Comment by pjc50
11 hours ago
I've observed this as well, that software in a hardware org is a bit of a second class citizen. The absolute worst case is AMD leaving a trillion dollars on the table because they can't compete with CUDA software APIs, but lots of places are like this.
But software in general - well, in America - got pulled up into the stratosphere by FAANG money. I feel that should have had more of an effect than it did on non-software orgs.
> The absolute worst case is AMD leaving a trillion dollars on the table because they can't compete with CUDA software APIs, but lots of places are like this.
I’m still so dumbfounded by this. It’s almost 20 years since NVidia introduced CUDA. Developer tooling / experience appears to be something AMD does not understand, for some reason.
20 years ago is also the time frame for when AMD acquired ATi (who IIRC were 99% on gaming graphics), and AMD was floundering in the following decade. They made the choice to prioritize the CPU side of the business, but on the GPU graphics/compute side it's hard to see that they've got much vision for how they want to steer the future to go and the ability to make that happen with their partners.
It might be related to patent portfolios. AMD might be reluctant to pursue something that can step on Nvidia's minefield. OTOH, you mention developer experience, and it'd be wonderful to have something less developer hostile than CUDA.
I remember my feelings when I learned how to use the Cell's SPUs and how much I didn't want to touch it with a barge pole after that.
If that is the only issue, any initial CUDA patents will be expiring now
> that software in a hardware org is a bit of a second class citizen
I noticed that with mainframes and banks.
IBM makes some really amazing hardware at the very top of the market, but the companies who own those machines don't seem to think any competitive advantage can come from them - they are the cost of doing business. Because of that, the mainframe teams are often neglected.
I would even be happy to write code on the least sexy language ever invented, COBOL, just so it could run on the sexiest hardware ever built.