ℹ

For context: Diese Plattform nutzt Unified Memory – die "VRAM" ist gemeinsamer System-RAM (APU/Superchip); das Modell teilt sich den Speicher mit dem System.

Generation31,25tok/s

Prefill938,92tok/s

Time to First Token14.944,00ms

Total duration52,32s

Concurrency5parallel

Ranking in the field

This GPU All GPUs

15of 16 systems

Performance benchmark · Primary metric: Generation-Speed (tok/s) · 5× concurrent

This run is better than 7 % of all comparable systems.

Generation 31,3 tok/s

-66 % vs Ø 91,4

Prefill 938,9 tok/s

-5 % vs Ø 984,2

Time to First Token 14.944 ms

-50 % vs Ø 30.097

Distribution in the field11 – 327 tok/s

Ø 91 Median Dieser Lauf

Wie schlägt sich dieser Benchmark mit anderen Modellen?

Nemotron-Cascade-2-30B-A3BAMD Radeon 8060S Graphics · run-20260820-184209-fe3575

326,9 tok/s

Qwen-AgentWorld-35B-A3BAMD Radeon 8060S Graphics · run-20260723-142109-7781ae

101,7 tok/s

Qwen3.6-35B-A3BAMD Radeon 8060S Graphics · run-20260820-184209-2e8259

98,3 tok/s

Qwen3-30B-A3B-Instruct-2507AMD Radeon 8060S Graphics · run-20260723-015518-6c8a89

97,8 tok/s

Qwen3.6-35B-A3BAMD Radeon 8060S Graphics · run-20260819-104341-2012e7

95,4 tok/s

Qwen3.6-35B-A3BAMD Radeon 8060S Graphics · run-20260820-184209-39ff2b

92,9 tok/s

Qwen3.6-35B-A3BAMD Radeon 8060S Graphics · run-20260819-104341-c511eb

86,4 tok/s

Qwen3.6-35B-A3BAMD Radeon 8060S Graphics · run-20260821-071254-6136b0

86,0 tok/s

Qwen3.6-35B-A3BAMD Radeon 8060S Graphics · run-20260819-104341-53e03b

81,9 tok/s

Gemma-4-26B-A4B-itAMD Radeon 8060S Graphics · run-20260725-031733-d3a49a

79,5 tok/s

Gemma-4-26B-A4B-itAMD Radeon 8060S Graphics · run-20260725-031733-6b9e97

77,7 tok/s

gpt-oss-20bAMD Radeon 8060S Graphics · run-20260724-082941-bd14ba

71,9 tok/s

gemma-4-26B-A4B-it-assistantAMD Radeon 8060S Graphics · run-20260725-031727-e0d1a8

67,3 tok/s

Ministral-3-14B-Reasoning-2512AMD Radeon 8060S Graphics · run-20260723-142058-33784d

56,8 tok/s

Mistral-Small-3.1-24B-Instruct-2503 this runAMD Radeon 8060S Graphics · run-20260723-142100-30c3b9

31,3 tok/s

How does this benchmark compare on other GPUs?

Same model on different hardware · 5× concurrent · Generation (tok/s)

AMD Radeon PRO W7900 Dual Slotrun-20260818-003841-9af4eb

119,7 tok/s

2× NVIDIA RTX A6000run-20260827-185904-fdd11a

110,7 tok/s

2× NVIDIA RTX A6000run-20260830-011927-6440e4

110,6 tok/s

NVIDIA RTX A6000run-20260826-065247-9822b4

94,6 tok/s

2× AMD Radeon PRO W7900 Dual Slotrun-20260818-115056-7e3a4d

89,5 tok/s

2× NVIDIA RTX A6000run-20260827-115206-4060a4

82,0 tok/s

2× NVIDIA RTX A6000run-20260829-163119-06f403

81,9 tok/s

2× AMD Radeon PRO W7800 48GBrun-20260902-105932-f21197

81,2 tok/s

2× AMD Radeon PRO W7900 Dual Slotrun-20260818-115056-aaf996

77,6 tok/s

NVIDIA RTX A6000run-20260826-121548-691de1

77,6 tok/s

AMD Radeon PRO W7900 Dual Slotrun-20260818-011053-d3ad58

77,6 tok/s

2× NVIDIA RTX A6000run-20260829-113415-4a5af1

51,4 tok/s

2× NVIDIA RTX A6000run-20260827-083601-8bffa1

51,2 tok/s

2× AMD Radeon PRO W7800 48GBrun-20260902-060236-f76384

47,2 tok/s

Hardware

GPU: AMD Radeon 8060S Graphics
CPU: AMD RYZEN AI MAX+ 395 w/ Radeon 8060S
RAM: 31 GB
Mainboard: Bosgame AXB35-02 (BeyondMax Series)

Setup

Runtime: llama.cpp
Quantization: Q4_K_M
Model: Mistral-Small-3.1-24B-Instruct-2503

Configuration

benchmark-konfiguration — run-20260723-142100-30c3b9

# LLM-Benchmark Konfiguration # Modell : Mistral-Small-3.1-24B-Instruct-2503 # Engine : llama.cpp # Run-ID : run-20260723-142100-30c3b9 # GPU : AMD Radeon 8060S Graphics # CPU : AMD RYZEN AI MAX+ 395 w/ Radeon 8060S # RAM : 31 GB bench@llm-benchmark:~$ /home/godcore/llama.cpp/build-rocm/bin/llama-server \ -m /home/godcore/models/dl/Mistral-Small-3.1-24B-Instruct-2503-Q4_K_M.gguf \ --host 0.0.0.0 \ --port 8000 \ --gpu-layers 99 \ --flash-attn on \ -c 26624 \ --parallel 12

Engine?Die Inferenz-Software, die das Modell ausliefert (z.B. vLLM oder llama.cpp). Sie bestimmt Geschwindigkeit, unterstuetzte Modellformate und welche Parameter ueberhaupt verfuegbar sind.	llamacpp
Kontextlaenge?Maximale Anzahl Tokens (Eingabe + erzeugte Ausgabe zusammen), die das Modell pro Anfrage verarbeiten kann.	26624
Modellpfad?Pfad zur GGUF-Modelldatei, die geladen und ausgeliefert wird.	/home/godcore/models/dl/Mistral-Small-3.1-24B-Instruct-2503-Q4_K_M.gguf
GPU-Layer?Anzahl der auf die GPU ausgelagerten Modell-Layer. Hoeher = mehr VRAM und schneller; der Rest laeuft auf der CPU. 999 = alles auf GPU.	99
Flash Attention?FlashAttention fuer schnellere und speichersparende Attention. Wert on/off/auto je nach Build.	on
Kontext?Groesse des Kontextfensters in Token. 0 = der beim Training verwendete Kontext des Modells.	26624
Parallel?Anzahl paralleler Slots/Sequenzen, die der Server gleichzeitig bedient. Der Kontext wird auf die Slots aufgeteilt.	12

All benchmarks of this model To leaderboard

Model comparison

Mistral-Small-3.1-24B-Instruct-2503 on various hardware

All published performance runs of this model – each bubble a variant: position = prefill (X) × generation (Y), bubble size = number of runs. Closer to the top right = faster. ★ Marked gold = this benchmark.

GPUby graphics card

NVIDIA RTX A6000 214,9 tok/sAMD Radeon PRO W7900 Dual Slot 150,4 tok/sAMD Radeon PRO W7800 48GB 141,9 tok/sNVIDIA RTX PRO 6000 Blackwell Workstation Edition 88,3 tok/sAMD Radeon AI PRO R9700 33,1 tok/s★ AMD Radeon 8060S Graphics 31,3 tok/s this runNVIDIA Tesla P100 PCIe 16GB 23,8 tok/s

CPUby processor

AMD Ryzen Threadripper PRO 7955WX 16-Cores 214,9 tok/sAMD Ryzen Threadripper PRO 5975WX 32-Cores 150,4 tok/sAMD Ryzen 9 9950X 16-Core Processor 88,3 tok/s★ AMD RYZEN AI MAX+ 395 w/ Radeon 8060S 31,3 tok/s this runAMD Ryzen 9 7945HX with Radeon Graphics 23,8 tok/s

MBby mainboard

ASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE 214,9 tok/sASUSTeK COMPUTER INC. Pro WS WRX80E-SAGE SE WIFI 150,4 tok/sASUSTeK COMPUTER INC. ProArt X870E-CREATOR WIFI 88,3 tok/s★ Bosgame AXB35-02 (BeyondMax Series) 31,3 tok/s this runShenzhen Meigao Electronic Equipment Co.,Ltd F1FXM (DeskMini Series) 23,8 tok/s

ENGby engine

★ llama.cpp 214,9 tok/s this runvLLM 99,3 tok/sunbekannt 33,1 tok/s

DRVby driver

unbekannt 214,9 tok/sAMD 7.0.0-27-generic 33,1 tok/s

💰 Economics

Economics of this run

Operating cost, TCO and comparison with the next-best runs of the same model at identical concurrency (5× concurrent). Methodology →

⚙️ ConfigurationAll metrics and charts below follow these settings – based on a 24-month runtime.Save to URL Reset

⚡ Electricity price EUR/kWh

⚙️ System utilization 100 %

🖥️ Acquisition EUR

🔌 Idle 18 W

⚡ TDP 120 W

☁️ External LLM (API) show

Input €/1M Output €/1M Input per output token

Electricity0.30 EUR/kWh

Avg power (incl. idle)120 W estimated (TDP)GPU 120 W full load

Avg cost / hourEUR 0.036

Electricity / 1M tokensEUR 0.32

Token / kWh937.50K

Acquisition (system)EUR 7,054 partial priceGPU EUR 3,000 · Board EUR 3,500 · RAM EUR 434 · PSU EUR 120

Electricity (2 years)–

TCO (2 years)EUR 7,685

Output tokens (2 years)1.97B

☁️ External LLM (API) – comparison

External LLM cost (2 years)–

Savings vs. external (2 years)–

All values above and the charts below take the configured system utilization into account: at X% the system generates only X% of the time, the rest it idles (18 W). Cost per hour drops (more idle), cost per token rises.

Cost over 2 years – electricity only

Cost over 2 years – incl. acquisition (TCO)

Speed vs. tokens per euro

Euro per 1M tokens

Comparison vs. API – economics per benchmark

	Mistral-Small-3.1-24B-Instruct-2503AMD Radeon 8060S Graphics	Mistral-Small-3.1-24B-Instruct-2503AMD Radeon PRO W7900 Dual Slot	Mistral-Small-3.1-24B-Instruct-25032x NVIDIA RTX A6000	Mistral-Small-3.1-24B-Instruct-25032x NVIDIA RTX A6000
Electricity cost (24 mo.)	–	–	–	–
Acquisition cost	–	–	–	–
Total cost (TCO)	–	–	–	–
Generated tokens (24 mo.)	–	–	–	–
Token price via API	–	–	–	–
Break-even point (days)	–	–	–	–
Result (savings / extra cost)	–	–	–	–

AMD Radeon 8060S Graphics · this run 31 tok/s AMD Radeon PRO W7900 Dual Slot 120 tok/s 2× NVIDIA RTX A6000 111 tok/s 2× NVIDIA RTX A6000 111 tok/s

Comparison with up to 3 next-best runs of this model at the same concurrency (at least one on different hardware). Power = GPU TDP + CPU (idle + 15 %) + board (estimated), acquisition = full system (GPU + CPU + board + RAM + PSU), prices = stored market prices.

Contributed by

Mario Alka Administrator

@marioalka

Ich bin Unternehmer, Softwareentwickler und KI-Enthusiast. Seit vielen Jahren entwickle ich Unternehmenssoftware und beschäftige mich inzwischen fast täglich mit lokalen LLMs, KI-Agenten und leistungsfähiger KI-Hardware.

Mit LLM-Benchmark.de möchte ich eine Plattform schaffen, auf der Modelle, GPUs und Agenten objektiv und reproduzierbar miteinander verglichen werden.