An adapter from HBM to DIMM (DDR5/6) is probably pretty easy, it’s basically downclocking. Of course they may need to destroy them for ‘tax purposes’ (AKA trying to destroy general purpose computing for sweet datacentre subscriptions in this case).
My understanding is that they’re fundamentally different - HBM is a 3D printed stack of memory with a massive data bus, and has to live on-chip (like, millimeters away from the processor).
DDR5 has 2x32 bit channels. HBM4 has a massive 2,048 bits bus. There’s absolutely no way to run the number of traces it would need through a motherboard…
HBM4 has a massive 2,048 bits bus. There’s absolutely no way to run the number of traces it would need through a motherboard
Presumably you would put a chip next to it to ‘downmix’. Just saying, if there’s enough of it floating around unused, a way will be found. Chip wouldn’t need to do much, and now we’re back to normal traces, bit of a waste of potential bandwidth, but better than losing GP compute for the masses.
Could always do something like stick a Blackwell next to it and pop it on the PCIe bus and do a kernel mapping ;) (yes that’s a video card, or AI accelerator, that’s the joke, but also…)
An adapter from HBM to DIMM (DDR5/6) is probably pretty easy, it’s basically downclocking. Of course they may need to destroy them for ‘tax purposes’ (AKA trying to destroy general purpose computing for sweet datacentre subscriptions in this case).
My understanding is that they’re fundamentally different - HBM is a 3D printed stack of memory with a massive data bus, and has to live on-chip (like, millimeters away from the processor).
DDR5 has 2x32 bit channels. HBM4 has a massive 2,048 bits bus. There’s absolutely no way to run the number of traces it would need through a motherboard…
Presumably you would put a chip next to it to ‘downmix’. Just saying, if there’s enough of it floating around unused, a way will be found. Chip wouldn’t need to do much, and now we’re back to normal traces, bit of a waste of potential bandwidth, but better than losing GP compute for the masses.
Could always do something like stick a Blackwell next to it and pop it on the PCIe bus and do a kernel mapping ;) (yes that’s a video card, or AI accelerator, that’s the joke, but also…)