HPCwire (Twitter)
After a brief rollback due to #AMD's request, #ZLUDA is now moving forward with a clean start, running CUDA apps on AMD #GPUs with near-native performance. A new sponsor is backing Andrzej Janik's vision for this game-changing project.
ow.ly/WrKi50TGhO4
After a brief rollback due to #AMD's request, #ZLUDA is now moving forward with a clean start, running CUDA apps on AMD #GPUs with near-native performance. A new sponsor is backing Andrzej Janik's vision for this game-changing project.
ow.ly/WrKi50TGhO4
HPCwire (Twitter)
Foxconn marks its 50th anniversary by announcing plans to build Taiwan’s most powerful #supercomputer with 4,608 Nvidia Blackwell #GPUs, boasting 90 AI exaflops. Set for completion in 2026, learn more: ow.ly/hzSg50TKmBE.
Foxconn marks its 50th anniversary by announcing plans to build Taiwan’s most powerful #supercomputer with 4,608 Nvidia Blackwell #GPUs, boasting 90 AI exaflops. Set for completion in 2026, learn more: ow.ly/hzSg50TKmBE.
HPC Guru (Twitter)
TSMC is expected to stop supplying chips made with 7nm or smaller processes to customers in China that are developing #AI processors or #GPUs
The move is reportedly to ensure it remains compliant with US export restrictions
https://www.theregister.com/2024/11/08/tsmc_chinese_ai_shipments/
#HPC via @TheRegister
TSMC is expected to stop supplying chips made with 7nm or smaller processes to customers in China that are developing #AI processors or #GPUs
The move is reportedly to ensure it remains compliant with US export restrictions
https://www.theregister.com/2024/11/08/tsmc_chinese_ai_shipments/
#HPC via @TheRegister
X (formerly Twitter)
#AI - Search / X
See posts about #AI on X. See what people are saying and join the conversation.
HPC Guru (Twitter)
Nvidia's #Blackwell #GPUs overheat in #NVL72 server racks
Design changes and delays raise concerns among customers
https://www.tomshardware.com/pc-components/gpus/nvidias-data-center-blackwell-gpus-reportedly-overheat-require-rack-redesigns-and-cause-delays-for-customers
#SC24 #HPC #AI via @reaktor_field
Nvidia's #Blackwell #GPUs overheat in #NVL72 server racks
Design changes and delays raise concerns among customers
https://www.tomshardware.com/pc-components/gpus/nvidias-data-center-blackwell-gpus-reportedly-overheat-require-rack-redesigns-and-cause-delays-for-customers
#SC24 #HPC #AI via @reaktor_field
HPC Guru (Twitter)
RT @EvanKirstel: xAI plans to expand its Colossus supercomputer tenfold to incorporate more than 1M #GPUs; work has already begun to increase the size of its facility in Memphis (Financial Times https://www.ft.com/content/9c0516cf-dd12-4665-aa22-712de854fe2f via @FT @HPC_Guru
RT @EvanKirstel: xAI plans to expand its Colossus supercomputer tenfold to incorporate more than 1M #GPUs; work has already begun to increase the size of its facility in Memphis (Financial Times https://www.ft.com/content/9c0516cf-dd12-4665-aa22-712de854fe2f via @FT @HPC_Guru
HPC Guru (Twitter)
Elon Musk’s artificial intelligence start-up @xai has pledged to expand its #Colossus supercomputer tenfold to incorporate more than 1 million #GPUs, in an effort to leap ahead of rivals such as Google, OpenAI and Anthropic
https://www.ft.com/content/9c0516cf-dd12-4665-aa22-712de854fe2f
#AI #HPC via @EvanKirstel
Elon Musk’s artificial intelligence start-up @xai has pledged to expand its #Colossus supercomputer tenfold to incorporate more than 1 million #GPUs, in an effort to leap ahead of rivals such as Google, OpenAI and Anthropic
https://www.ft.com/content/9c0516cf-dd12-4665-aa22-712de854fe2f
#AI #HPC via @EvanKirstel
X (formerly Twitter)
#Colossus - Search / X
See posts about #Colossus on X. See what people are saying and join the conversation.
HPC Guru (Twitter)
Look ma, no #GPUs
AWS is building a machine with hundreds of thousands of #Tranium2 processors to achieve performance of ~65 ExaFLOPS for #AI
AWS also unveiled its Trainium3 processor which will quadruple performance compared to the Trainium2
https://www.tomshardware.com/tech-industry/artificial-intelligence/aws-building-exaflops-class-supercomputer-for-ai-with-hundreds-of-thousands-trainium2-processors
#HPC
Look ma, no #GPUs
AWS is building a machine with hundreds of thousands of #Tranium2 processors to achieve performance of ~65 ExaFLOPS for #AI
AWS also unveiled its Trainium3 processor which will quadruple performance compared to the Trainium2
https://www.tomshardware.com/tech-industry/artificial-intelligence/aws-building-exaflops-class-supercomputer-for-ai-with-hundreds-of-thousands-trainium2-processors
#HPC
HPC Guru (Twitter)
RT @HatemLtaief: @HPC_Guru @ernstdj @GeorgHager: 2!!!
Even worse with a wider gap for #GPUs.
Credits: graph from Thomas Gruber (#likwid)
RT @HatemLtaief: @HPC_Guru @ernstdj @GeorgHager: 2!!!
Even worse with a wider gap for #GPUs.
Credits: graph from Thomas Gruber (#likwid)
HPC Guru (Twitter)
Nvidia is constructing a 30MW R&D #supercomputer stuffed with its latest-generation #Blackwell #GPUs in northern Israel at an estimated cost of half a billion dollars
https://www.theregister.com/2025/01/16/nvidia_israel_blackwell/
#AI #HPC via @TheRegister
Nvidia is constructing a 30MW R&D #supercomputer stuffed with its latest-generation #Blackwell #GPUs in northern Israel at an estimated cost of half a billion dollars
https://www.theregister.com/2025/01/16/nvidia_israel_blackwell/
#AI #HPC via @TheRegister
X (formerly Twitter)
#supercomputer - Search / X
See posts about #supercomputer on X. See what people are saying and join the conversation.