-17 karma · joined August 27, 2019
I’ve got both (single R9700, dual B70) and they do nicely for about anything I throw at them, such that the latter has a visible improvement when the model is well-cached.
One of the notable figures in that, Carly Fiorina, made a point to “forget the engineers”.
We aren’t captured by environmental activists that force the poor to shoulder the compliance burden while the rich get to defer and delay.
As for software, anything that has a SYCL or Vulkan backend, and/or can be Intel optimized (especially to the same degree as llama.cpp) can run well.
I have a pair of them with a 9480 and the only thing I have to do is keep the cache happy.
No thanks, but I’d rather not have mysterymeat patches from effectively impossible to replicate models.
Did they just follow on from New Zealand?
At this point, they’re at the Extinguish stage.
I had to rebuild llama.cpp from source with the SYCL and CPU specific backends.
Started with a barebones Ubuntu Server 24 LTS install, used the HWE kernel, pulled in the Intel dependencies for hardware support/oneapi/libze, then built llama.cpp with the Intel compiler (icx?) for the SYCL and NATIVE backends (CPU specific support).
In short, built it based mostly on the Intel instructions.
For some workloads, the Arc Pro B70 actually does reasonably well when cached.
With some reasonable bring-up, it also seems to be more usable versus the 32gb R9700.
IRGC sympathizers across the world that would rather have the current government than the more progressive predecessor.
…at the end stage.
You would have to openly make them a protected class before that would happen.
Aside from public figures, that is very much in the realm of doxxing.