AMD-Pro-BenchCommunity

13,812 benchmarks in 32 GPU combinations documented

Qwen3.5-122B-A10B on AMD GPUs

46 results · 4 sources · 79 measurements

Explore all results

Grouped by reported model identity. Exact weights, quantizations and software can differ. How results are reviewed

Documented results

24 selected tests
View
ReportLocalMaxxing		Qwen3.5-122B-A10BMXFP4_MOEModel ref. ↗Token speed—
155.93 tok/sPrefill · median · unknown
15.03 tok/sGeneration · median · unknown
1 × Ryzen AI Max+ 395Single GPU · Unified memory
Configuration
Memory
Unified memory
OS
Not reported
Engine
llama.cpp
Backend
vulkan
Harness
LocalMaxxing community speed test; exact client where reported
Prompt / output
533 / 176 tokens
Context
709
Concurrency
Not reported
Date
2026-07-11
Full record →
View
ReportLocalMaxxing		Qwen3.5-122B-A10BMXFP4_MOEModel ref. ↗Token speed—
432.15 tok/sPrefill · unknown
26.41 tok/sGeneration · peak · unknown
3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Arch Linux
Engine
llama.cpp
Backend
rocm
Harness
LocalMaxxing community speed test; exact client where reported
Prompt / output
531 / 256 tokens
Context
787
Concurrency
Not reported
Date
2026-06-27
Full record →
View
ReportGeneric LLM		Qwen3.5-122B-A10BGPTQ Int4Model ref. ↗Token speed
39.9 tok/sThroughput · per request
——4 × R9700Tensor parallel · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Linux (distribution not reported)
Engine
vLLM
Backend
ROCm 7.0.51831 container / 6.4.3 host
Harness
Not reported
Prompt / output
? / ? tokens
Context
Not reported
Concurrency
1
Date
2026-05-22
Full record →
View
Reportllama-bench		Qwen3.5-122B-A10BQ5_K_XLModel ref. ↗Token speed512 input tokens · 0 output tokens · 0 existing KV tokens—
336.69 tok/sPrefill · mean · per request
—1 × Ryzen AI Max+ 395Single GPU · Unified memory
Configuration
Memory
Unified memory
OS
Fedora Linux 43 (Workstation Edition)
Engine
llama.cpp
Backend
ROCm / rocm-7_2_3
Harness
llama-bench
Prompt / output
512 / 0 tokens
Context
Not reported
Concurrency
1
Date
2026-05-16
Full record →
View
ReportLocalMaxxing		Qwen3.5-122B-A10BQ4_K_XLModel ref. ↗Token speed——
20.9 tok/sGeneration · unknown
1 × Ryzen AI Max+ 395Single GPU · Unified memory
Configuration
Memory
Unified memory
OS
Linux
Engine
llama.cpp
Backend
rocm
Harness
LocalMaxxing community speed test; exact client where reported
Prompt / output
512 / 128 tokens
Context
262144
Concurrency
Not reported
Date
2026-05-17
Full record →
View
ReportLocalMaxxing		Qwen3.5-122B-A10BMXFP4_MOEModel ref. ↗Token speed—
527.68 tok/sPrefill · unknown
26.62 tok/sGeneration · peak · unknown
3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Arch Linux
Engine
llama.cpp
Backend
vulkan
Harness
LocalMaxxing community speed test; exact client where reported
Prompt / output
533 / 158 tokens
Context
691
Concurrency
Not reported
Date
2026-08-04
Full record →
View
ReportGeneric LLM		Qwen3.5-122B-A10BGPTQ Int4Model ref. ↗Token speed——
53.8 tok/sGeneration · per request
4 × R9700Tensor parallel · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Linux (distribution not reported)
Engine
vLLM
Backend
ROCm 7.0.51831 container / 6.4.3 host
Harness
Not reported
Prompt / output
? / ? tokens
Context
Not reported
Concurrency
1
Date
2026-05-22
Full record →
View
Reportllama-bench		Qwen3.5-122B-A10BQ5_K_XLModel ref. ↗Token speed0 input tokens · 128 output tokens · 0 existing KV tokens——
20.12 tok/sGeneration · mean · per request
1 × Ryzen AI Max+ 395Single GPU · Unified memory
Configuration
Memory
Unified memory
OS
Fedora Linux 43 (Workstation Edition)
Engine
llama.cpp
Backend
ROCm / rocm-7_2_3
Harness
llama-bench
Prompt / output
0 / 128 tokens
Context
Not reported
Concurrency
1
Date
2026-05-16
Full record →
View
ReportLocalMaxxing		Qwen3.5-122B-A10BUD-IQ4_NLModel ref. ↗Token speed——
19.02 tok/sGeneration · unknown
1 × Ryzen AI Max+ 395Single GPU · Unified memory
Configuration
Memory
Unified memory
OS
Linux
Engine
llama.cpp
Backend
rocm
Harness
LocalMaxxing community speed test; exact client where reported
Prompt / output
3663 / 2048 tokens
Context
262144
Concurrency
Not reported
Date
2026-04-29
Full record →
View
ReportLocalMaxxing		Qwen3.5-122B-A10BMXFP4_MOEModel ref. ↗Token speed—
312.56 tok/sPrefill · unknown
21 tok/sGeneration · peak · unknown
3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Arch Linux
Engine
llama.cpp
Backend
vulkan
Harness
LocalMaxxing community speed test; exact client where reported
Prompt / output
533 / 156 tokens
Context
689
Concurrency
Not reported
Date
2026-06-27
Full record →
View
ReportGeneric LLM		Qwen3.5-122B-A10BGPTQ Int4Model ref. ↗Latency1 concurrent request
1,810 msTime to first token · Per request
——4 × R9700Tensor parallel · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Linux (distribution not reported)
Engine
vLLM
Backend
ROCm 7.0.51831 container / 6.4.3 host
Harness
Not reported
Prompt / output
? / ? tokens
Context
Not reported
Concurrency
1
Date
2026-05-22
Full record →
View
Reportllama-bench		Qwen3.5-122B-A10BQ5_K_XLModel ref. ↗Token speed2,048 input tokens · 0 output tokens · 32,768 existing KV tokens—
276.15 tok/sPrefill · mean · per request
—1 × Ryzen AI Max+ 395Single GPU · Unified memory
Configuration
Memory
Unified memory
OS
Fedora Linux 43 (Workstation Edition)
Engine
llama.cpp
Backend
ROCm / rocm-7_2_3
Harness
llama-bench
Prompt / output
2048 / 0 tokens
Context
Not reported
Concurrency
1
Date
2026-05-16
Full record →
View
ReportLocalMaxxing		Qwen3.5-122B-A10BUD-IQ4_NLModel ref. ↗Token speed——
21.25 tok/sGeneration · unknown
1 × Ryzen AI Max+ 395Single GPU · Unified memory
Configuration
Memory
Unified memory
OS
Linux
Engine
llama.cpp
Backend
vulkan
Harness
LocalMaxxing community speed test; exact client where reported
Prompt / output
3663 / 2048 tokens
Context
262144
Concurrency
Not reported
Date
2026-04-29
Full record →
View
ReportLocalMaxxing		Qwen3.5-122B-A10BMXFP4_MOEModel ref. ↗Token speed—
390.99 tok/sPrefill · unknown
21.33 tok/sGeneration · peak · unknown
3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Arch Linux
Engine
llama.cpp
Backend
vulkan
Harness
LocalMaxxing community speed test; exact client where reported
Prompt / output
533 / 158 tokens
Context
691
Concurrency
Not reported
Date
2026-06-27
Full record →
View
Reportllama-bench		Qwen3.5-122B-A10BQ5_K_XLModel ref. ↗Token speed0 input tokens · 32 output tokens · 32,768 existing KV tokens——
17.73 tok/sGeneration · mean · per request
1 × Ryzen AI Max+ 395Single GPU · Unified memory
Configuration
Memory
Unified memory
OS
Fedora Linux 43 (Workstation Edition)
Engine
llama.cpp
Backend
ROCm / rocm-7_2_3
Harness
llama-bench
Prompt / output
0 / 32 tokens
Context
Not reported
Concurrency
1
Date
2026-05-16
Full record →
View
ReportLocalMaxxing		Qwen3.5-122B-A10BMXFP4_MOEModel ref. ↗Token speed——
26.85 tok/sGeneration · peak · unknown
3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Arch Linux
Engine
llama.cpp
Backend
vulkan
Harness
LocalMaxxing community speed test; exact client where reported
Prompt / output
531 / 256 tokens
Context
65536
Concurrency
Not reported
Date
2026-05-04
Full record →
View
Reportllama-bench		Qwen3.5-122B-A10BQ5_K_XLModel ref. ↗Token speed2,048 input tokens · 0 output tokens · 65,536 existing KV tokens—
192.28 tok/sPrefill · mean · per request
—1 × Ryzen AI Max+ 395Single GPU · Unified memory
Configuration
Memory
Unified memory
OS
Fedora Linux 43 (Workstation Edition)
Engine
llama.cpp
Backend
ROCm / rocm-7_2_3
Harness
llama-bench
Prompt / output
2048 / 0 tokens
Context
Not reported
Concurrency
1
Date
2026-05-16
Full record →
View
ReportLocalMaxxing		Qwen3.5-122B-A10BMXFP4_MOEModel ref. ↗Token speed——
27.31 tok/sGeneration · peak · unknown
3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Arch Linux
Engine
llama.cpp
Backend
vulkan
Harness
LocalMaxxing community speed test; exact client where reported
Prompt / output
533 / 161 tokens
Context
694
Concurrency
Not reported
Date
2026-05-04
Full record →
View
Reportllama-bench		Qwen3.5-122B-A10BQ5_K_XLModel ref. ↗Token speed0 input tokens · 32 output tokens · 65,536 existing KV tokens——
15.74 tok/sGeneration · mean · per request
1 × Ryzen AI Max+ 395Single GPU · Unified memory
Configuration
Memory
Unified memory
OS
Fedora Linux 43 (Workstation Edition)
Engine
llama.cpp
Backend
ROCm / rocm-7_2_3
Harness
llama-bench
Prompt / output
0 / 32 tokens
Context
Not reported
Concurrency
1
Date
2026-05-16
Full record →
View
ReportLocalMaxxing		Qwen3.5-122B-A10BMXFP4_MOEModel ref. ↗Token speed——
27.19 tok/sGeneration · peak · unknown
3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Arch Linux
Engine
llama.cpp
Backend
vulkan
Harness
LocalMaxxing community speed test; exact client where reported
Prompt / output
533 / 161 tokens
Context
65536
Concurrency
Not reported
Date
2026-05-05
Full record →
View
Reportllama-bench		Qwen3.5-122B-A10BQ5_K_XLModel ref. ↗Token speed512 input tokens · 0 output tokens · 0 existing KV tokens—
336.02 tok/sPrefill · mean · per request
—1 × Ryzen AI Max+ 395Single GPU · Unified memory
Configuration
Memory
Unified memory
OS
Fedora Linux 43 (Workstation Edition)
Engine
llama.cpp
Backend
ROCm / rocm6_4_4
Harness
llama-bench
Prompt / output
512 / 0 tokens
Context
Not reported
Concurrency
1
Date
2026-05-16
Full record →
View
ReportLocalMaxxing		Qwen3.5-122B-A10BMXFP4_MOEModel ref. ↗Token speed——
26.81 tok/sGeneration · peak · unknown
3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Arch Linux
Engine
llama.cpp
Backend
vulkan
Harness
LocalMaxxing community speed test; exact client where reported
Prompt / output
533 / 161 tokens
Context
694
Concurrency
Not reported
Date
2026-05-05
Full record →
View
Reportllama-bench		Qwen3.5-122B-A10BQ5_K_XLModel ref. ↗Token speed0 input tokens · 128 output tokens · 0 existing KV tokens——
19.73 tok/sGeneration · mean · per request
1 × Ryzen AI Max+ 395Single GPU · Unified memory
Configuration
Memory
Unified memory
OS
Fedora Linux 43 (Workstation Edition)
Engine
llama.cpp
Backend
ROCm / rocm6_4_4
Harness
llama-bench
Prompt / output
0 / 128 tokens
Context
Not reported
Concurrency
1
Date
2026-05-16
Full record →
View
ReportLocalMaxxing		Qwen3.5-122B-A10BF16Model ref. ↗Token speed——
25.43 tok/sGeneration · peak · unknown
3 × R9700Multi-GPU · split unknown · 32 GB VRAM/GPU
Configuration
Memory
32 GB VRAM/GPU
OS
Arch Linux
Engine
llama.cpp
Backend
vulkan
Harness
LocalMaxxing community speed test; exact client where reported
Prompt / output
541 / 256 tokens
Context
797
Concurrency
Not reported
Date
2026-05-11
Full record →
Quantizations and software

Quantizations: F16, GPTQ Int4, MXFP4_MOE, Q4_K_XL, Q5_K_XL, UD-IQ4_NL
Engines: llama.cpp, vLLM

Explore all 46 results →

This is a selection across sources, not a ranking or a matched experiment. Reading benchmark reports

GPU benchmark pages: R9700 · Ryzen AI Max+ 395