Do we have any owners of one of those Ryzen AI Max+ 395 128GB UMA boxes here that operate them on the daily for at least a few months as a claude LLM coding server and are capable of giving a comparative run down on their performance vs the OG claude and its collection of formal prose generators?

Also: Especially curious to hear any numbers that came out of a watt meter in daily consumption and base/peak numbers. Same with the used models, their size, their respective achieved tok/s and response times.

And should you have had the opportunity of comparing this against non-UMA beefy dGPUs on the above parameters that'd also be quote interesting.

#claude #aicoding #AIAsssisted #ollama #onprem #selfhosing #StrixPoint #ryzenaimaxplus395 #ryzenAiMax #powerconsumption #costefficiency #uma

404 Media: Pinterest Is Drowning in a Sea of AI Slop and Auto-Moderation. “Since 2009, the image sharing social media site has been a place for people to share their art, recipes, home renovation inspiration, corny motivational quotes, and more, but in the last year users, especially artists, say the site has gotten worse. AI-powered mods are pulling down posts and banning accounts, […]

The Conversation: Forensics tool ‘reanimates’ the ‘brains’ of AIs that fail in order to understand what went wrong. “We are computer scientists who study digital forensics. Our team at the Georgia Institute of Technology has built a system, AI Psychiatry, or AIP, that can recreate the scenario in which an AI failed in order to determine what went wrong. The system addresses the […]

https://rbfirehose.com/2025/05/02/the-conversation-forensics-tool-reanimates-the-brains-of-ais-that-fail-in-order-to-understand-what-went-wrong/