llamaperf

Qwen3.8 27B

on RTX 5000 Pro · SGLang

Tone: positive
Sep 7, 2026
Throughput
130.0 t/s gen · 4000.0 t/s pp
VRAM reported
48 GB

Summary

User reports Qwen3.8-27B at 130 t/s decode and 4000 t/s prefill on an RTX 5000 Pro with 48 GB VRAM, at 5 concurrencies. Setup is SGLang with prefix offload to storage. The user compares the result to a Claude subscription and notes the card's price has doubled since purchase.