As can be seen from the diagramHowever, it is not clear how to use multiple CCDs to run one instance faster.
the first generation Epyc servers are essentially a cluster of eight 8-core processors.
This makes me wonder whether this thread is somehow back on topic and distributed llama
https://github.com/b4rtaz/distributed-llama
could also be used to obtain faster speeds for the Epyc. It seems worth a try, especially as most of the required steps would be the same for a cluster of Raspberry Pi computers.
Statistics: Posted by ejolson — Mon Apr 14, 2025 8:51 pm