On April 2, 2025, industry-standard MLPerf Inference v5.0 datacenter results were published by MLCommons. Red Hat and Supermicro submitted strong results for the popular llama2 70B model with Red Hat OpenShift running on their dual GPU GH200 Grace Hopper Superchip 144GB server. This was the first time anyone has submitted an MLPerf result with OpenShift on GH200. You can view these results at mlcommons.org. Llama2-70b Meta released the Llama2-70b model on July 18, 2023. This model is open source and part of the very popular Llama family of models that range from 7 billion to 70 billion paramet
via Red Hat Blog https://ift.tt/ah5FP0I
>>Click here to continue<<
