llamaperf

Qwen3.8 27B Swift-1.5

on AMD RX 9060 XT 16GB · llama.cpp · 98,304 ctx

Tone: positiveAMD hardware
Nov 5, 2026
Throughput
20-29 t/s gen
Quant
IQ3_S (GGUF)
KV cache
q5_0/q4_1
System RAM
32 GB
VRAM reported
16 GB

Use cases

codingagentictool-usevision

Summary

User reports running Swift-1.5-Qwen3.8-27B at IQ3_S on an RX 9060 XT 16GB, with eval in the mid 20s t/s and prompt processing between 550 and 231 t/s. Setup is llama.cpp (llama-server) with MTP draft speculation, q5_0 K and q4_1 V caches, 98304 context, all layers on the GPU and the mmproj kept on the CPU. Draft acceptance was around 0.6-0.8. A web-search agent request took 2m 10s end to end; the user has seen mid-40s eval at other settings. They ask whether to shrink context, add draft KV quantisation, and how to handle vision.

Compare this run

Medians of plain runs: one device, one request, no speculative decoding, the whole model in the device's memory. The report above is one setup, so read it beside them and not as one of them.

Qwen3.8 27B at 4-bit by hardware

Bars are medians on one scale, 0 to 67.0 tokens/s. Press ▶ to watch text arrive at a row's speed.

Models most run on the AMD RX 9060 XT 16GB

Bars are medians on one scale, 0 to 33.5 tokens/s. Press ▶ to watch text arrive at a row's speed.

Everything reported on the AMD RX 9060 XT 16GB →Every Qwen3.8 report →Estimate Qwen3.8 on this card →