Listening to Ancient Greek is good for the soul youtube.com/watch?v=WcF3iFBV…
Replying to @NVIDIAAIInfra
I made this image in march with @perplexity_ai hope you like it
I decided to write an article with my exact architecture I would use for what I am calling Oxide Computers: The Brains of an AI Fleet vibecodingagency.com/resourc… As an NVIDIA Certified AI Infrastructure Professional I am sure I could deploy this stack within one day with complete SBOM in the bill of materials. I would use chainguard.dev/ Docker containers, this produces a full verfiable non-root execution enviornment and pods will run with the highest security standards. If becoming "Oxide Certified" was a thing I would be happy to sign up.
Replying to @KVCacheStore @TechMDAI
Sometime soon someone is going to be able to fan out NVIDIA Workstations as L1-L3 cache clusters and it is going to be amazing historically unprecedented compute. For example a 10-node, 80-GPU Vera Rubin cluster with 23 TB HBM4 and 800G InfiniBand scale-out can handle up to 36T trillion parameters in NVFP4 models and train 30B–70B dense foundation models from scratch and it can only get optimized from there. I would never watch TV again if I had this.
@llm_wizard
Replying to @ratatui_rs
Fruit fly brains playing Doom in the terminal github.com/awdemos/replicant…
Replying to @bcantrill @ShaunRatcliff
"well actually, because 1% of factories do this, your point is totally invalid" I think it is fair to say Shaun has no idea what he is talking about. this response pattern has been elevated to a meme decades ago.
Replying to @llm_wizard @Mayhem4Markets
me looking at my mountain of Hail Caesar kits "but there's more room on the table for a historically accurate battle"













