HPC Guru (Twitter)
Why SYCL: Elephants in the #SYCL Room
@JamesReinders and Michael Wong answer important questions asked by software developers looking to understand if SYCL matters to them
https://www.hpcwire.com/2022/02/03/why-sycl-elephants-in-the-sycl-room/
#HPC #AI #GPU #FPGA #CUDA #HIP #OpenCL via @HPCwire
Why SYCL: Elephants in the #SYCL Room
@JamesReinders and Michael Wong answer important questions asked by software developers looking to understand if SYCL matters to them
https://www.hpcwire.com/2022/02/03/why-sycl-elephants-in-the-sycl-room/
#HPC #AI #GPU #FPGA #CUDA #HIP #OpenCL via @HPCwire
HPC Guru (Twitter)
RT @IWOCL: IWOCL & SYCLcon 2022 marks the 10th anniversary of the annual gathering of the international community of #OpenCL and #SYCL developers (May 10-12, Online). The submission deadline for papers, talks and posters is Feb 25. https://www.iwocl.org/call-for-submissions/ @openclapi @SYCLstd
RT @IWOCL: IWOCL & SYCLcon 2022 marks the 10th anniversary of the annual gathering of the international community of #OpenCL and #SYCL developers (May 10-12, Online). The submission deadline for papers, talks and posters is Feb 25. https://www.iwocl.org/call-for-submissions/ @openclapi @SYCLstd
IWOCL
10th Int'l. Workshop on OpenCL and SYCL - May 10-12 | ONLINE
You are cordially invited to contribute and participate in the workshop through a paper, talk, special session, tutorial or poster. Submission deadline: FEB 25, 2021
HPC Guru (Twitter)
RT @ProjectPhysX: @geomark I got similar bandwidth values for the MI100 with a custom #OpenCL benchmark tool. Performs not quite to spec. Nvidia seems almost 2 generations ahead in terms of efficiency.
https://arxiv.org/abs/2112.08926 https://twitter.com/ProjectPhysX/status/1506271703084810240/photo/1
RT @ProjectPhysX: @geomark I got similar bandwidth values for the MI100 with a custom #OpenCL benchmark tool. Performs not quite to spec. Nvidia seems almost 2 generations ahead in terms of efficiency.
https://arxiv.org/abs/2112.08926 https://twitter.com/ProjectPhysX/status/1506271703084810240/photo/1
HPC Guru (Twitter)
RT @ProjectPhysX: @ProfMatsuoka At this point, arithmetic performance is so far ahead of memory bandwidth that it makes sense - even for non-AI workloads - to use "spare" arithmetic cycles to do data compression, such that the algorithm uses less memory and bandwidth.
https://t.co/3i9YdeXnOr
RT @ProjectPhysX: @ProfMatsuoka At this point, arithmetic performance is so far ahead of memory bandwidth that it makes sense - even for non-AI workloads - to use "spare" arithmetic cycles to do data compression, such that the algorithm uses less memory and bandwidth.
https://t.co/3i9YdeXnOr
Twitter
Moritz Lehmann
Make the lattice Boltzmann method (#LBM) use only ½ the memory and run 80% faster on #GPUs? #FP32/16-bit mixed precision makes it possible. Implementation takes 20 lines in existing code. Accuracy often is as good as #FP64.🖖🧐📃 Read my preprint: arxiv.or…
HPC Guru (Twitter)
RT @IWOCL: Delighted to confirm that Prof. @simonmcs from @BristolUni will present a keynote at this year's #opencl and #sycl conference titled "10 years of IWOCL and SYCLcon: A decade of breakthroughs, but is the best yet to come? " Full program and register at: https://www.iwocl.org/iwocl-2022/program/ https://twitter.com/IWOCL/status/1514561922091339781/photo/1
RT @IWOCL: Delighted to confirm that Prof. @simonmcs from @BristolUni will present a keynote at this year's #opencl and #sycl conference titled "10 years of IWOCL and SYCLcon: A decade of breakthroughs, but is the best yet to come? " Full program and register at: https://www.iwocl.org/iwocl-2022/program/ https://twitter.com/IWOCL/status/1514561922091339781/photo/1
HPC Guru (Twitter)
RT @ProjectPhysX: Big news everyone!! 🖖🥳💻🌊
I have published the @FluidX3D #CFD source code:
https://github.com/ProjectPhysX/FluidX3D
It's ~10k lines of very dense C++/#OpenCL code, optimized to the max and compatible with all datacenter/gaming/pro #GPUs. Looking forward to your feedback! 🖖😎 https://twitter.com/ProjectPhysX/status/1555172553689337863/photo/1
RT @ProjectPhysX: Big news everyone!! 🖖🥳💻🌊
I have published the @FluidX3D #CFD source code:
https://github.com/ProjectPhysX/FluidX3D
It's ~10k lines of very dense C++/#OpenCL code, optimized to the max and compatible with all datacenter/gaming/pro #GPUs. Looking forward to your feedback! 🖖😎 https://twitter.com/ProjectPhysX/status/1555172553689337863/photo/1
HPC Guru (Twitter)
👏👏 Great accomplishment, Dr. Moritz Lehmann (@ProjectPhysX), and an inspiration for others!
#HPC #CFD #GPU #OpenCL #FluidX3D
-----------
@ProjectPhysX:
5 years ago I had this wild idea to write my own #CFD software from scratch in #OpenCL. I wanted to know how fluid simulations work, and make them ridiculously fast on any #GPU. Today #FluidX3D has ⭐1.4k on #GitHub: https://t.co/c92kURQxiH
how it started how it's going https://t.co/aoJCKSR3sY
👏👏 Great accomplishment, Dr. Moritz Lehmann (@ProjectPhysX), and an inspiration for others!
#HPC #CFD #GPU #OpenCL #FluidX3D
-----------
@ProjectPhysX:
5 years ago I had this wild idea to write my own #CFD software from scratch in #OpenCL. I wanted to know how fluid simulations work, and make them ridiculously fast on any #GPU. Today #FluidX3D has ⭐1.4k on #GitHub: https://t.co/c92kURQxiH
how it started how it's going https://t.co/aoJCKSR3sY
HPC Guru (Twitter)
RT @ProjectPhysX: How fast are #HPC CPUs in #CFD? I benchmarked @FluidX3D on Sapphire Rapids and Ice/Cascade/Sky Lake on @intelhpc DevCloud. SPR-HBM (#IntelMaxSeries 9480) in cache mode are slower than non-HBM SPR (8480+) in #OpenCL. But both beat the EPYC 9654. GTX 1080 Ti for scale. 🖖😇🔥
🧵1/3 https://twitter.com/ProjectPhysX/status/1663942677283450880/photo/1
RT @ProjectPhysX: How fast are #HPC CPUs in #CFD? I benchmarked @FluidX3D on Sapphire Rapids and Ice/Cascade/Sky Lake on @intelhpc DevCloud. SPR-HBM (#IntelMaxSeries 9480) in cache mode are slower than non-HBM SPR (8480+) in #OpenCL. But both beat the EPYC 9654. GTX 1080 Ti for scale. 🖖😇🔥
🧵1/3 https://twitter.com/ProjectPhysX/status/1663942677283450880/photo/1
HPCwire (Twitter)
Primarily by using #CUDA (and #OpenCL, #OpenACC, etc.), #applications that are amenable to accelerated #processing have been optimized for #GPUs with great success. https://ow.ly/mjAk50PtGHR
Primarily by using #CUDA (and #OpenCL, #OpenACC, etc.), #applications that are amenable to accelerated #processing have been optimized for #GPUs with great success. https://ow.ly/mjAk50PtGHR
HPC Guru (Twitter)
Peachy Assignments serve as effective tools for #teaching specific programming models and provide insights into the portability of concepts and techniques across different models
https://gamuva.infor.uva.es/peachy-assignments/
#HPC #SC24 #MPI #CUDA #OpenCL #OpenMP
via @jgarciames.bsky.social
Peachy Assignments serve as effective tools for #teaching specific programming models and provide insights into the portability of concepts and techniques across different models
https://gamuva.infor.uva.es/peachy-assignments/
#HPC #SC24 #MPI #CUDA #OpenCL #OpenMP
via @jgarciames.bsky.social