AMD MI300 Performance – Faster Than H100, but How Much?(semianalysis.com)

63 points by klelatti 2 years ago | 47 comments

AMD AI Software Solved – MI300X Pricing, Perf, PyTorch, FlashAttention, Triton(semianalysis.com)

32 points by another 3 years ago | 11 comments

OpenAI Jalapeño: Better than Nvidia Blackwell(newsletter.semianalysis.com)

555 points by bmulholland 2 months ago | 351 comments

SemiAnalysis article analyzing OpenAI's self-designed Jalapeño inference ASIC, comparing its TCO and throughput per MW against Nvidia Blackwell and Rubin.

AMD MI300x GPUs with GEMM tuning improves throughput and latency by up to 7.2x(nscale.com)

60 points by latchkey 2 years ago | 43 comments

AWS Trainium3 Deep Dive – A Potential Challenger Approaching(newsletter.semianalysis.com)

58 points by Symmetry 10 months ago | 20 comments

CXL Is Dead in the AI Era(semianalysis.com)

12 points by wmf 2 years ago | 5 comments