElemExtent: (1)}
WorkDivCopy :{gridBlockExtent: (1048576), blockThreadExtent: (1), threadElemExtent: (1)}
WorkDivMult :{gridBlockExtent: (1048576), blockThreadExtent: (1), threadElemExtent: (1)}
WorkDivAdd :{gridBlockExtent: (1048576), blockThreadExtent: (1), threadElemExtent: (1)}
WorkDivTriad:{gridBlockExtent: (1048576), blockThreadExtent: (1), threadElemExtent: (1)}
AccToHost Memcpy Time(sec):0.0107734
Kernels Bandwidths(GB/s) MinTime(s) MaxTime(s) AvgTime(s) DataUsage(MB)
AddKernel 0.026882 0.46807 0.46807 0.46807 12.583
CopyKernel 0.019559 0.42889 0.42889 0.42889 8.3886
InitKernel 0.029203 0.43088 0.43088 0.43088 12.583
MultKernel 0.019739 0.42498 0.42498 0.42498 8.3886
TriadKernel 0.025445 0.49452 0.49452 0.49452 12.583

Kernels: Init, Copy, Mul, Add, Triad Kernels (and Dot kernel, if acc is multi-thread per block.)

AcceleratorType:AccGpuCudaRt<1,unsigned int>
NumberOfRuns:2
Precision:single
DataSize(items):1048576
DeviceName:NVIDIA RTX A500 Laptop GPU
WorkDivInit :{gridBlockExtent: (1024), blockThreadExtent: (1024), threadElemExtent: (1)}
WorkDivCopy :{gridBlockExtent: (1024), blockThreadExtent: (1024), threadElemExtent: (1)}
WorkDivMult :{gridBlockExtent: (1024), blockThreadExtent: (1024), threadElemExtent: (1)}
WorkDivAdd :{gridBlockExtent: (1024), blockThreadExtent: (1024), threadElemExtent: (1)}
WorkDivTriad:{gridBlockExtent: (1024), blockThreadExtent: (1024), threadElemExtent: (1)}
WorkDivDot :{gridBlockExtent: (256), blockThreadExtent: (1024), threadElemExtent: (1)}
AccToHost Memcpy Time(sec):0.0135214
Kernels Bandwidths(GB/s) MinTime(s) MaxTime(s) AvgTime(s) DataUsage(MB)
AddKernel 80.998 0.00015535 0.00015535 0.00015535 12.583
CopyKernel 69.936 0.00011995 0.00011995 0.00011995 8.3886
DotKernel 48.621 0.00017253 0.00017253 0.00017253 8.3886
InitKernel 51.814 0.00024285 0.00024285 0.00024285 12.583
MultKernel 76.158 0.00011015 0.00011015 0.00011015 8.3886
TriadKernel 81.478 0.00015443 0.00015443 0.00015443 12.583

Kernels: Init, Copy, Mul, Add, Triad Kernels (and Dot kernel, if acc is multi-thread per block.)

AcceleratorType:AccCpuSerial<1,unsigned int>
NumberOfRuns:2
Precision:double
DataSize(items):1048576
DeviceName:13th Gen Intel(R) Core(TM) i7-1360P
WorkDivInit :{gridBlockExtent: (1048576), blockThreadExtent: (1), threadElemExtent: (1)}
WorkDivCopy :{gridBlockExtent: (1048576), blockThreadExtent: (1), threadElemExtent: (1)}
WorkDivMult :{gridBlockExtent: (1048576), blockThreadExtent: (1), threadElemExtent: (1)}
WorkDivAdd :{gridBlockExtent: (1048576), blockThreadExtent: (1), threadElemExtent: (1)}
WorkDivTriad:{gridBlockExtent: (1048576), blockThreadExtent: (1), threadElemExtent: (1)}
AccToHost Memcpy Time(sec):0.0151765
Kernels Bandwidths(GB/s) MinTime(s) MaxTime(s) AvgTime(s) DataUsage(MB)
AddKernel 0.059712 0.42146 0.42146 0.42146 25.166
CopyKernel 0.042238 0.39721 0.39721 0.39721 16.777
InitKernel 0.03913 0.64314 0.64314 0.64314 25.166
MultKernel 0.04646 0.36111 0.36111 0.36111 16.777
TriadKernel 0.062699 0.40138 0.40138 0.40138 25.166

Kernels: Init, Copy, Mul, Add, Triad Kernels (and Dot kernel, if acc is multi-thread per block.)

AcceleratorType:AccGpuCudaRt<1,unsigned int>
NumberOfRuns:2
Precision:double
DataSize(items):1048576
DeviceName:NVIDIA RTX A500 Laptop GPU
WorkDivInit :{gridBlockExtent: (1024), blockThreadExtent: (1024), threadElemExtent: (1)}
WorkDivCopy :{gridBlockExtent: (1024), blockThreadExtent: (1024), threadElemExtent: (1)}
WorkDivMult :{gridBlockExtent: (1024), blockThreadExtent: (1024), threadElemExtent: (1)}
WorkDivAdd :{gridBlockExtent: (1024), blockThreadExtent: (1024), threadElemExtent: (1)}
WorkDivTriad:{gridBlockExtent: (1024), blockThreadExtent: (1024), threadElemExtent: (1)}
WorkDivDot :{gridBlockExtent: (256), blockThreadExtent: (1024), threadElemExtent: (1)}
AccToHost Memcpy Time(sec):0.0173797
Kernels Bandwidths(GB/s) MinTime(s) MaxTime(s) AvgTime(s) DataUsage(MB)
AddKernel 87.21 0.00028857 0.00028857 0.00028857 25.166
CopyKernel 80.442 0.00020856 0.00020856 0.00020856 16.777
DotKernel 73.262 0.000229 0.000229 0.000229 16.777
InitKernel 85.267 0.00029514 0.00029514 0.00029514 25.166
MultKernel 85.196 0.00019693 0.00019693 0.00019693 16.777
TriadKernel 87.512 0.00028757 0.00028757 0.00028757 25.166

===============================================================================
All tests passed (18 assertions in 4 test cases)

mehmetyusufoglu · 2024-11-19T18:05:31Z

tbSum is reference because the function return type is -> T& and returns a dereferenced value return *data;

Thanks for the explanation! That makes sense.

Concerning the reduce implementation, I had an offline discussion with @psychocoderHPC: The concept to benchmark here is any implementation of a reduction based on alpaka. In that sense, we are not required to follow the reference implementation precisely. Not hammering on shared memory with every thread is probably a worthwhile change.

Ok I reverted it back. (Yes accessing shared memory at each thread many times is not needed at such case)

mehmetyusufoglu force-pushed the updateBabelStr branch from 3777a36 to 98aac95 Compare November 9, 2024 15:03

mehmetyusufoglu marked this pull request as draft November 10, 2024 20:03

mehmetyusufoglu force-pushed the updateBabelStr branch 8 times, most recently from 07496fa to d7982fb Compare November 11, 2024 14:08

mehmetyusufoglu force-pushed the updateBabelStr branch 2 times, most recently from 2bfcf11 to 3aa4303 Compare November 12, 2024 13:40

mehmetyusufoglu marked this pull request as ready for review November 12, 2024 13:42

mehmetyusufoglu force-pushed the updateBabelStr branch from 3aa4303 to 862a390 Compare November 12, 2024 13:53

psychocoderHPC requested changes Nov 13, 2024

View reviewed changes

psychocoderHPC reviewed Nov 13, 2024

View reviewed changes

mehmetyusufoglu marked this pull request as draft November 13, 2024 15:35

mehmetyusufoglu force-pushed the updateBabelStr branch 6 times, most recently from cf19b90 to e2b8eae Compare November 18, 2024 13:28

psychocoderHPC changed the title ~~Make kernel results depend each other directly~~ fix bablestream benchmark Nov 18, 2024

psychocoderHPC added the Type:Bug label Nov 18, 2024

psychocoderHPC added this to the 2.0.0 milestone Nov 18, 2024

chillenzer suggested changes Nov 18, 2024

View reviewed changes

mehmetyusufoglu force-pushed the updateBabelStr branch from e2b8eae to 254dc3d Compare November 18, 2024 15:46

mehmetyusufoglu force-pushed the updateBabelStr branch 2 times, most recently from e014dca to efbd007 Compare November 18, 2024 18:56

mehmetyusufoglu force-pushed the updateBabelStr branch 4 times, most recently from 7ec9872 to dd2b920 Compare November 19, 2024 18:02

mehmetyusufoglu force-pushed the updateBabelStr branch 2 times, most recently from 5d704c4 to 1f4aeeb Compare November 21, 2024 13:54

make kernels depend each other, use original variables and access order

34fb823

mehmetyusufoglu force-pushed the updateBabelStr branch from 1f4aeeb to 34fb823 Compare November 22, 2024 18:46

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

fix bablestream benchmark #2420

fix bablestream benchmark #2420

mehmetyusufoglu commented Nov 7, 2024 •

edited

Loading

psychocoderHPC commented Nov 12, 2024

psychocoderHPC Nov 13, 2024

mehmetyusufoglu Nov 13, 2024

mehmetyusufoglu Nov 13, 2024 •

edited

Loading

psychocoderHPC Nov 13, 2024

mehmetyusufoglu Nov 19, 2024

chillenzer left a comment

chillenzer Nov 13, 2024

mehmetyusufoglu Nov 18, 2024

chillenzer Nov 18, 2024

mehmetyusufoglu Nov 18, 2024

chillenzer Nov 19, 2024

mehmetyusufoglu Nov 19, 2024

chillenzer Nov 18, 2024

mehmetyusufoglu Nov 19, 2024

mehmetyusufoglu commented Nov 18, 2024

mehmetyusufoglu commented Nov 18, 2024

chillenzer commented Nov 19, 2024

mehmetyusufoglu commented Nov 19, 2024

mehmetyusufoglu commented Nov 19, 2024


		DataType const* sumPtr = std::data(bufHostSumPerBlock);
		float const result = std::reduce(sumPtr, sumPtr + gridBlockExtent, 0.0f);

	if(kernelsToBeExecuted == KernelsToRun::All \|\| kernelsToBeExecuted == KernelsToRun::Triad)
	else if(kernelsToBeExecuted == KernelsToRun::Triad)

fix bablestream benchmark #2420

Are you sure you want to change the base?

fix bablestream benchmark #2420

Conversation

mehmetyusufoglu commented Nov 7, 2024 • edited Loading

psychocoderHPC commented Nov 12, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

mehmetyusufoglu Nov 13, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

chillenzer left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

mehmetyusufoglu commented Nov 18, 2024

mehmetyusufoglu commented Nov 18, 2024

chillenzer commented Nov 19, 2024

mehmetyusufoglu commented Nov 19, 2024

mehmetyusufoglu commented Nov 19, 2024

mehmetyusufoglu commented Nov 7, 2024 •

edited

Loading

mehmetyusufoglu Nov 13, 2024 •

edited

Loading