Hacker News new | past | comments | ask | show | jobs | submit login

Real business-class features we want to know about:

Will they auto-detect workloads and cripple performance (like the mining stuff recently)? Only work through special drivers with extra licensing feeds depending on the name of the building it is in (data center vs office)?




Market segmentation is practiced by every chip company that you use. Intel: ECC. AMD: ROCM. Qualcomm: cost as percentage of the phone price.


I still think Nvidia takes it further.


Every company does market segmentation: it makes sense to have customers that want a feature pay more for it.

Still, every company does it differently.

For example, both NVIDIA and AMD compute GPUs are necessarily more expensive than gamer GPUs because of hardware costs (e.g. HBM).

However, NVIDIA gamer GPUs can do CUDA, while AMD gamer GPUs can't do ROCm.

The reason is that NVIDIA has 1 architecture for gaming and compute (Ampere), while AMD has two different architectures (RDNA and CDNA).


It's common, but only possible in a very dominant position or with competitors that are borderline colluding.


You must be the only gamer in the world that wants an HBM2e GPU for gaming that's 10x more expensive while only delivering a negligible improvement in FPS.


I'm only talking about driver/license locks, not different ram types.


Can the CDNA GPUs from AMD even connect to a monitor ?

I don't think they even have display ports.

Not sure what good would a "gaming driver" do you on those cards.

Same for the opposite. Do the RDNA GFX cards have even hardware for compute? They don't even have tensor cores, so why would AMD invest money into creating a compute driver for hardware that's bad at compute?

> I'm only talking about driver/license locks,

Not "locked" is a big understatement. A driver release for some hardware needs at least some QA, so the assumption that doing this is just "free" because its software is incorrect.


>A driver release for some hardware needs at least some QA, so the assumption that doing this is just "free" because its software is incorrect.

Nvidia detects mining workloads in software based on heuristics and disables them. Probably causes more support burden than less and took extra engineering time to implement, not less.


> Probably causes more support burden than less and took extra engineering time to implement, not less.

Given how well that turn out, I have a hard time believing they put any effort into this.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: