Posted by networked 5 hours ago
I understand why the AI people want it to help with the latency between the CPU and GPU. But having people in some cases use it as for marketing is kinda like getting watered down beer and getting told its a good thing because alcohol is bad for you.
The reason AI users like it is because it provides a large amount of higher-bandwidth memory at a price much lower than an equivalent GPU.
If you had 16GB of DDR5 in a normal 2-channel consumer board it would not be this fast.
This gives 16GB of high speed memory for a very low price compared to a full GPU.
The larger unified memory systems like Strix Halo or Apple Silicon provide access to high bandwidth memory at scales you can’t get with any consumer GPU.
> But having people in some cases use it as for marketing is kinda like getting watered down beer and getting told its a good thing because alcohol is bad for you
I don’t think you understand the benefits of having high speed memory attached at a low price point.
Yes we would all prefer to have the same amount of even faster memory attached to a dedicated GPU, but that’s not happening at this price point.
Better analogy would be that this is an entry-level sports car for the price of a Honda Civic but you’re complaining that it’s not a Lamborghini.
But unified memory isn't providing high bandwidth. Apple's specific implementation on pro or higher chips does. There's a zillion desktops and laptops running unified memory at low to medium-low speeds, and the baseline M series is part of that with its 128 bit bus.
Most unified memory devices are handicapped because of the unification, not boosted. Yet people keep praising unified memory.
Primarily the bottleneck actually. PCIe bandwidth is extremely limited.
GDDR has worse latency than DDR and scaling it up to larger capacities faces challenges. So if you unify memory your options are small GDDR, large DDR, or expensive HBM.
Unusually, the BC-250 and PS5 both use GDDR with the CPU so the memory bandwidth is great but the latency not so much AFAIK. It's like using a dGPU as a full computer.
I guffawed. And then closed the tab.
Been really fun watching this go from e-waste to having mesa support and CU/CPU unlocks.
Not discussed much in the article, but it looks like:
TSMC → AMD controlled/qualified the usable off-spec parts → AMD supplied BC-250 parts to ASRock → ASRock sold mining hardware.