Qwen 3.5: small models with impressive performancetwitter.com6 points·by moondistance·vor 4 Monaten·0 comments
Language Models Are Injective and Hence Invertiblearxiv.org1 points·by moondistance·vor 9 Monaten·2 comments
DeepSeek-R1 at 3,872 tokens / second on a single Nvidia HGX H200blogs.nvidia.com13 points·by moondistance·letztes Jahr·1 comments
ByteDance Doubao-1.5-pro matches GPT 4o benchmarks at 50x cheapertwitter.com4 points·by moondistance·letztes Jahr·0 comments
moondistance·vor 11 Monaten·discussHow does this compare with Exowatt? Assume fresnel lenses are more efficient.
moondistance·letztes Jahr·discussChris Crawford is also famous for the Dragon Speech :) https://youtu.be/CBrj4S24074?si=Ph12RpW8BKsh8-qS
moondistance·letztes Jahr·discussYes, big time, but there continues to be lots of progress.Most importantly, models are maturing, and this means less custom optimization is required.
moondistance·letztes Jahr·discuss(this is also why AMD popped 10% at open yesterday - this is a new development and talks from their 2025 "Advancing AI" event were published late last week + over the weekend)
moondistance·letztes Jahr·discussYes, for many applications.Meta, OpenAI, Crusoe, and xAI recently announced large purchases of MI300 chips for inference.MI400, which will be available next year, also looks to be at least on par with Nvidia's roadmap.
moondistance·letztes Jahr·discussWhere do you find/buy land? How do you vet purchases? Can you point to a few websites, etc.? Thanks!
moondistance·vor 2 Jahren·discussHappy New Year!I know so few of you but feel like you’re my best friends. Love to all.