⚠

Configuration note: Es sind 2 Grafikkarten installiert, aber nur 1 wurde fuer diesen Lauf genutzt. Mit allen 2 GPUs koennte der Durchsatz deutlich hoeher liegen.

Generation129,35tok/s

Prefill4.876,19tok/s

Time to First Token464,50ms

Total duration16,76s

Concurrency1parallel

Ranking in the field

This GPU All GPUs

246of 1264 systems

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

This run is better than 81 % of all comparable systems.

Generation 129,4 tok/s

+71 % vs Ø 75,6

Prefill 4.876,2 tok/s

+95 % vs Ø 2.498,4

Time to First Token 465 ms

-99 % vs Ø 32.833

Distribution in the field0 – 405 tok/s

Ø 76 Median Dieser Lauf

Wie schlägt sich dieser Benchmark mit anderen Modellen?

gpt-oss-20bNVIDIA GeForce RTX 5090 · run-20260724-004608-16af9b

404,6 tok/s

gpt-oss-20bNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260730-035052-9f766a

395,2 tok/s

Nemotron-3-Nano-4BNVIDIA GeForce RTX 5090 · run-20260728-194135-d456e7

393,5 tok/s

Nemotron-3-Nano-4BNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260727-181607-bffb11

393,5 tok/s

gpt-oss-20bNVIDIA GeForce RTX 5090 · run-20260724-004607-29bd16

388,9 tok/s

gpt-oss-20bNVIDIA GeForce RTX 5090 · run-20260729-032121-1c779f

388,2 tok/s

gemma-4-E2B-itNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260727-181606-6ea089

378,6 tok/s

gemma-4-E2B-itNVIDIA GeForce RTX 5090 · run-20260728-184455-e8b129

377,6 tok/s

Nemotron-3-Nano-4B3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260728-194136-ac388e

362,3 tok/s

Nemotron-3-Nano-4B3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260728-194136-5e5085

361,4 tok/s

Nemotron-3-Nano-30B-A3BNVIDIA GeForce RTX 5090 · run-20260729-105335-653788

359,4 tok/s

Nemotron-3-Nano-Omni-30B-A3B-Reasoning same modelNVIDIA GeForce RTX 5090 · run-20260729-105334-2c4905

359,3 tok/s

Nemotron-Cascade-2-30B-A3BNVIDIA GeForce RTX 5090 · run-20260729-105334-44d1e7

359,0 tok/s

Nemotron-3-Nano-Omni-30B-A3B-Reasoning same modelNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260729-032119-4852bf

357,1 tok/s

Nemotron-3-Nano-Omni-30B-A3B-Reasoning this run2× NVIDIA RTX A6000 · run-20260827-024202-60bd13

129,4 tok/s

How does this benchmark compare on other GPUs?

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

NVIDIA GeForce RTX 5090run-20260729-105334-2c4905

359,3 tok/s

NVIDIA RTX PRO 6000 Blackwell Workstation Editionrun-20260729-032119-4852bf

357,1 tok/s

3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20260729-105339-e213a2

331,0 tok/s

3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20260729-105339-133b6a

330,1 tok/s

3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20260729-105340-0a8541

326,8 tok/s

NVIDIA GeForce RTX 3090 Tirun-20260729-035835-89abde

221,8 tok/s

NVIDIA RTX A6000run-20260826-131708-823486

146,2 tok/s

NVIDIA RTX A6000run-20260826-131708-6158f8

132,4 tok/s

2× NVIDIA RTX A6000 this runrun-20260827-024202-60bd13

129,4 tok/s

NVIDIA RTX A6000run-20260826-225451-804f23

120,9 tok/s

3× AMD Radeon AI PRO R9700run-20260722-182228-95c528

102,9 tok/s

NVIDIA GeForce RTX 5070 Tirun-20260729-032125-286e08

68,9 tok/s

3× AMD Radeon AI PRO R9700run-20260806-094441-5e76b5

32,1 tok/s

3× AMD Radeon AI PRO R9700run-20260806-094441-50cc3e

31,7 tok/s

Hardware

GPU: 2x NVIDIA RTX A6000 · 48 GB VRAM · only 1 of 2 used
CPU: AMD Ryzen Threadripper PRO 7955WX 16-Cores
RAM: 184 GB
Mainboard: ASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE

Setup

Runtime: vLLM
Quantization: FP8
Model: Nemotron-3-Nano-Omni-30B-A3B-Reasoning

Configuration

benchmark-konfiguration — run-20260827-024202-60bd13

# LLM-Benchmark Konfiguration # Modell : Nemotron-3-Nano-Omni-30B-A3B-Reasoning # Engine : vLLM # Run-ID : run-20260827-024202-60bd13 # GPU : 2x NVIDIA RTX A6000 # CPU : AMD Ryzen Threadripper PRO 7955WX 16-Cores # RAM : 184 GB bench@llm-benchmark:~$ vllm serve \ --model Nemotron-3-Nano-Omni-30B-A3B-Reasoning

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.	vllm
Modellalias?Der Name, unter dem das Modell ueber die API angesprochen wird. Genau dieser Wert muss im Request-Feld 'model' stehen.	Nemotron-3-Nano-Omni-30B-A3B-Reasoning

All benchmarks of this model To leaderboard

Model comparison

Nemotron-3-Nano-Omni-30B-A3B-Reasoning 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 PRO 6000 Blackwell Workstation Edition 1.791,5 tok/sNVIDIA GeForce RTX 5090 1.706,7 tok/sNVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition 1.651,0 tok/sNVIDIA GeForce RTX 3090 Ti 896,6 tok/s★ NVIDIA RTX A6000 527,0 tok/s this runNVIDIA GeForce RTX 5070 Ti 262,5 tok/sAMD Radeon AI PRO R9700 122,4 tok/s

CPUby processor

AMD Ryzen 9 9950X 16-Core Processor 1.791,5 tok/sAMD Ryzen 7 5800X3D 8-Core Processor 1.706,7 tok/sAMD Ryzen Threadripper PRO 9965WX 24-Cores 1.651,0 tok/sAMD Ryzen 9 8945HX with Radeon Graphics 896,6 tok/s★ AMD Ryzen Threadripper PRO 7955WX 16-Cores 527,0 tok/s this runAMD Ryzen Threadripper PRO 5975WX 32-Cores 262,5 tok/s

MBby mainboard

ASUSTeK COMPUTER INC. ProArt X870E-CREATOR WIFI 1.791,5 tok/sASUSTeK COMPUTER INC. ROG STRIX B550-A GAMING 1.706,7 tok/s★ ASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE 1.651,0 tok/s this runMeigao Innovation Technology (Shen Zhen) Co., Ltd DRFXL (MotherBoard Series) 896,6 tok/sASUSTeK COMPUTER INC. Pro WS WRX80E-SAGE SE WIFI 262,5 tok/s

ENGby engine

llama.cpp 1.791,5 tok/s★ vLLM 527,0 tok/s this run

DRVby driver

unbekannt 1.791,5 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 URL Reset

⚡ Electricity price EUR/kWh

⚙️ System utilization 100 %

🖥️ Acquisition EUR

🔌 Idle 80 W

⚡ TDP 371 W

☁️ External LLM (API) show

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

Electricity0.30 EUR/kWh

Avg power (incl. idle)371 W estimated (TDP)GPU 300 + CPU 46 + Board 25 W full load

Avg cost / hourEUR 0.11

Electricity / 1M tokensEUR 0.24

Token / kWh1.26M

Acquisition (system)EUR 8,424 full priceGPU EUR 3,000 · CPU EUR 1,399 · Board EUR 1,299 · RAM EUR 2,576 · PSU EUR 150

Electricity (2 years)–

TCO (2 years)EUR 10,374

Output tokens (2 years)8.16B

☁️ 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 (80 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

	Nemotron-3-Nano-Omni-30B-A3B-ReasoningNVIDIA RTX A6000	Nemotron-3-Nano-Omni-30B-A3B-ReasoningNVIDIA GeForce RTX 5090	Nemotron-3-Nano-Omni-30B-A3B-ReasoningNVIDIA RTX PRO 6000 Blackwell Workstation Edition	Nemotron-3-Nano-Omni-30B-A3B-Reasoning3x 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)	–	–	–	–

NVIDIA RTX A6000 · this run 129 tok/s NVIDIA GeForce RTX 5090 359 tok/s NVIDIA RTX PRO 6000 Blackwell Workstation Edition 357 tok/s 3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition 331 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.