Radeon AI PRO R9700
Four Radeon GPUs Power an 8-User Local AI Server for C$13,000
A four-GPU Radeon server costing about C$13,000 initially served 8 users with 128K-token contexts, but worst-case first-token latency hit 2.5 minutes.
1 article tagged with this topic
A four-GPU Radeon server costing about C$13,000 initially served 8 users with 128K-token contexts, but worst-case first-token latency hit 2.5 minutes.