Llamabench6.2 Statistics v2

GPUs GPUsx Model TG PP TTFT Watts LDU System Data Created
[ROCm0] AMD Radeon Graphics (32732 MiB free)
[ROCm1] AMD Radeon VII (16340 MiB free)
2 Bielik-11B-v3.0-Instruct-Q8_0.gguf TG = 43.32 PP = 295.76 TTFT = 391.09 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-31 17:41:35
[ROCm0] AMD Radeon VII (16116 MiB) [[max Watt:27W]] 1 Bielik-11B-v3.0-Instruct-Q8_0.gguf TG = 37.79 PP = 287.32 TTFT = 472.98 308.0 W ROCM
OS: Linux-6.17.0-22-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-23 13:43:08
[ROCm0] AMD Radeon Graphics (32428 MiB free) 1 Bielik-11B-v3.0-Instruct-Q8_0.gguf TG = 44.39 PP = 300.90 TTFT = 383.81 160 W ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-29 20:55:05
[CUDA0] NVIDIA GeForce RTX 4070 (11127 MiB free)
[RPC0] 127.0.0.1:50052 (32732 MiB free)
2 Bielik-11B-v3.0-Instruct-Q8_0.gguf TG = 37.43 PP = 359.90 TTFT = 572.42 N/A CUDA + RPC
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 18:41:46
[Vulkan0] AMD Radeon VII (RADV VEGA20) (15572 MiB free)
[Vulkan1] AMD Radeon Graphics (RADV VEGA20) (32751 MiB free)
2 Bielik-11B-v3.0-Instruct-Q8_0.gguf TG = 27.94 PP = 56.86 TTFT = 235.69 N/A VULKAN
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-01 15:02:36
[CUDA0] NVIDIA GeForce RTX 4070 (11066 MiB free) 1 Bielik-11B-v3.0-Instruct.Q4_K_M.gguf TG = 63.40 PP = 2716.68 TTFT = 55.13 200 W CUDA
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 19:47:36
[CUDA0] NVIDIA GeForce RTX 4070 (11294 MiB free) 1 Bielik-11B-v3.0-Instruct.Q4_K_M.gguf TG = 30.04 PP = 1373.55 TTFT = 94.35 100 W CUDA
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-29 19:57:50
[ROCm0] AMD Radeon Graphics (32732 MiB free)
[ROCm1] AMD Radeon VII (16340 MiB free)
2 Bielik-11B-v3.0-Instruct.Q4_K_M.gguf TG = 48.51 PP = 501.29 TTFT = 251.94 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-31 17:36:25
[ROCm0] AMD Radeon Graphics (32428 MiB free) 1 Bielik-11B-v3.0-Instruct.Q4_K_M.gguf TG = 50.14 PP = 509.49 TTFT = 247.66 160 W ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-29 20:53:16
[CUDA0] NVIDIA GeForce RTX 4070 (11011 MiB free)
[RPC0] 127.0.0.1:50052 (32732 MiB free)
2 Bielik-11B-v3.0-Instruct.Q4_K_M.gguf TG = 47.69 PP = 597.77 TTFT = 359.39 N/A CUDA + RPC
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 19:17:46
[Vulkan0] AMD Radeon VII (RADV VEGA20) (15698 MiB free)
[Vulkan1] AMD Radeon Graphics (RADV VEGA20) (32751 MiB free)
2 Bielik-11B-v3.0-Instruct.Q4_K_M.gguf TG = 27.79 PP = 188.96 TTFT = 729.17 N/A VULKAN
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-03 18:11:29
[CUDA0] NVIDIA GeForce RTX 4070 (11062 MiB free) 1 Bielik-4.5B-v3.0-Instruct.Q8_0.gguf TG = 80.13 PP = 5415.43 TTFT = 50.63 N/A CUDA
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 19:53:38
[ROCm0] AMD Radeon Graphics (32732 MiB free)
[ROCm1] AMD Radeon VII (16340 MiB free)
2 Bielik-4.5B-v3.0-Instruct.Q8_0.gguf TG = 41.35 PP = 780.28 TTFT = 292.35 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-31 17:54:21
[ROCm0] AMD Radeon Graphics (32428 MiB free) 1 Bielik-4.5B-v3.0-Instruct.Q8_0.gguf TG = 70.92 PP = 646.84 TTFT = 340.81 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-29 20:55:41
[CUDA0] NVIDIA GeForce RTX 4070 (10979 MiB free)
[RPC0] 127.0.0.1:50052 (32732 MiB free)
2 Bielik-4.5B-v3.0-Instruct.Q8_0.gguf TG = 56.89 PP = 796.45 TTFT = 502.65 N/A CUDA + RPC
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 19:14:37
[ROCm0] AMD Radeon Graphics (32732 MiB free) 1 Bielik-7B-v3.0-Instruct-GGUF.Q8_0.gguf TG = 59.52 PP = 185.65 TTFT = 71.09 N/A ROCM
OS: Linux-6.17.0-20-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-08 09:53:07
[ROCm0] AMD Radeon Graphics (32732 MiB) [[max Watt:225W]] 1 Bielik-7B-v3.0-Instruct-GGUF.Q8_0.gguf TG = 58.86 PP = 167.73 TTFT = 71.55 299.0 W ROCM
OS: Linux-6.17.0-20-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-08 19:26:06
[ROCm0] AMD Radeon Graphics (32732 MiB free)
[ROCm1] AMD Radeon VII (16340 MiB free)
2 FINAL-Bench_Darwin-35B-A3B-Opus-Q8_0.gguf TG = 58.03 PP = 645.62 TTFT = 723.84 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-02 22:10:38
[CUDA0] NVIDIA GeForce RTX 4070 (11111 MiB free)
[RPC0] 127.0.0.1:50052 (32732 MiB free)
2 GLM-4.7-Flash-MXFP4_MOE.gguf TG = 46.93 PP = 810.10 TTFT = 304.27 N/A CUDA + RPC
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 18:42:52
[ROCm0] AMD Radeon Graphics (32732 MiB free)
[ROCm1] AMD Radeon VII (16340 MiB free)
2 GLM-4.7-Flash-Q8_0.gguf TG = 70.95 PP = 616.02 TTFT = 246.40 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-31 17:56:50
[CUDA0] NVIDIA GeForce RTX 4070 (11112 MiB free)
[RPC0] 127.0.0.1:50052 (32732 MiB free)
2 GLM-4.7-Flash-Q8_0.gguf TG = 45.91 PP = 692.57 TTFT = 347.59 N/A CUDA + RPC
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 18:44:42
[ROCm0] AMD Radeon Graphics (32732 MiB free)
[ROCm1] AMD Radeon VII (16340 MiB free)
2 GLM-4.7-Flash-UD-Q4_K_XL.gguf TG = 59.01 PP = 912.12 TTFT = 195.31 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-31 17:58:21
[ROCm0] AMD Radeon Graphics (32732 MiB free) 1 GLM-4.7-Flash-UD-Q4_K_XL.gguf TG = 64.92 PP = 942.26 TTFT = 191.84 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 20:17:03
[CUDA0] NVIDIA GeForce RTX 4070 (11104 MiB free)
[RPC0] 127.0.0.1:50052 (32732 MiB free)
2 GLM-4.7-Flash-UD-Q4_K_XL.gguf TG = 47.62 PP = 173.79 TTFT = 116.21 N/A CUDA + RPC
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 18:47:56
[CUDA0] NVIDIA GeForce RTX 3050 Laptop GPU (3299 MiB free) [[max Watt:69W]] 1 Llama-3.2-3B-Instruct-Q4_K_M.gguf TG = 63.23 PP = 372.00 TTFT = 26.37 122.4 W CUDA
OS: Windows-10-10.0.19045-SP0
CUDA: 591.59
MMAP: false
2026-04-08 17:45:33
[CUDA0] NVIDIA GeForce RTX 4070 (11059 MiB free) 1 NVIDIA-Nemotron3-Nano-4B-Q4_K_M.gguf TG = 135.53 PP = 4508.39 TTFT = 104.29 N/A CUDA
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 19:53:59
[ROCm0] AMD Radeon Graphics (32732 MiB free)
[ROCm1] AMD Radeon VII (16340 MiB free)
2 NVIDIA-Nemotron3-Nano-4B-Q4_K_M.gguf TG = 91.72 PP = 868.68 TTFT = 541.07 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-31 18:07:48
[ROCm0] AMD Radeon Graphics (32732 MiB free)
[ROCm1] AMD Radeon VII (16340 MiB free)
2 Nemotron-3-Nano-30B-A3B-IQ4_NL.gguf TG = 93.63 PP = 765.51 TTFT = 613.47 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-31 18:02:35
[ROCm0] AMD Radeon Graphics (32732 MiB free) 1 Nemotron-3-Nano-30B-A3B-IQ4_NL.gguf TG = 104.73 PP = 778.66 TTFT = 603.41 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 20:17:44
[CUDA0] NVIDIA GeForce RTX 4070 (11100 MiB free)
[RPC0] 127.0.0.1:50052 (32586 MiB free)
2 Nemotron-3-Nano-30B-A3B-IQ4_NL.gguf TG = 80.77 PP = 915.43 TTFT = 513.22 N/A CUDA + RPC
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 18:49:24
[ROCm0] AMD Radeon Graphics (32732 MiB free)
[ROCm1] AMD Radeon VII (16340 MiB free)
2 Nemotron-3-Nano-30B-A3B-Q4_K_M.gguf TG = 87.93 PP = 734.42 TTFT = 638.89 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-31 18:03:56
[ROCm0] AMD Radeon Graphics (32732 MiB free) 1 Nemotron-3-Nano-30B-A3B-Q4_K_M.gguf TG = 97.48 PP = 745.46 TTFT = 630.17 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 20:18:37
[CUDA0] NVIDIA GeForce RTX 3050 Laptop GPU (3299 MiB free)
[RPC0] 192.168.0.222:50052 (27365 MiB free)
2 Nemotron-3-Nano-30B-A3B-Q4_K_M.gguf TG = 20.72 PP = 135.97 TTFT = 3459.82 N/A CUDA + RPC
OS: Windows-10-10.0.19045-SP0
CUDA: 591.59
MMAP: false
2026-03-28 13:56:23
[CUDA0] NVIDIA GeForce RTX 4070 (11096 MiB free)
[RPC0] 127.0.0.1:50052 (32586 MiB free)
2 Nemotron-3-Nano-30B-A3B-Q4_K_M.gguf TG = 78.32 PP = 877.17 TTFT = 535.84 N/A CUDA + RPC
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 18:51:23
[ROCm0] AMD Radeon Graphics (32732 MiB free)
[ROCm1] AMD Radeon VII (16340 MiB free)
2 Nemotron-3-Nano-30B-A3B-UD-Q8_K_XL.gguf TG = 76.36 PP = 656.73 TTFT = 714.76 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-31 18:04:58
[CUDA0] NVIDIA GeForce RTX 4070 (11126 MiB free)
[RPC0] 127.0.0.1:50052 (32586 MiB free)
2 Nemotron-3-Nano-30B-A3B-UD-Q8_K_XL.gguf TG = 66.50 PP = 776.82 TTFT = 604.57 N/A CUDA + RPC
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 16:10:40
[CUDA0] NVIDIA GeForce RTX 4070 (11108 MiB free)
[RPC0] 127.0.0.1:50052 (32586 MiB free)
2 Nemotron-3-Nano-30B-A3B-UD-Q8_K_XL.gguf TG = 66.32 PP = 769.75 TTFT = 610.64 N/A CUDA + RPC
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 18:53:15
[ROCm0] AMD Radeon Graphics (32732 MiB free)
[ROCm1] AMD Radeon VII (16340 MiB free)
2 Qwen3-VL-30B-A3B-Instruct-Q5_K_M.gguf TG = 63.70 PP = 716.64 TTFT = 196.32 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-31 18:15:19
[ROCm0] AMD Radeon Graphics (32732 MiB free) 1 Qwen3-VL-30B-A3B-Instruct-Q5_K_M.gguf TG = 74.18 PP = 735.59 TTFT = 191.47 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 20:28:45
[CUDA0] NVIDIA GeForce RTX 4070 (11049 MiB free)
[RPC0] 127.0.0.1:50052 (32732 MiB free)
2 Qwen3-VL-30B-A3B-Instruct-Q5_K_M.gguf TG = 59.39 PP = 820.84 TTFT = 279.30 N/A CUDA + RPC
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 19:12:22
[ROCm0] AMD Radeon Graphics (32732 MiB free)
[ROCm1] AMD Radeon VII (16340 MiB free)
2 Qwen3.5-35B-A3B-Q4_K_M.gguf TG = 54.89 PP = 714.86 TTFT = 654.72 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-31 18:13:30
[ROCm0] AMD Radeon Graphics (32732 MiB free) 1 Qwen3.5-35B-A3B-Q4_K_M.gguf TG = 61.48 PP = 730.12 TTFT = 640.03 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 20:27:56
[CUDA0] NVIDIA GeForce RTX 4070 (11046 MiB free)
[RPC0] 127.0.0.1:50052 (32586 MiB free)
2 Qwen3.5-35B-A3B-Q4_K_M.gguf TG = 50.66 PP = 833.31 TTFT = 560.42 N/A CUDA + RPC
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 19:09:09
[ROCm0] AMD Radeon Graphics (32732 MiB free)
[ROCm1] AMD Radeon VII (16340 MiB free)
2 Qwen3.5-35B-A3B-Q8_0.gguf TG = 59.42 PP = 631.67 TTFT = 739.64 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-31 18:14:21
[CUDA0] NVIDIA GeForce RTX 4070 (11030 MiB free)
[RPC0] 127.0.0.1:50052 (32586 MiB free)
2 Qwen3.5-35B-A3B-Q8_0.gguf TG = 49.72 PP = 726.29 TTFT = 643.84 N/A CUDA + RPC
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 19:10:58
[ROCm0] AMD Radeon Graphics (32732 MiB free)
[ROCm1] AMD Radeon VII (16340 MiB free)
2 Qwen3.5-35B-A3B-heretic-v2.i1-Q6_K.gguf TG = 54.38 PP = 681.80 TTFT = 687.01 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-31 18:12:31
[ROCm0] AMD Radeon Graphics (32732 MiB) [[max Watt:211W]]
[ROCm1] AMD Radeon VII (16340 MiB) [[max Watt:199W]]
2 Qwen3.5-35B-A3B-heretic-v2.i1-Q6_K.gguf TG = 32.57 PP = 504.08 TTFT = 56662.46 438.0 W ROCM
OS: Linux-6.17.0-20-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-14 12:49:59
[ROCm0] AMD Radeon Graphics (32732 MiB) [[max Watt:179W]]
[ROCm1] AMD Radeon VII (16340 MiB) [[max Watt:145W]]
2 Qwen3.5-35B-A3B-heretic-v2.i1-Q6_K.gguf TG = 51.94 PP = 636.24 TTFT = 778.00 342.0 W ROCM
OS: Linux-6.17.0-20-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-14 17:46:40
[ROCm0] AMD Radeon Graphics (32732 MiB free) 1 Qwen3.5-35B-A3B-heretic-v2.i1-Q6_K.gguf TG = 58.28 PP = 693.10 TTFT = 674.69 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 20:26:33
[CUDA0] NVIDIA GeForce RTX 4070 (11129 MiB free)
[RPC0] 127.0.0.1:50052 (32586 MiB free)
2 Qwen3.5-35B-A3B-heretic-v2.i1-Q6_K.gguf TG = 46.34 PP = 781.76 TTFT = 597.65 N/A CUDA + RPC
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 19:03:02
[ROCm0] AMD Radeon Graphics (32732 MiB) [[max Watt:211W]] 1 Qwen3.6-27B-UD-Q6_K_XL.gguf TG = 16.54 PP = 143.69 TTFT = 3448.35 504.0 W ROCM
OS: Linux-6.17.0-22-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-23 13:26:59
[ROCm0] AMD Radeon Graphics (32732 MiB) [[max Watt:245W]]
[ROCm1] AMD Radeon VII (16340 MiB) [[max Watt:270W]]
2 Qwen3.6-27B-UD-Q8_K_XL.gguf TG = 14.96 PP = 132.46 TTFT = 3744.43 584.0 W ROCM
OS: Linux-6.17.0-22-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-23 13:09:07
[ROCm0] AMD Radeon Graphics (32732 MiB) [[max Watt:177W]]
[ROCm1] AMD Radeon VII (16340 MiB) [[max Watt:156W]]
2 Qwen3.6-35B-A3B-Q8_0.gguf TG = 49.91 PP = 596.94 TTFT = 830.90 401.0 W ROCM
OS: Linux-6.17.0-22-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-23 13:06:38
[ROCm0] AMD Radeon Graphics (32732 MiB) [[max Watt:202W]] 1 Qwen3.6-35B-A3B-UD-Q6_K.gguf TG = 60.11 PP = 667.46 TTFT = 743.22 475.0 W ROCM
OS: Linux-6.17.0-22-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-23 13:24:55
[ROCm0] AMD Radeon Graphics (32732 MiB) [[max Watt:174W]] 1 Qwen3.6-35B-A3B-UD-Q6_K.gguf TG = 59.91 PP = 649.84 TTFT = 780.80 274.0 W ROCM
OS: Linux-6.17.0-20-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-17 16:20:29
[ROCm0] AMD Radeon Graphics (32732 MiB) [[max Watt:187W]]
[ROCm1] AMD Radeon VII (16332 MiB) [[max Watt:140W]]
2 Qwen3.6-35B-A3B-UD-Q8_K_XL.gguf TG = 50.14 PP = 490.98 TTFT = 1009.52 402.0 W ROCM
OS: Linux-6.17.0-22-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-23 13:04:37
No data 0 Ternary-Bonsai-8B-Q2_0.gguf TG = 4.38 PP = 5.07 TTFT = 6515.15 123.0 W No data
OS: Linux-6.17.0-22-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-26 00:20:33
[ROCm0] AMD Radeon VII (16340 MiB) [[max Watt:23W]] 1 Ternary-Bonsai-8B-Q2_0.gguf TG = 80.01 PP = 499.48 TTFT = 307.82 286.0 W ROCM
OS: Linux-6.17.0-22-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-26 00:28:34
[ROCm0] AMD Radeon Graphics (32732 MiB) [[max Watt:155W]] 1 gemma-4-26B-A4B-it-UD-Q4_K_M.gguf TG = 10.04 PP = 268.05 TTFT = 1764.79 250.0 W ROCM
OS: Linux-6.17.0-20-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-11 22:12:20
[ROCm0] AMD Radeon Graphics (32732 MiB) [[max Watt:173W]] 1 gemma-4-26B-A4B-it-UD-Q4_K_M.gguf TG = 74.49 PP = 837.44 TTFT = 564.34 268.0 W ROCM
OS: Linux-6.17.0-20-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-11 22:51:45
[ROCm0] AMD Radeon Graphics (32732 MiB) [[max Watt:156W]]
[ROCm1] AMD Radeon VII (16340 MiB) [[max Watt:149W]]
2 gemma-4-26B-A4B-it-uncensored-Q8_0.gguf TG = 62.51 PP = 724.50 TTFT = 651.49 327.3 W ROCM
OS: Linux-6.17.0-20-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-08 19:17:21
[ROCm0] AMD Radeon Graphics (32732 MiB free) 1 gemma-4-26B-A4B-it-uncensored-Q8_0.gguf TG = 75.79 PP = 736.27 TTFT = 640.54 N/A ROCM
OS: Linux-6.17.0-20-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-08 09:55:31
[ROCm0] AMD Radeon Graphics (32732 MiB free) [[max Watt:243W]] 1 gemma-4-26B-A4B-it-uncensored-Q8_0.gguf TG = 75.97 PP = 736.97 TTFT = 640.73 320.9 W ROCM
OS: Linux-6.17.0-20-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-08 18:25:39
[ROCm0] AMD Radeon VII (16340 MiB free)
[ROCm1] AMD Radeon Graphics (32732 MiB free)
2 gemma-4-31B-it-UD-Q8_K_XL.gguf TG = 17.41 PP = 120.35 TTFT = 4008.50 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-03 16:22:10
[ROCm0] AMD Radeon Graphics (32732 MiB) [[max Watt:171W]] 1 gemma-4-31B-it-UD-Q8_K_XL.gguf TG = 1.08 PP = 58.58 TTFT = 8049.66 266.0 W ROCM
OS: Linux-6.17.0-20-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-11 23:18:16
[ROCm0] AMD Radeon Graphics (32732 MiB) [[max Watt:188W]] 1 gemma-4-E4B-it-Q4_K_M.gguf TG = 73.22 PP = 1214.60 TTFT = 388.61 287.0 W ROCM
OS: Linux-6.17.0-20-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-11 21:54:06
[ROCm0] AMD Radeon Graphics (32732 MiB) [[max Watt:198W]] 1 gemma-4-E4B-it-Q4_K_M.gguf TG = 73.26 PP = 1217.48 TTFT = 387.52 293.0 W ROCM
OS: Linux-6.17.0-20-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-11 22:53:23
[ROCm0] AMD Radeon Graphics (32732 MiB) [[max Watt:169W]] 1 gemma-4-E4B-it-Q4_K_M.gguf TG = 12.15 PP = 665.72 TTFT = 710.11 264.0 W ROCM
OS: Linux-6.17.0-20-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-12 00:05:49
[ROCm0] AMD Radeon VII (16340 MiB free)
[ROCm1] AMD Radeon Graphics (32732 MiB free)
2 gemma-4-E4B-it-UD-Q8_K_XL.gguf TG = 56.87 PP = 760.57 TTFT = 633.78 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-03 16:18:23
[ROCm0] AMD Radeon VII (16332 MiB free)
[ROCm1] AMD Radeon Graphics (32732 MiB free)
2 google_gemma-4-26B-A4B-it-IQ4_NL.gguf TG = 64.97 PP = 791.07 TTFT = 611.32 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-03 17:20:46
[ROCm0] AMD Radeon Graphics (32732 MiB free) 1 google_gemma-4-26B-A4B-it-IQ4_NL.gguf TG = 75.79 PP = 832.76 TTFT = 565.59 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-03 22:22:40
[Vulkan0] AMD Radeon VII (RADV VEGA20) (15612 MiB free)
[Vulkan1] AMD Radeon Graphics (RADV VEGA20) (32751 MiB free)
2 google_gemma-4-26B-A4B-it-IQ4_NL.gguf TG = 37.92 PP = 493.11 TTFT = 956.44 N/A VULKAN
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-03 21:48:12
[Vulkan0] AMD Radeon Graphics (RADV VEGA20) (32751 MiB free) 1 google_gemma-4-26B-A4B-it-IQ4_NL.gguf TG = 74.64 PP = 474.21 TTFT = 993.24 N/A VULKAN
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-03 22:15:37
[ROCm0] AMD Radeon Graphics (32732 MiB) [[max Watt:128W]] 1 google_gemma-4-26B-A4B-it-Q8_0.gguf TG = 8.58 PP = 245.90 TTFT = 1915.95 223.0 W ROCM
OS: Linux-6.17.0-20-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-11 22:40:13
[ROCm0] AMD Radeon Graphics (32732 MiB) [[max Watt:179W]] 1 google_gemma-4-26B-A4B-it-Q8_0.gguf TG = 75.73 PP = 744.75 TTFT = 631.90 274.0 W ROCM
OS: Linux-6.17.0-20-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-11 22:59:35
[ROCm0] AMD Radeon Graphics (32732 MiB free) 1 google_gemma-4-26B-A4B-it-Q8_0.gguf TG = 75.57 PP = 741.64 TTFT = 635.62 N/A ROCM
OS: Linux-6.17.0-20-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-08 09:58:38
[ROCm0] AMD Radeon Graphics (32732 MiB) [[max Watt:87W]] 1 gpt-oss-20b-Q4_K_M.gguf TG = 13.99 PP = 364.90 TTFT = 540.81 182.0 W ROCM
OS: Linux-6.17.0-20-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-11 23:02:09
[ROCm0] AMD Radeon VII (16340 MiB) [[max Watt:24W]] 1 gpt-oss-20b-Q4_K_M.gguf TG = 82.34 PP = 862.71 TTFT = 249.70 275.0 W ROCM
OS: Linux-6.17.0-22-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-23 13:12:15
[ROCm0] AMD Radeon Graphics (32732 MiB) [[max Watt:242W]] 1 gpt-oss-20b-Q4_K_M.gguf TG = 59.92 PP = 642.60 TTFT = 4759.14 337.0 W ROCM
OS: Linux-6.17.0-20-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-14 20:25:08
[CUDA0] NVIDIA GeForce RTX 3050 Laptop GPU (3299 MiB free)
[RPC0] 192.168.0.222:50052 (27365 MiB free)
2 gpt-oss-20b-Q4_K_M.gguf TG = 28.10 PP = 64.49 TTFT = 1542.54 N/A CUDA + RPC
OS: Windows-10-10.0.19045-SP0
CUDA: 591.59
MMAP: false
2026-03-28 13:41:12
[ROCm0] AMD Radeon Graphics (32732 MiB free)
[ROCm1] AMD Radeon VII (16340 MiB free)
2 nvidia_Nemotron-3-Nano-30B-A3B-Q8_0.gguf TG = 84.38 PP = 656.95 TTFT = 711.84 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-31 18:06:02
[ROCm0] AMD Radeon Graphics (32732 MiB free) 1 nvidia_Nemotron-3-Nano-30B-A3B-Q8_0.gguf TG = 93.42 PP = 664.88 TTFT = 705.16 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 20:42:20
[CUDA0] NVIDIA GeForce RTX 4070 (11090 MiB free)
[RPC0] 127.0.0.1:50052 (32586 MiB free)
2 nvidia_Nemotron-3-Nano-30B-A3B-Q8_0.gguf TG = 73.28 PP = 789.68 TTFT = 593.66 N/A CUDA + RPC
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 18:55:25
[ROCm0] AMD Radeon Graphics (32732 MiB free)
[ROCm1] AMD Radeon VII (16340 MiB free)
2 nvidia_Nemotron-Cascade-2-30B-A3B-Q4_K_M.gguf TG = 84.52 PP = 723.37 TTFT = 648.90 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-31 18:08:27
[ROCm0] AMD Radeon Graphics (32520 MiB) [[max Watt:169W]] 1 nvidia_Nemotron-Cascade-2-30B-A3B-Q4_K_M.gguf TG = 11.87 PP = 234.64 TTFT = 2009.42 264.0 W ROCM
OS: Linux-6.17.0-20-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-11 22:44:19
[ROCm0] AMD Radeon Graphics (32732 MiB) [[max Watt:191W]] 1 nvidia_Nemotron-Cascade-2-30B-A3B-Q4_K_M.gguf TG = 12.02 PP = 232.52 TTFT = 2030.64 286.0 W ROCM
OS: Linux-6.17.0-20-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-04-11 23:04:41
[ROCm0] AMD Radeon Graphics (32732 MiB free) 1 nvidia_Nemotron-Cascade-2-30B-A3B-Q4_K_M.gguf TG = 96.39 PP = 743.60 TTFT = 630.18 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 20:19:24
[CUDA0] NVIDIA GeForce RTX 4070 (11046 MiB free)
[RPC0] 127.0.0.1:50052 (32586 MiB free)
2 nvidia_Nemotron-Cascade-2-30B-A3B-Q4_K_M.gguf TG = 75.20 PP = 869.88 TTFT = 538.94 N/A CUDA + RPC
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 19:22:25
[ROCm0] AMD Radeon Graphics (32732 MiB free)
[ROCm1] AMD Radeon VII (16340 MiB free)
2 nvidia_Nemotron-Cascade-2-30B-A3B-Q8_0.gguf TG = 85.21 PP = 648.49 TTFT = 722.34 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-31 18:09:11
[ROCm0] AMD Radeon Graphics (32732 MiB free) 1 nvidia_Nemotron-Cascade-2-30B-A3B-Q8_0.gguf TG = 91.76 PP = 661.61 TTFT = 709.25 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 20:41:21
[CUDA0] NVIDIA GeForce RTX 4070 (11000 MiB free)
[RPC0] 127.0.0.1:50052 (32586 MiB free)
2 nvidia_Nemotron-Cascade-2-30B-A3B-Q8_0.gguf TG = 72.76 PP = 782.02 TTFT = 599.79 N/A CUDA + RPC
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 19:20:35
[ROCm0] AMD Radeon Graphics (32732 MiB free)
[ROCm1] AMD Radeon VII (16340 MiB free)
2 omnicoder-9b-bf16.gguf TG = 25.30 PP = 200.43 TTFT = 2332.04 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-31 18:10:22
[ROCm0] AMD Radeon Graphics (32732 MiB free) 1 omnicoder-9b-bf16.gguf TG = 27.42 PP = 210.59 TTFT = 2216.90 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 20:24:52
[CUDA0] NVIDIA GeForce RTX 4070 (11122 MiB free)
[RPC0] 127.0.0.1:50052 (32586 MiB free)
2 omnicoder-9b-bf16.gguf TG = 25.54 PP = 256.80 TTFT = 1820.93 N/A CUDA + RPC
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 18:57:51
[CUDA0] NVIDIA GeForce RTX 4070 (11062 MiB free) 1 omnicoder-9b-q8_0.gguf TG = 49.22 PP = 3040.54 TTFT = 153.63 N/A CUDA
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
CUDA: 580.126.09 / ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-27 19:54:39
[ROCm0] AMD Radeon Graphics (32732 MiB free)
[ROCm1] AMD Radeon VII (16340 MiB free)
2 omnicoder-9b-q8_0.gguf TG = 50.62 PP = 417.33 TTFT = 1120.01 N/A ROCM
OS: Linux-6.17.0-19-generic-x86_64-with-glibc2.39
ROCM: 7.2.1.70201 / VULKAN: 1.3.275
MMAP: true
2026-03-31 18:11:14