#FP64

@nextplatform.com.web.brid.gy

3 months ago

Driving HPC Performance Up Is Easier Than Keeping The Spending Constant We are still mulling over all of the new HPC-AI supercomputer systems that were announced in recent months before and during the SC25 supercomputing conference in St Louis, particularly how the slew of new machines announced by the HPC national labs will be advancing not just the state of the art, but also pushing down the cost of the FP64 floating point operations that still drives a lot of HPC simulation and modeling work. … Driving HPC Performance Up Is Easier Than Keeping The Spending Constant was written by Timothy Prickett Morgan at The Next Platform.

0 0 0 0

HGPU group

@hgpu.bsky.social

5 months ago

Performance and Numerical Aspects of Decompositional Factorizations with FP64 Floating-Point Emulation in INT8 Mixing precisions for performance has been an ongoing trend as the modern hardware accelerators started including new, and mostly lower-precision, data formats. The advantage of using them is a gre…

Performance and Numerical Aspects of Decompositional Factorizations with FP64 Floating-Point Emulation in INT8

#NVIDIA #Int8 #FP64 #Factorization

hgpu.org?p=30278

0 0 0 0

HGPU group

@hgpu.bsky.social

7 months ago

DGEMM without FP64 Arithmetic – using FP64 Emulation and FP8 Tensor Cores with Ozaki Scheme Since AI computations require low-precision matrix multiplications, processors with enhanced performance for these operations are increasing along with the growing demand for AI computations. Howev…

DGEMM without FP64 Arithmetic – using FP64 Emulation and FP8 Tensor Cores with Ozaki Scheme

#CUDA #Performance #DGEMM #FP64 #Package

hgpu.org?p=30081

0 0 0 0

Max Hawkins

@hawkinsmax.bsky.social

8 months ago

NVIDIA HGX Platform Enabling advanced AI and HPC in every data center.

Looks like #NVIDIA nerfed Blackwell Ultra #FP64 performance to make room for low-precision FP formats. Each B300 GPU now has ~1.25 TFLOPs of dense FP64 vector and tensor core performance (a ~90% decrease from the B200). Not exactly unexpected, but drastic.

Source: www.nvidia.com/en-us/data-c...

2 0 1 0

Andrew Jones (hpcnotes)

@hpcnotes.bsky.social

9 months ago

Provocative question to debate at #ISC25 - in talks, meetings, booths, over beers:

What is so special about #FP64 that science holds it as a gold standard?

If #HPC had never got as far as FP64, eg had got stuck at FP32, surely we would still have been able to do science?

8 1 2 1

Andrew Jones (hpcnotes)

@hpcnotes.bsky.social

1 year ago

#HPC and other scientific computing folks should like this image ... it's a floating point 😀

#FP64 #FLOPS

19 0 4 0

Andrew Jones (hpcnotes)

@hpcnotes.bsky.social

1 year ago

Experiment comparing reactions across social media platforms ...

One or more of #Fortran or #FP64 or #onprem #hpc will be obsolete by 2030.
Ignoring this message counts as agreeing :-)