Nvidia reportedly prioritizes model distillation over heavy engineering.
"Nvidia could release a GLM 5.2 or better equivalent model any time it wants" I love the Nemotron guys, but this isn't true
Nvidia could release a GLM 5.2 or better equivalent model any time it wants, and nuke OpenAI's margins 🤯
Many users challenged the claim that Nvidia cannot release a competitive LLM to undercut OpenAI, calling the skepticism biased or unrealistic and pointing to Chinese model progress, while others agreed it was wishful thinking.
No Digg Deeper questions have been answered for this story yet.
yeah, Jensen would sure like to do just that The reason Baker believes Jensen is choosing to not do that is that he has absolute faith in the powers of distillation. The Chinese are not even doing any real engineering, so it stands to reason Nemotron team also doesn't have to
Nvidia could release a GLM 5.2 or better equivalent model any time it wants, and nuke OpenAI's margins 🤯
@zephyr_z9 You are wrong , this is completely true
@zephyr_z9 Totally agree!
No one can even try reaching the capability levels of the Chinese OS models. This is just wishful day dreaming.
@riskdeity @zephyr_z9 Nemotron Ultra dropped few weeks ago and no one really swears by it
Nvidia drops NVFP4 quants of GLM 5.2 and it’s not like it took the timeline by storm these past few days
What are you seeing that we aren’t ?
@zephyr_z9 I hadn't listened to all in in a while and their reaction to him saying anthropic was $3T was the first time i've ever heard everyone gasp like that lmao.
Yeah he's huffing farts
@zephyr_z9 That guy just hates Chinese, anytime they do anything he will try to drag them down. If its so easy perhaps he can explain why Meta cant do it.
@zephyr_z9 But the way @GavinSBaker said it... and his facial expressions...
@advented_ @zephyr_z9 Have you used it? Or just commenting on it
@zephyr_z9 Nemotron is solid but having compute and building frontier models are different games. GPU inventory alone isn't a model strategy.
