There was an obvious problem in the original release, they re issued it after like a week with the reasoning looping supposedly fixed.

No it doesn't follow instructions and is substantially slower than ds4.

Yeah I think it got bad press because the chat templates (or something?) were messed up on first release, but I've been using a quant of it and it's a powerhouse, better than qwen 3.6 27b for local on a 3090, which is saying a lot.

I think glancing at a random snapshot from today misses all the context. Nemotron 3 is far more significant than you're giving it credit for.

At this point, Nemotron 3 is really an 8 month old model series. That's when Nemotron 3 Nano was released, and the Nemotron 3 Super/Ultra models this year are obviously based on that recipe, mostly just bigger with a few tweaks here and there. Against today's models, no, not that interesting. Each of the Nemotron 3 models were briefly competitive when they launched, but never exceptional, and less competitive with each scale up. The fact that it took so long for Nemotron 3 Ultra to launch really hampered its competitiveness.

The Nemotron 3 series is extremely open about training recipes and training data, far more open than most open weight models, and that is valuable.

Before Nemotron 3, Nvidia had never released a single LLM that I would consider interesting at all, so Nemotron 3 was a big step up. The closest thing was Mistral NeMo, but a significant part of the credit there goes to the Mistral team, not Nvidia.

Given how much Nemotron 3 improved, I'm curious to see if Nemotron 4 will take them to a leading edge level instead of just briefly competitive.

(Nvidia released a Nemotron 3 and a Nemotron 4 like 3 years ago... this year's Nemotron 3 is entirely unrelated. Nvidia's naming schemes leave a little bit to be desired.)

In the 1980s databases were dominated by private companies like Oracle and IBM and over time we ended up with MySQL/Postgres.

If open models end up being the main endgame I don't see why American companies won't just end up funding talent to build open models, just like databases and languages and other fundamental tech they build off of. There will always be cutting-edge commercial variations for the important stuff and maybe open models will take over general business and public work, while Apple and Google will eventually have models built into phones, etc.

Ultimately we have no idea what the future will look like because the market is so immature. It's fair for the US government to experiment early on with open models, but at this point we're pretty far from a situation where the US needs to rush in. China's government put billions into their own private industry just to help them compete, but they will have to pay those bills eventually just the same.