I found it absolutely hilarious that this award is sponsored by Intel! Anyways, love where AMD is going. Super curious to see if their next GPUs will finally give Nvidia a run for their money at the high end.
Even if their hardware catches up with and beats that of NVidia, they will still have a huge problem with the software ecosystem.
Everyone who does anything serious with GPGPU and supercomputing is using CUDA, (a proprietary abomination) these days and AMD does not have a competitive product in that space.
That said, the software ecosystem around GPUs is an extraordinary mess, and there is therefore a huge opportunity for the company that can fix it.
Unfortunately, this is a deep cultural problem: as witnessed by the giant mess that is the Nvidia software ecosystem (nightmarish driver version management, proprietary technology, huge unwieldy binary distribution to install anytime you want to run 20 lines of GPGPU code, lack of properly working and debugging and profiling tools, etc...) companies run by EEs (such as Dr. Su) simply don't have the DNA to produce simple, clean, elegant, working software.
The kind of brain it takes to produce cutting edge silicon and the kind of brain it takes to produce cutting edge software are imo simply incompatible.
The list continues, and is a long one worthy of a short documentary. Apple excels at a few interface consistencies and at selling premium hardware with high margins in part due to an admittedly deserved proficiency with supply chain consistency & scale, and with convincing customers to pay with their kidneys for marginal upgrades e.g. 400% up charges on RAM and Storage.
The M1 is impressive, certainly agree, but it won’t be too long before others catch-up. Still, let’s not get ahead of ourselves about Apple and software prowess. It’s been a long time since I preferred Cupertino’s code to Redmond’s or Mountain View’s.
As someone on the outskirts of Fuchsia, looking on with longing, desperation, and confusion... do you have some connection to the mysterious pink project?
We are still waiting for other arm chips to catch up to A series iPhones.
Also, your hyperbole use of “sickening” betrays your extreme bias. Are you actually sickened by the overwrought animations?
I don't think I agree.
Apple is not really a great software company.
Especially if you consider how closed their ecosystem generally is: a large part of their stack is basically completely opaque, closed-source and worst of all, prone to change at the whim of the company.
I've always had a hard time understanding why anyone in the business of building technical infrastructure (other than user-facing retail stuff) would ever consider using their products.
iPod changed the world.
The iPhone Changed the world
Apple II changed the world.
Those successes were largely driven by good software.
Apple’s M series chips are only used in low end systems at the moment, so not sure what you meant by markup.
The ecosystem around AMD GPUs is quite small - and now that they seem to have abandoned OpenCL (possibly not their own fault though) - even that is put into question.
But things are bad even on the NVIDIA side. Example of how bad: I had to write my own C++ bindings for the CUDA runtime API (https://github.com/eyalroz/cuda-api-wrappers/). You'd think they would have that after 13 years of CUDA being available, right? Wrong. I repeatedly tried to pitch this to them, but they seem to suffer from the "Not Invented Here" syndrome (https://learnosity.com/not-invented-here-syndrome-explained/). This despite me having a lot of respect for people like Mark Harris, Bryce Lelbach, Duane Merrill et alia, and their work.
You're also rights about the "two kinds of brains" - or rather, it's not clear to me that the brains creating the silicon and the brains creating the software are in close enough cooperation.
By the way - it is possible to extract a pretty distribution of CUDA to justify run 20 lines of GPGPU code, from their installer. But they won't be bothered to package this nicely for you.
According to leaks their next generation GPUs will be multichip like their CPUs, and Nvidia needs 1 more generation to get there, so they might indeed get a step up on them.