Created bymario-alka.dePowered bygodcore.denoob2claw.detricoma.de
Contributed byMario AlkaQwen (Alibaba)

Qwen3-30B-A3B-Instruct-2507

Performance benchmark · measured on 18.08.2026 13:46

Benchmark-IDrun-20260818-141454-3bf049
Timebench 3 - Kombi (Prefill + Generation)MoE30BRuntime: llama.cppQuantisierung: Q8_0
Generation88,04tok/s
Prefill2.446,75tok/s
Time to First Token1.039,00ms
Total duration25,34s
Concurrency1parallel
Ranking in the field
17of 81 systems

Performance benchmark · Primary metric: Generation-Speed (tok/s) · 1× concurrent

This run is better than 80 % of all comparable systems.
Generation 88,0 tok/s
+65 % vs Ø 53,3
Prefill 2.446,8 tok/s
+77 % vs Ø 1.382,6
Time to First Token 1.039 ms
-75 % vs Ø 4.232
Distribution in the field10 – 155 tok/s
Ø 53 Median Dieser Lauf

Wie schlägt sich dieser Benchmark mit anderen Modellen?

How does this benchmark compare on other GPUs?

Same model on different hardware · 1× concurrent · Generation (tok/s)

Hardware

GPU: 2x AMD Radeon PRO W7900 Dual Slot · 48 GB VRAM
CPU: AMD Ryzen Threadripper PRO 5975WX 32-Cores
RAM: 247 GB
Mainboard: ASUSTeK COMPUTER INC. Pro WS WRX80E-SAGE SE WIFI

Setup

Runtime: llama.cpp
Quantization: Q8_0
Model: Qwen3-30B-A3B-Instruct-2507

Configuration

benchmark-konfiguration — run-20260818-141454-3bf049
# LLM-Benchmark Konfiguration # Modell : Qwen3-30B-A3B-Instruct-2507 # Engine : llama.cpp # Run-ID : run-20260818-141454-3bf049 # GPU : 2x AMD Radeon PRO W7900 Dual Slot # CPU : AMD Ryzen Threadripper PRO 5975WX 32-Cores # RAM : 247 GB bench@llm-benchmark:~$ /home/godcore/llama.cpp/build_hip/bin/llama-server \ -m /home/godcore/bench_scratch/Qwen_Qwen3-30B-A3B-Instruct-2507-Q8_0.gguf \ --alias Qwen3-30B-A3B-Instruct-2507 \ -ngl 999 \ -fa on \ -sm layer \ --tensor-split 1,1 \ -c 49152 \ -np 12 \ --jinja \ --host 0.0.0.0 \ --port 8000
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
Modellalias?Der Name, unter dem das Modell ueber die API angesprochen wird. Genau dieser Wert muss im Request-Feld 'model' stehen.Qwen3-30B-A3B-Instruct-2507
Kontextlaenge?Maximale Anzahl Tokens (Eingabe + erzeugte Ausgabe zusammen), die das Modell pro Anfrage verarbeiten kann.49152
Modellpfad?Pfad zur GGUF-Modelldatei, die geladen und ausgeliefert wird./home/godcore/bench_scratch/Qwen_Qwen3-30B-A3B-Instruct-2507-Q8_0.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.999
faon
Split-Mode?Verteilung ueber mehrere GPUs: none (nur eine GPU), layer (Layer aufteilen) oder row (Tensoren zeilenweise).layer
Tensor-Split?Verhaeltnis, in dem die Modell-Layer auf mehrere GPUs verteilt werden, z.B. 3,1.1,1
Kontext?Groesse des Kontextfensters in Token. 0 = der beim Training verwendete Kontext des Modells.49152
np12

All benchmarks of this model To leaderboard

Anzeige
Model comparison

Qwen3-30B-A3B-Instruct-2507 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

2.1711.6281.0855430,006.26012.52018.780Prefill (tok/s)Generation (tok/s)NVIDIA GeForce RTX 5090 - 1.671,0 tok/s Generation, 15.160 tok/s Prefill, TTFT 2.212 ms (6 Laufe)NVIDIA GeForce RTX 50...NVIDIA RTX PRO 6000 Blackwell Workstation Edition - 1.658,1 tok/s Generation, 15.099 tok/s Prefill, TTFT 2.030 ms (6 Laufe)NVIDIA RTX PRO 6000 B...NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition - 1.515,7 tok/s Generation, 13.445 tok/s Prefill, TTFT 3.300 ms (9 Laufe)NVIDIA RTX PRO 6000 B...NVIDIA GeForce RTX 3090 Ti - 905,1 tok/s Generation, 8.498 tok/s Prefill, TTFT 5.639 ms (6 Laufe)NVIDIA GeForce RTX 30...NVIDIA GB10 (DGX Spark) - 169,9 tok/s Generation, 5.183 tok/s Prefill, TTFT 2.923 ms (3 Laufe)NVIDIA GB10 (DGX Spar...NVIDIA GeForce RTX 5070 Ti - 152,7 tok/s Generation, 1.209 tok/s Prefill, TTFT 22.640 ms (6 Laufe)NVIDIA GeForce RTX 50...Intel Arc Pro B70 - 143,6 tok/s Generation, 1.123 tok/s Prefill, TTFT 2.220 ms (3 Laufe)Intel Arc Pro B70AMD Radeon 8060S Graphics - 107,6 tok/s Generation, 2.228 tok/s Prefill, TTFT 8.963 ms (2 Laufe)AMD Radeon 8060S Grap...AMD Radeon AI PRO R9700 - 97,6 tok/s Generation, 2.543 tok/s Prefill, TTFT 36.082 ms (26 Laufe)AMD Radeon AI PRO R97...CPU-only - 7,1 tok/s Generation, 102 tok/s Prefill, TTFT 109.338 ms (3 Laufe)CPU-onlyAMD Radeon PRO W7900 Dual Slot - 264,2 tok/s Generation, 4.310 tok/s Prefill, TTFT 3.293 ms (9 Laufe) | DIESER LAUF★ AMD Radeon PRO W7900 ...
NVIDIA GeForce RTX 5090 1.671,0 tok/sNVIDIA RTX PRO 6000 Blackwell Workstation Edition 1.658,1 tok/sNVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition 1.515,7 tok/sNVIDIA GeForce RTX 3090 Ti 905,1 tok/s★ AMD Radeon PRO W7900 Dual Slot 264,2 tok/s this runNVIDIA GB10 (DGX Spark) 169,9 tok/sNVIDIA GeForce RTX 5070 Ti 152,7 tok/sIntel Arc Pro B70 143,6 tok/sAMD Radeon 8060S Graphics 107,6 tok/sAMD Radeon AI PRO R9700 97,6 tok/sCPU-only 7,1 tok/s

CPUby processor

2.1711.6281.0855430,006.26012.52018.780Prefill (tok/s)Generation (tok/s)AMD Ryzen 7 5800X3D 8-Core Processor - 1.671,0 tok/s Generation, 15.160 tok/s Prefill, TTFT 2.212 ms (6 Laufe)AMD Ryzen 7 5800X3D 8...AMD Ryzen 9 9950X 16-Core Processor - 1.658,1 tok/s Generation, 15.099 tok/s Prefill, TTFT 2.030 ms (6 Laufe)AMD Ryzen 9 9950X 16-...AMD Ryzen Threadripper PRO 9965WX 24-Cores - 1.515,7 tok/s Generation, 13.445 tok/s Prefill, TTFT 3.300 ms (9 Laufe)AMD Ryzen Threadrippe...AMD Ryzen 9 8945HX with Radeon Graphics - 905,1 tok/s Generation, 8.498 tok/s Prefill, TTFT 5.639 ms (6 Laufe)AMD Ryzen 9 8945HX wi...NVIDIA Grace - 169,9 tok/s Generation, 5.183 tok/s Prefill, TTFT 2.923 ms (3 Laufe)NVIDIA GraceAMD Ryzen 9 7945HX with Radeon Graphics - 143,6 tok/s Generation, 1.123 tok/s Prefill, TTFT 2.220 ms (3 Laufe)AMD Ryzen 9 7945HX wi...AMD RYZEN AI MAX+ 395 w/ Radeon 8060S - 107,6 tok/s Generation, 2.228 tok/s Prefill, TTFT 8.963 ms (2 Laufe)AMD RYZEN AI MAX+ 395...AMD Ryzen Threadripper PRO 7955WX 16-Cores - 97,6 tok/s Generation, 2.543 tok/s Prefill, TTFT 36.082 ms (26 Laufe)AMD Ryzen Threadrippe...Intel(R) Xeon(R) CPU E5-4620 v2 @ 2.60GHz - 7,1 tok/s Generation, 102 tok/s Prefill, TTFT 109.338 ms (3 Laufe)Intel(R) Xeon(R) CPU ...AMD Ryzen Threadripper PRO 5975WX 32-Cores - 264,2 tok/s Generation, 3.069 tok/s Prefill, TTFT 11.032 ms (15 Laufe) | DIESER LAUF★ AMD Ryzen Threadrippe...
AMD Ryzen 7 5800X3D 8-Core Processor 1.671,0 tok/sAMD Ryzen 9 9950X 16-Core Processor 1.658,1 tok/sAMD Ryzen Threadripper PRO 9965WX 24-Cores 1.515,7 tok/sAMD Ryzen 9 8945HX with Radeon Graphics 905,1 tok/s★ AMD Ryzen Threadripper PRO 5975WX 32-Cores 264,2 tok/s this runNVIDIA Grace 169,9 tok/sAMD Ryzen 9 7945HX with Radeon Graphics 143,6 tok/sAMD RYZEN AI MAX+ 395 w/ Radeon 8060S 107,6 tok/sAMD Ryzen Threadripper PRO 7955WX 16-Cores 97,6 tok/sIntel(R) Xeon(R) CPU E5-4620 v2 @ 2.60GHz 7,1 tok/s

MBby mainboard

2.1711.6281.0855430,006.26012.52018.780Prefill (tok/s)Generation (tok/s)ASUSTeK COMPUTER INC. ROG STRIX B550-A GAMING - 1.671,0 tok/s Generation, 15.160 tok/s Prefill, TTFT 2.212 ms (6 Laufe)ASUSTeK COMPUTER INC....ASUSTeK COMPUTER INC. ProArt X870E-CREATOR WIFI - 1.658,1 tok/s Generation, 15.099 tok/s Prefill, TTFT 2.030 ms (6 Laufe)ASUSTeK COMPUTER INC....ASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE - 1.515,7 tok/s Generation, 5.347 tok/s Prefill, TTFT 27.652 ms (35 Laufe)ASUSTeK COMPUTER INC....Meigao Innovation Technology (Shen Zhen) Co., Ltd DRFXL (MotherBoard Series) - 905,1 tok/s Generation, 8.498 tok/s Prefill, TTFT 5.639 ms (6 Laufe)Meigao Innovation Tec...ASUSTeK COMPUTER INC. GX10 - 169,9 tok/s Generation, 5.183 tok/s Prefill, TTFT 2.923 ms (3 Laufe)ASUSTeK COMPUTER INC....Shenzhen Meigao Electronic Equipment Co.,Ltd DRFXI (MotherBoard Series) - 143,6 tok/s Generation, 1.123 tok/s Prefill, TTFT 2.220 ms (3 Laufe)Shenzhen Meigao Elect...Bosgame AXB35-02 (BeyondMax Series) - 107,6 tok/s Generation, 2.228 tok/s Prefill, TTFT 8.963 ms (2 Laufe)Bosgame AXB35-02 (Bey...Dell Inc. PowerEdge R820 - 7,1 tok/s Generation, 102 tok/s Prefill, TTFT 109.338 ms (3 Laufe)Dell Inc. PowerEdge R...ASUSTeK COMPUTER INC. Pro WS WRX80E-SAGE SE WIFI - 264,2 tok/s Generation, 3.069 tok/s Prefill, TTFT 11.032 ms (15 Laufe) | DIESER LAUF★ ASUSTeK COMPUTER INC....
ASUSTeK COMPUTER INC. ROG STRIX B550-A GAMING 1.671,0 tok/sASUSTeK COMPUTER INC. ProArt X870E-CREATOR WIFI 1.658,1 tok/sASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE 1.515,7 tok/sMeigao Innovation Technology (Shen Zhen) Co., Ltd DRFXL (MotherBoard Series) 905,1 tok/s★ ASUSTeK COMPUTER INC. Pro WS WRX80E-SAGE SE WIFI 264,2 tok/s this runASUSTeK COMPUTER INC. GX10 169,9 tok/sShenzhen Meigao Electronic Equipment Co.,Ltd DRFXI (MotherBoard Series) 143,6 tok/sBosgame AXB35-02 (BeyondMax Series) 107,6 tok/sDell Inc. PowerEdge R820 7,1 tok/s

ENGby engine

2.1531.6151.0765380,01.8015.2038.60612.008Prefill (tok/s)Generation (tok/s)vLLM - 961,5 tok/s Generation, 10.207 tok/s Prefill, TTFT 3.750 ms (17 Laufe)vLLMunbekannt - 97,6 tok/s Generation, 3.602 tok/s Prefill, TTFT 1.021 ms (9 Laufe)unbekanntllama.cpp - 1.671,0 tok/s Generation, 5.348 tok/s Prefill, TTFT 27.944 ms (53 Laufe) | DIESER LAUF★ llama.cpp
★ llama.cpp 1.671,0 tok/s this runvLLM 961,5 tok/sunbekannt 97,6 tok/s

DRVby driver

2.1531.6151.0765380,002.7625.5248.285Prefill (tok/s)Generation (tok/s)unbekannt - 1.671,0 tok/s Generation, 6.845 tok/s Prefill, TTFT 23.896 ms (64 Laufe)unbekanntNVIDIA 590.48.01 / CUDA 13.1 - 169,9 tok/s Generation, 5.183 tok/s Prefill, TTFT 2.923 ms (3 Laufe)NVIDIA 590.48.01 / CU...Intel 26.18.38308.4 - 143,6 tok/s Generation, 1.123 tok/s Prefill, TTFT 2.220 ms (3 Laufe)Intel 26.18.38308.4AMD 7.0.0-27-generic - 97,6 tok/s Generation, 3.602 tok/s Prefill, TTFT 1.021 ms (9 Laufe)AMD 7.0.0-27-generic
unbekannt 1.671,0 tok/sNVIDIA 590.48.01 / CUDA 13.1 169,9 tok/sIntel 26.18.38308.4 143,6 tok/sAMD 7.0.0-27-generic 97,6 tok/s
💰 Economics

Economics of this run

Operating cost, TCO and comparison with the next-best runs of the same model at identical concurrency (1× concurrent). Methodology →

⚙️ ConfigurationAll metrics and charts below follow these settings – based on a 24-month runtime.Save to URLReset
⚡ Electricity price EUR/kWh
⚙️ System utilization 100 %
🖥️ Acquisition EUR
🔌 Idle 10 W
⚡ TDP 600 W
☁️ External LLM (API)
Electricity0.30 EUR/kWh
Avg power (incl. idle)600 W estimated (TDP)GPU 590 + Board 10 W full load
Avg cost / hourEUR 0.18
Electricity / 1M tokensEUR 0.57
Token / kWh528.24K
Acquisition (system)EUR 2,156 missingRAM EUR 1,976 · PSU EUR 180
Electricity (2 years)
TCO (2 years)EUR 5,310
Output tokens (2 years)5.55B
☁️ 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 (10 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

Qwen3-30B-A3B-Instruct-25072x AMD Radeon PRO W7900 Dual SlotQwen3-30B-A3B-Instruct-2507NVIDIA GeForce RTX 5090Qwen3-30B-A3B-Instruct-2507NVIDIA RTX PRO 6000 Blackwell Workstation EditionQwen3-30B-A3B-Instruct-25073x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition
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)

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.