Contributed byMario AlkaNVIDIA NVIDIA-Nemotron-3-Nano-4B
Performance benchmark · measured on 29.07.2026 23:42
Benchmark-ID
run-20260730-035052-93e207Timebench 3 - Kombi (Prefill + Generation)Dense4BRuntime: llama.cppQuantisierung: Q4_K_M
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.
Generation354,29tok/s
Prefill3.544,48tok/s
Time to First Token30.044,50ms
Total duration180,96s
Concurrency10parallel
Ranking in the field
271of 464 systems
Performance benchmark · Primary metric: Generation-Speed (tok/s) · 10× concurrent
This run is better than 42 % of all comparable systems.
Generation
354,3 tok/s
-49 % vs Ø 701,3
Prefill
3.544,5 tok/s
-46 % vs Ø 6.593,0
Time to First Token
30.045 ms
-30 % vs Ø 43.008
Distribution in the field0 – 2.491 tok/s
Wie schlägt sich dieser Benchmark mit anderen Modellen?
gemma-4-E2B-itNVIDIA GeForce RTX 5090 · run-20260728-184455-e28775
2.491,2 tok/s
gemma-4-E2B-itNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260727-181606-b37ed9
2.414,4 tok/s
NVIDIA-Nemotron-3-Nano-4B same modelNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260727-181607-4cca42
2.182,7 tok/s
gemma-4-E2B-it3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260728-184456-d87c73
2.143,6 tok/s
gpt-oss-20bNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260730-035052-9c966e
1.995,4 tok/s
gpt-oss-20bNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260728-140954-8007c4
1.993,5 tok/s
NVIDIA-Nemotron-3-Nano-4B same model3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260728-194136-0aed86
1.937,7 tok/s
NVIDIA-Nemotron-3-Nano-4B same model3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260728-194136-8db0fa
1.911,6 tok/s
NVIDIA-Nemotron-3-Nano-4B same model3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260727-181606-e0eafd
1.897,0 tok/s
gpt-oss-20bNVIDIA GeForce RTX 5090 · run-20260729-032121-7c961d
1.890,7 tok/s
NVIDIA-Nemotron-3-Nano-4B same modelNVIDIA GeForce RTX 5090 · run-20260728-194135-2ed9dd
1.878,7 tok/s
gpt-oss-20b3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260728-034052-26063b
1.835,7 tok/s
NVIDIA-Nemotron-3-Nano-4B same model3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260728-194136-a77046
1.830,3 tok/s
gpt-oss-20b3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260729-032121-c6e7e0
1.819,5 tok/s
NVIDIA-Nemotron-3-Nano-4B this runNVIDIA GB10 · run-20260730-035052-93e207
354,3 tok/s
How does this benchmark compare on other GPUs?
Same model on different hardware · 10× concurrent · Generation (tok/s)
NVIDIA RTX PRO 6000 Blackwell Workstation Editionrun-20260727-181607-4cca42
2.182,7 tok/s
3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20260728-194136-0aed86
1.937,7 tok/s
3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20260728-194136-8db0fa
1.911,6 tok/s
3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20260727-181606-e0eafd
1.897,0 tok/s
NVIDIA GeForce RTX 5090run-20260728-194135-2ed9dd
1.878,7 tok/s
3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20260728-194136-a77046
1.830,3 tok/s
NVIDIA GeForce RTX 5070 Tirun-20260727-090153-f232e7
1.197,0 tok/s
NVIDIA GeForce RTX 3090 Tirun-20260727-145956-3936c2
992,1 tok/s
NVIDIA GB10 this runrun-20260730-035052-93e207
354,3 tok/s
3× AMD Radeon AI PRO R9700run-20260723-023427-9684b2
218,9 tok/s
AMD Radeon 8060S Graphicsrun-20260723-015518-0bdcee
108,0 tok/s
Keine GPU (CPU-only)run-20260723-015518-1486fe
0,3 tok/s
Hardware
GPU: NVIDIA GB10 · 128 GB VRAM
CPU: NVIDIA Grace
RAM: 120 GB
Mainboard: ASUSTeK COMPUTER INC. GX10
Setup
Runtime: llama.cpp
Quantization: Q4_K_M
Driver: NVIDIA 590.48.01 / CUDA 13.1
Model: NVIDIA-Nemotron-3-Nano-4B
Configuration
# LLM-Benchmark Konfiguration
# Modell : NVIDIA-Nemotron-3-Nano-4B
# Engine : llama.cpp
# Run-ID : run-20260730-035052-93e207
# GPU : NVIDIA GB10
# CPU : NVIDIA Grace
# RAM : 120 GB
bench@llm-benchmark:~$ /home/godcore/llama.cpp/build/bin/llama-server \
-m /home/godcore/.cache/huggingface/hub/models--DevQuasar--nvidia.NVIDIA-Nemotron-3-Nano-4B-BF16-GGUF/snapshots/86e3ddc0b34d94852cae725ade1b9b015b516f38/nvidia.NVIDIA-Nemotron-3-Nano-4B-BF16.f16.gguf.Q4_K_M.gguf \
--alias NVIDIA-Nemotron-3-Nano-4B \
--host 0.0.0.0 \
--port 8000 \
-ngl 999 \
-c 16384 \
-np 4 \
--jinja
| 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. | NVIDIA-Nemotron-3-Nano-4B |
| Kontextlaenge?Maximale Anzahl Tokens (Eingabe + erzeugte Ausgabe zusammen), die das Modell pro Anfrage verarbeiten kann. | 16384 |
| Modellpfad?Pfad zur GGUF-Modelldatei, die geladen und ausgeliefert wird. | /home/godcore/.cache/huggingface/hub/models--DevQuasar--nvidia.NVIDIA-Nemotron-3-Nano-4B-BF16-GGUF/snapshots/86e3ddc0b34d94852cae725ade1b9b015b516f38/nvidia.NVIDIA-Nemotron-3-Nano-4B-BF16.f16.gguf.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. | 999 |
| Kontext?Groesse des Kontextfensters in Token. 0 = der beim Training verwendete Kontext des Modells. | 16384 |
| np | 4 |
Anzeige
Model comparison
NVIDIA-Nemotron-3-Nano-4B 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 2.182,7 tok/sNVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition 1.937,7 tok/sNVIDIA GeForce RTX 5090 1.878,7 tok/sNVIDIA GeForce RTX 5070 Ti 1.197,0 tok/sNVIDIA GeForce RTX 3090 Ti 992,1 tok/s★ NVIDIA GB10 (DGX Spark) 354,3 tok/s this runAMD Radeon AI PRO R9700 261,2 tok/sAMD Radeon 8060S Graphics 113,4 tok/sCPU-only 8,0 tok/s
CPUby processor
AMD Ryzen 9 9950X 16-Core Processor 2.182,7 tok/sAMD Ryzen Threadripper PRO 9965WX 24-Cores 1.937,7 tok/sAMD Ryzen 7 5800X3D 8-Core Processor 1.878,7 tok/sAMD Ryzen Threadripper PRO 5975WX 32-Cores 1.197,0 tok/sAMD Ryzen 9 8945HX with Radeon Graphics 992,1 tok/s★ NVIDIA Grace 354,3 tok/s this runAMD Ryzen Threadripper PRO 7955WX 16-Cores 261,2 tok/sAMD RYZEN AI MAX+ 395 w/ Radeon 8060S 113,4 tok/sIntel(R) Xeon(R) CPU E5-4620 v2 @ 2.60GHz 8,0 tok/s
MBby mainboard
ASUSTeK COMPUTER INC. ProArt X870E-CREATOR WIFI 2.182,7 tok/sASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE 1.937,7 tok/sASUSTeK COMPUTER INC. ROG STRIX B550-A GAMING 1.878,7 tok/sASUSTeK COMPUTER INC. Pro WS WRX80E-SAGE SE WIFI 1.197,0 tok/sMeigao Innovation Technology (Shen Zhen) Co., Ltd DRFXL (MotherBoard Series) 992,1 tok/s★ ASUSTeK COMPUTER INC. GX10 354,3 tok/s this runBosgame AXB35-02 (BeyondMax Series) 113,4 tok/sDell Inc. PowerEdge R820 8,0 tok/s
DRVby driver
unbekannt 2.182,7 tok/s★ NVIDIA 590.48.01 / CUDA 13.1 354,3 tok/s this run
💰 Economics
Economics of this run
Operating cost, TCO and comparison with the next-best runs of the same model at identical concurrency (10× 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
15 W
⚡ TDP
140 W
☁️ External LLM (API)
Electricity0.30 EUR/kWh
Avg power (incl. idle)140 W estimated (TDP)GPU 140 W full load
Avg cost / hourEUR 0.042
Electricity / 1M tokensEUR 0.033
Token / kWh9.11M
Acquisition (system)EUR 5,800 partial priceGPU EUR 4,000 · RAM EUR 1,680 · PSU EUR 120
Electricity (2 years)–
TCO (2 years)EUR 6,536
Output tokens (2 years)22.35B
☁️ 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 (15 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
| NVIDIA-Nemotron-3-Nano-4BNVIDIA GB10 (DGX Spark) | NVIDIA-Nemotron-3-Nano-4BNVIDIA RTX PRO 6000 Blackwell Workstation Edition | NVIDIA-Nemotron-3-Nano-4B3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition | NVIDIA-Nemotron-3-Nano-4B3x 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 GB10 (DGX Spark) · this run 354 tok/sNVIDIA RTX PRO 6000 Blackwell Workstation Edition 2,183 tok/s3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition 1,938 tok/s3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition 1,912 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.