Qwen3.6 27B

Q5_K_M·27B params·GGUF
reasoning
checkpoint: unsloth/Qwen3.6-27B-GGUF:Q5_K_M
commit: 82d411acf4a0
weights 18.17 GiB

All runs (19)

HardwareBackendModeShapeConc.Gen tok/sPrefill tok/sTTFTTPOT (ms)Prompt tokOut tokTotalVRAM Δ
GeForce RTX 3090 · 24 GiB450 Wdrv 590
llama.cpp cuda-4f13cb7 (cuda)baseline-pl-450wcodegen1
36.1
206.6348ms26.562100027.70s0.000 GiB
GeForce RTX 3090 · 24 GiB350 Wdrv 590
llama.cpp cuda-4f13cb7 (cuda)baseline-pl-350wcodegen1
35.7
193.1323ms27.062100028.01s0.000 GiB
GeForce RTX 3090 · 24 GiB450 Wdrv 590
llama.cpp cuda-4f13cb7 (cuda)baseline-pl-450wagent1
35.1
1124.9533ms26.659950014.23s0.000 GiB
GeForce RTX 3090 · 24 GiB350 Wdrv 590
llama.cpp cuda-4f13cb7 (cuda)baseline-pl-350wagent1
34.8
1161.2516ms27.059950014.37s0.000 GiB
GeForce RTX 3090 · 24 GiB450 Wdrv 590
llama.cpp cuda-4f13cb7 (cuda)baseline-pl-450wchat1
34.2
117.8256ms26.2301002.93s0.000 GiB
GeForce RTX 3090 · 24 GiB350 Wdrv 590
llama.cpp cuda-4f13cb7 (cuda)baseline-pl-350wchat1
33.5
119.4251ms26.7301002.99s0.000 GiB
GeForce RTX 3090 · 24 GiB450 Wdrv 590
llama.cpp cuda-4f13cb7 (cuda)baseline-pl-450wrag1
32.7
1218.8751ms26.48422006.12s0.000 GiB
GeForce RTX 3090 · 24 GiB350 Wdrv 590
llama.cpp cuda-4f13cb7 (cuda)baseline-pl-350wrag1
31.1
1316.7895ms26.98422006.42s0.000 GiB
GeForce RTX 3090 · 24 GiB200 Wdrv 590
llama.cpp 59778f0 (cuda)baselinecodegen1
18.9
187.0375ms52.462100052.77s0.000 GiB
GeForce RTX 3090 · 24 GiB200 Wdrv 590
llama.cpp 59778f0 (cuda)baselinechat1
18.6
112.9268ms50.1301005.38s0.000 GiB
GeForce RTX 3090 · 24 GiB200 Wdrv 590
llama.cpp 59778f0 (cuda)baselineagent1
18.2
517.01.16s52.759950027.44s0.000 GiB
GeForce RTX 3090 · 24 GiB200 Wdrv 590
llama.cpp 59778f0 (cuda)baselinerag1
17.3
756.41.20s51.884220011.56s0.000 GiB
GeForce RTX 3090 · 24 GiB450 Wdrv 590
llama.cpp cuda-4f13cb7 (cuda)baseline-pl-450wagent4
14.6
32.721.75s26.659950035.52s0.000 GiB
GeForce RTX 3090 · 24 GiB350 Wdrv 590
llama.cpp cuda-4f13cb7 (cuda)baseline-pl-350wagent4
14.6
33.621.54s27.059950035.39s0.000 GiB
Strix Halo · Radeon 8060S · 128 GiB unified (96 GiB VRAM)unifieddrv 7
llama.cpp rocm-4f13cb7 (rocm)baselinecodegen1
10.6
145.5440ms93.862100094.20s0.017 GiB
Strix Halo · Radeon 8060S · 128 GiB unified (96 GiB VRAM)unifieddrv 7
llama.cpp rocm-4f13cb7 (rocm)baselineagent1
10.6
2036.6294ms93.859950047.21s0.010 GiB
Strix Halo · Radeon 8060S · 128 GiB unified (96 GiB VRAM)unifieddrv 7
llama.cpp rocm-4f13cb7 (rocm)baselinechat1
10.4
83.9358ms93.5301009.63s0.004 GiB
Strix Halo · Radeon 8060S · 128 GiB unified (96 GiB VRAM)unifieddrv 7
llama.cpp rocm-4f13cb7 (rocm)baselinerag1
9.9
452.51.61s93.884220020.26s0.007 GiB
Strix Halo · Radeon 8060S · 128 GiB unified (96 GiB VRAM)unifieddrv 7
llama.cpp rocm-4f13cb7 (rocm)baselineagent4
4.4
8.870.91s93.8599500117.83s0.037 GiB

Environment

GeForce RTX 3090 · 24 GiB
cpuAMD EPYC 7302P 16-Core Processor
gpuNVIDIA GeForce RTX 3090
archNVIDIA
vram24 GiB (system 64.0 GiB)
power350 W / 450 W max(78% cap)
pcieGen 4 x16 / Gen 4 x16 max
clocksgfx 1965/2100 MHz · mem 9501 MHz
temp44°C idle · 64°C peak
peak draw337 W
backendllama.cpp cuda-4f13cb7 (cuda)
serverlemonade unknown
osUbuntu 24.04 LTS
kernel6.17.13-7-pve
driverNVIDIA 590.48.01 + CUDA 13.1
libc2.39
python3.12.3
containerizedtrue
llama.cppversion: 18 (4f13cb7) built with GNU 13.3.0 for Linux x86_64
build flagsGGML_CUDA=ON CMAKE_BUILD_TYPE=Release
runs/cell5
warmups2
endpoint/v1/chat/completions
streamingtrue
GeForce RTX 3090 · 24 GiB
cpuAMD EPYC 7302P 16-Core Processor
gpuNVIDIA GeForce RTX 3090
archNVIDIA
vram24 GiB (system 64.0 GiB)
power450 W / 450 W max
pcieGen 4 x16 / Gen 4 x16 max
clocksgfx 1965/2100 MHz · mem 9501 MHz
temp43°C idle · 83°C peak
peak draw430 W
backendllama.cpp cuda-4f13cb7 (cuda)
serverlemonade unknown
osUbuntu 24.04 LTS
kernel6.17.13-7-pve
driverNVIDIA 590.48.01 + CUDA 13.1
libc2.39
python3.12.3
containerizedtrue
llama.cppversion: 18 (4f13cb7) built with GNU 13.3.0 for Linux x86_64
build flagsGGML_CUDA=ON CMAKE_BUILD_TYPE=Release
runs/cell5
warmups2
endpoint/v1/chat/completions
streamingtrue
GeForce RTX 3090 · 24 GiB
cpuAMD EPYC 7302P 16-Core Processor
gpuNVIDIA GeForce RTX 3090
archNVIDIA
vram24 GiB (system 64.0 GiB)
power200 W / 450 W max(44% cap)
backendllama.cpp 59778f0 (cuda)
serverlemonade unknown
osUbuntu 24.04 LTS
kernel6.17.13-7-pve
driver590.48.01
python3.12.3
containerizedtrue
runs/cell5
warmups2
endpoint/v1/chat/completions
streamingtrue
Strix Halo · Radeon 8060S · 128 GiB unified (96 GiB VRAM)
cpuAMD RYZEN AI MAX+ 395 w/ Radeon 8060S
gpuAMD Radeon 8060S
archStrix Halo (gfx1151)
vram96 GiB (system 31.1 GiB, unified)
pcieGen 4 x16 / Gen 4 x16 max
clocksgfx 1320 MHz · mem 1000 MHz
temp47°C idle · 75°C peak
peak draw99 W
backendllama.cpp rocm-4f13cb7 (rocm)
serverlemonade unknown
osUbuntu 24.04 LTS
kernel7.0.2-2-pve
driverROCm 7.2.3
libc2.39
python3.12.3
containerizedtrue
llama.cppversion: 1 (4f13cb7) built with Clang 22.0.0 for Linux x86_64
build flagsGGML_HIP=ON AMDGPU_TARGETS=gfx1151 CMAKE_BUILD_TYPE=Release
runs/cell5
warmups2
endpoint/v1/chat/completions
streamingtrue