[Thefuckrr]

Hey folks,

The recent 40- and 50-series cards boast insane specs, but they’re also priced higher than ever, and many of the architectural improvements (like tensor cores and massive VRAM) seem clearly geared toward AI and deep learning use cases.

NVIDIA is slowly pivoting from being a gamer-first company to an AI-focused hardware giant. Gamers benefit from the trickle-down, but we’re no longer the main audience and the pricing reflects that.

Are we as gamers and creators still the target market for NVIDIA GPUs—or are we just along for the ride while the AI industry drives the future of graphics hardware?

[donPo]

AI is defintely (becoming) the bigger market.
NVIDIA has an API (lay term: library) which allows to easily program directly on the GPU. It's called CUDA.

As NVIDIA was the first to publish such an API, many other libraries that allow to write AI program, are built on this 'NVIDIA-only' API. This is the reason why NVIDIA is by far the biggest in AI development.
I don't think it will happen soon that other such APIs of other companies will be implemented in those higher level libraries, because of the overhead.

The market for AI is way bigger than that for gaming. The amount of GPU's e.g. openai needs is both bigger and easier to handle logistically compared to individual costumers wanting a GPU for gaming.

AMD will probably have to accept their loss, for now, in this AI battle, meaning that they will now probably focus on gaming mainly.

Tldr: for the short period, NVIDIA will be focusing on AI and AMD on gaming. On the longer period, it's waiting untill another company can break NVIDIA's dominance on the AI market.