You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
Dell and HPE have both captured strong growth in the AI Servers market, benefiting from increasing demand from Tier-2 CSP's, enterprises and sovereign AI's.
Wondering what your thoughts are on their differentiation--when would one pick Dell/HPE over the other when considering AI Servers?
Dell and HPE have both captured strong growth in the AI Servers market, benefiting from increasing demand from Tier-2 CSP's, enterprises and sovereign AI's.
Wondering what your thoughts are on their differentiation--when would one pick Dell/HPE over the other when considering AI Servers?
Maybe I’m missing something, but it sure looks like core AI inference is moving to server racks and systems built by AI chip providers - the systems showing the fastest and highest efficiency inference benchmarks are all from chip company built hardware. The last benchmark that popped out last week from Artifical Analysis shows Cerebras, NVIDIA, Sambanova and Groq built hardware dominating on the Llama Maverick LLM, though I think we’ll see continued back and forth between Cerebras and NVIDIA. Other players aren’t even really in the running.
Plus it looks like both AMD and Intel realize they have to jump fully into the AI rack-level system design and manufacturing game. To that end, AMD bought ZY systems, and Lip-bu at Intel has highlighted that “ there's no question we need to strengthen our position in the cloud-based Al data center market by developing competitive rack-scale system solutions, which will be a key priority for me and the team.”
So I would question your question if you are focused on systems that actually do the AI work.
Now if you are talking about HPC (high performant computing) leveraging GPUs, that’s a different story.
The NERSC-10 system at Berkeley Lab will accelerate DOE Office of Science research in an era of rapidly advancing simulation, data, and AI capabilities.
