dev.to: An MI300X Over MCP: What the Matrix Cores Execute, and What They Don't
retained text
One AMD Instinct MI300X on AMD Developer Cloud, managed entirely through a tag-scoped Python MCP server, with every figure read off the card rather than a spec sheet. fp8 e4m3fnuz runs 1.77x bf16; int8, which AMD rates identically to fp8, runs 0.69x; fp4 is not on this silicon at all. One droplet, $1.99 an hour, and two readings that were wrong the first time. tags: amd, mcp, rocm, machinelearning reactions=11 | comments=5
document fetched 2026-09-24 02:33