captjack on Nostr: Qwen3.6 35B-A3B on 4 GPUs how it performs ? CUDA TENSORFLOW TORCH 🟣 RTX 3090 — ...
Qwen3.6 35B-A3B on 4 GPUs
how it performs ? CUDA TENSORFLOW TORCH
🟣 RTX 3090 — 49.78 tok/s, TTFT 852ms
🟡 RTX 4090 — 118.93 tok/s, TTFT 686ms
🟢 RTX 5090 — 160.37 tok/s, TTFT 409ms
Published at
2026-04-17 15:36:27 UTCEvent JSON
{
"id": "2b449beeeacb67283328e740ca278ef921d34e4abd37c51f0effa40ce413ecf6",
"pubkey": "5e5fc1434c928bcdcba6f801859d5238341093291980fd36e33b7416393d5a2c",
"created_at": 1776440187,
"kind": 1,
"tags": [],
"content": "Qwen3.6 35B-A3B on 4 GPUs \n how it performs ? CUDA TENSORFLOW TORCH\n🟣 RTX 3090 — 49.78 tok/s, TTFT 852ms\n🟡 RTX 4090 — 118.93 tok/s, TTFT 686ms\n🟢 RTX 5090 — 160.37 tok/s, TTFT 409ms",
"sig": "fba6135e4e947959b675e5f8d8a6ddb50066075380a52a856eda4a84a4a07b3304a9c28d7470010f2f057fb8b1274bd80ee9df8d92f6144db52744be0c743aa6"
}
