Not to sound too negative but IMO NVidia has largely been "right place, right time" and I don't think Huang has had much to do with it. NVidia made computer graphics cards and CUDA long ago; they just happened to be in the right place, twice, in quick succession: first crypto/blockchain, then AI, neither of which they really had any part in boosting. It's like they set up a general store that sells pickaxes and drills, then someone found gold in the mountain right next to them, and then the goldmine ran out but someone found oil right under them.

The one big thing I will given Huang credit for is not going all-in on blockchain and recognizing it as a bubble. NVidia never really pivoted from their core business/customers to support crypto miners, unlike what they're doing with AI.

There were a lot of computer graphics cards manufacturers back in the 90's. Nvidia survived amongst the whole pack of them. 3dfx, Matrox, S3, ATI, Cirrus. How much of that is soley due to Jenson vs outside factors outside of his control, only he really knows, but I think her deserves a bit more credit than right place right time. CUDA was a very expensive investment with zero guarantee of payoff. The market could have gone an entirely different way and Nvidia would have failed.

> first crypto/blockchain, then AI, neither of which they really had any part in boosting.

Nvidia boosts AI a lot. They are the central bank of AI. https://news.ycombinator.com/item?id=49673098

Yeah, they had been working on CUDA for a very long time before AI turned out to be interesting again.

To be honest other hardware companies had years to catch up to Cuda and ML, but they invested shit into their software stacks.

Even when AI boom was imminient but 10 times smaller 2-3 years ago they did not invested heavily enough and Nvidia went all in.

I have no clue what percent of that decision making was Huang, but all the competition leadership was certainly incompetent.

So at least he did built company with a good leadership. Its a fact. Even if I personally dont like Nvidia or its products.

> they just happened to be in the right place, twice, in quick succession

that is generally what it looks like when someone is smarter than you

The had the insight to develop and promote CUDA long before DL was a thing. IIRC AlexNet was when CUDA really paid off.

Yeah, but in this case they also built the pickaxes and drills and did it better than any other pickaxe and drill builder. In your analogy, they are commodities and a shop like that can be set up by anyone. But in this case, that doesn't apply. And if you look at their sales composition, they never pivoted away from gaming, but data centers clearly became ludicrously more attractive in the AI era. But they were always very good at computing triangles, perhaps the best in the world. They were lucky that market forces made that computation very valuable but they were not lucky in their ability to build it.