News Tiny Corp is 70% confident that AMD will make at least some of its GPU firmware open source

[endocine]

I’m just interested to know why you consider a dude selling a rig with 6 consumer gaming GPUs as a commercial AI development platform is a user? He’s a reselling huckster ripping people off (including AMD realistically). He’s able to sell $8000 worth of hardware for $15k because it’s still cheaper than buying hardware with commercial support. But then he expects commercial support. AMD needs to just stick to their game plan instead of dealing with all the crooks cashing in on the AI boom.

Who are they ripping off and how, is there something criminal about their activities? And what exactly is AMD's game plan, to be a purveyor of bad software that has no quality control that no one dare question? AI isn't a dubious activity like cryptomining.

 

[bit_user]

Who are they ripping off and how, is there something criminal about their activities?

I think @Pierce2623 overstated their case, but I take the point as being that Tiny shouldn't act like they're doing AMD a huge favor, since what they're selling will likely be a fairly insignificant volume of lower-margin gaming GPUs. We'll see, but they're certainly not as beneficial as all the workstation & server partners AMD has that are actually selling Pro and MI-series cards.

And what exactly is AMD's game plan, to be a purveyor of bad software that has no quality control

If it were that bad, then nobody could use it for anything! Software of this complexity needs to be of a pretty high standard, in order to be at all usable. Furthermore, every production codebase of such a size and complexity will always have bugs. This is just a fact of life. As a business - or even just an end user grounded in the real world - one cannot have an expectation that there are no bugs. Rather, the expectation should be that the developer is reasonably responsive and successful at providing workable solutions. In order for that to happen, cooperation (and a little patience) is needed from both parties.

As I mentioned, Lamini has shown that ROCm and AMD hardware is indeed capable of providing a production-grade platform for LLMs.

Lamini - Enterprise LLM Platform

Lamini is the enterprise LLM platform for existing software teams to quickly develop and control their own LLMs. Lamini has built-in best practices for specializing LLMs on billions of proprietary documents to improve performance, reduce hallucinations, offer citations, and ensure safety. Lamini...

www.lamini.ai

I feel like you two are both being hyperbolic, at opposite ends of the spectrum, and that doesn't help anyone.

... that no one dare question?

Plenty of people have written much criticism of ROCm and AMD's GPU Compute strategy, over the years - myself included! I don't see anyone in this thread saying that Tiny (or anyone else) should just "shut up and accept what you're given", if that's what you're implying.