Contributed byMario AlkaNVIDIA Nemotron-3-Super-120B-A12B
Performance benchmark · measured on 09.10.2026 06:42
Benchmark-ID
run-20261009-090853-dd2e27Timebench 3 - Kombi (Prefill + Generation)MoE120BRuntime: llama.cppQuantisierung: Q4_K_M
ℹ
Public note: Dell PowerEdge R820 . 2x AMD Radeon RX 7900 XTX (48 GB VRAM gesamt, ROCm/HIP) . RAM: 32x 16 GB SK Hynix DDR3-1600 ECC RDIMM (HMT42GR7AFR4A-PB) = 512 GB @ 1600 MT/s
Generation6,03tok/s
Prefill58,31tok/s
Time to First Token186.063,00ms
Total duration1.200,00s
Concurrency5parallel
Ranking in the field
115of 120 systems
Performance benchmark · Primary metric: Generation-Speed (tok/s) · 5× concurrent
This run is better than 4 % of all comparable systems.
Generation
6,0 tok/s
-96 % vs Ø 165,1
Prefill
58,3 tok/s
-99 % vs Ø 6.346,8
Time to First Token
186.063 ms
+984 % vs Ø 17.171
Distribution in the field2 – 496 tok/s
Wie schlägt sich dieser Benchmark mit anderen Modellen?
Qwen3-0.6BAMD Radeon RX 7900 XTX · run-20260922-095735-570c64
495,5 tok/s
Llama-3.2-1B-Instruct2× 2x AMD Radeon RX 7900 XTX · run-20260927-015653-2a7127
479,2 tok/s
Llama-3.2-1B-Instruct2× 2x AMD Radeon RX 7900 XTX · run-20260927-001710-b1912f
443,7 tok/s
Llama-3.2-1B-Instruct2× 2x AMD Radeon RX 7900 XTX · run-20261008-105916-852cbe
376,5 tok/s
SmolLM2-1.7B-Instruct2× 2x AMD Radeon RX 7900 XTX · run-20260927-023029-88b3c3
338,7 tok/s
gemma-4-E2B-itAMD Radeon RX 7900 XTX · run-20260922-115730-99b6cb
331,5 tok/s
Qwen2.5-7B-Instruct2× 2x AMD Radeon RX 7900 XTX · run-20260926-171705-b0ed03
326,7 tok/s
Llama-3.2-3B-Instruct2× 2x AMD Radeon RX 7900 XTX · run-20260927-015653-1ca7fc
318,5 tok/s
Llama-3.1-Nemotron-Nano-4B-v1.1AMD Radeon RX 7900 XTX · run-20260922-163818-387a23
318,2 tok/s
gemma-4-E2B-itAMD Radeon RX 7900 XTX · run-20260922-115730-c821a3
316,4 tok/s
gpt-oss-20bAMD Radeon RX 7900 XTX · run-20260922-115731-1da105
314,8 tok/s
Qwen3-1.7B2× 2x AMD Radeon RX 7900 XTX · run-20260926-214312-100b47
314,1 tok/s
gpt-oss-20bAMD Radeon RX 7900 XTX · run-20260922-115731-94445e
310,6 tok/s
Llama-3.1-Nemotron-Nano-4B-v1.1AMD Radeon RX 7900 XTX · run-20260922-163818-e33391
294,9 tok/s
Nemotron-3-Super-120B-A12B this run2× 2x AMD Radeon RX 7900 XTX · run-20261009-090853-dd2e27
6,0 tok/s
How does this benchmark compare on other GPUs?
Same model on different hardware · 5× concurrent · Generation (tok/s)
NVIDIA RTX PRO 6000 Blackwell Workstation Editionrun-20260729-060802-f577cc
348,0 tok/s
3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20260803-043845-cedaac
238,5 tok/s
3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20260803-043845-bfdd78
238,3 tok/s
3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20260803-043844-2b6ba6
238,0 tok/s
3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20260803-043844-9284ba
236,1 tok/s
4× AMD Radeon AI PRO R9700run-20260813-073605-dff616
69,4 tok/s
NVIDIA GeForce RTX 5070 Tirun-20260729-105341-2f8335
29,8 tok/s
3× AMD Radeon AI PRO R9700run-20260805-053415-888eb2
23,2 tok/s
3× AMD Radeon AI PRO R9700run-20260805-053414-aa99c6
23,0 tok/s
3× AMD Radeon AI PRO R9700run-20260805-053414-fe8ee7
23,0 tok/s
4× AMD Radeon AI PRO R9700run-20260817-111922-0cbac1
20,2 tok/s
NVIDIA GeForce RTX 5070 Tirun-20260803-043844-bc091e
16,8 tok/s
NVIDIA GeForce RTX 3090 Tirun-20260731-071555-0da588
16,1 tok/s
NVIDIA RTX PRO 6000 Blackwell Workstation Editionrun-20260803-043843-36721b
12,2 tok/s
Hardware
GPU: 2x AMD Radeon RX 7900 XTX · 24 GB VRAM
CPU: 4x Intel(R) Xeon(R) CPU E5-4657L v2 @ 2.40GHz
RAM: 504 GB
Mainboard: Dell Inc. PowerEdge R820
Setup
Runtime: llama.cpp
Quantization: Q4_K_M
Model: Nemotron-3-Super-120B-A12B
Configuration
# LLM-Benchmark Konfiguration
# Modell : Nemotron-3-Super-120B-A12B
# Engine : llama.cpp
# Run-ID : run-20261009-090853-dd2e27
# GPU : 2x AMD Radeon RX 7900 XTX
# CPU : 4x Intel(R) Xeon(R) CPU E5-4657L v2 @ 2.40GHz
# RAM : 504 GB
bench@llm-benchmark:~$ llama-server \
-m nvidia_Nemotron-3-Super-120B-A12B-IQ4_XS/nvidia_Nemotron-3-Super-120B-A12B-IQ4_XS-00001-of-00002.gguf \
--alias NVIDIA-Nemotron-3-Super-120B-A12B \
-ngl 999 \
-sm layer \
-fa on \
-c 40960 \
-np 10 \
--n-cpu-moe 99 '(2x' RX 7900 XTX ROCm/HIP, 'moe-offload)'
| 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-Super-120B-A12B |
| Kontextlaenge?Maximale Anzahl Tokens (Eingabe + erzeugte Ausgabe zusammen), die das Modell pro Anfrage verarbeiten kann. | 40960 |
| Modellpfad?Pfad zur GGUF-Modelldatei, die geladen und ausgeliefert wird. | nvidia_Nemotron-3-Super-120B-A12B-IQ4_XS/nvidia_Nemotron-3-Super-120B-A12B-IQ4_XS-00001-of-00002.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 |
| Split-Mode?Verteilung ueber mehrere GPUs: none (nur eine GPU), layer (Layer aufteilen) oder row (Tensoren zeilenweise). | layer |
| fa | on |
| Kontext?Groesse des Kontextfensters in Token. 0 = der beim Training verwendete Kontext des Modells. | 40960 |
| np | 10 |
| n-cpu-moe | 99 |
| execution_type | local |
Anzeige
Model comparison
Nemotron-3-Super-120B-A12B 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 600,9 tok/sNVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition 423,4 tok/sAMD Radeon AI PRO R9700 80,5 tok/sNVIDIA GeForce RTX 5070 Ti 32,6 tok/sNVIDIA GeForce RTX 3090 Ti 18,9 tok/sNVIDIA GB10 (DGX Spark) 11,4 tok/s★ AMD Radeon RX 7900 XTX 6,1 tok/s this runNVIDIA GeForce RTX 5090 5,5 tok/s
CPUby processor
AMD Ryzen 9 9950X 16-Core Processor 600,9 tok/sAMD Ryzen Threadripper PRO 9965WX 24-Cores 423,4 tok/sAMD Ryzen Threadripper PRO 7955WX 16-Cores 80,5 tok/sAMD Ryzen Threadripper PRO 5975WX 32-Cores 32,6 tok/sAMD Ryzen 9 8945HX with Radeon Graphics 18,9 tok/sNVIDIA Grace 11,4 tok/s★ Intel(R) Xeon(R) CPU E5-4657L v2 @ 2.40GHz 6,1 tok/s this runAMD Ryzen 7 5800X3D 8-Core Processor 5,5 tok/s
MBby mainboard
ASUSTeK COMPUTER INC. ProArt X870E-CREATOR WIFI 600,9 tok/sASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE 423,4 tok/sASUSTeK COMPUTER INC. Pro WS WRX80E-SAGE SE WIFI 32,6 tok/sMeigao Innovation Technology (Shen Zhen) Co., Ltd DRFXL (MotherBoard Series) 18,9 tok/sASUSTeK COMPUTER INC. GX10 11,4 tok/s★ Dell Inc. PowerEdge R820 6,1 tok/s this runASUSTeK COMPUTER INC. ROG STRIX B550-A GAMING 5,5 tok/s
DRVby driver
unbekannt 600,9 tok/sNVIDIA 590.48.01 / CUDA 13.1 11,4 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 URLReset
⚡ Electricity price
EUR/kWh
⚙️ System utilization
100 %
🖥️ Acquisition
EUR
🔌 Idle
35 W
⚡ TDP
749 W
☁️ External LLM (API)
Electricity0.30 EUR/kWh
Avg power (incl. idle)749 W estimated (TDP)GPU 710 + CPU 29 + Board 10 W full load
Avg cost / hourEUR 0.22
Electricity / 1M tokensEUR 10.35
Token / kWh28.99K
Acquisition (system)EUR 4,353 partial priceGPU EUR 2,098 · CPU EUR 59 · RAM EUR 2,016 · PSU EUR 180
Electricity (2 years)–
TCO (2 years)EUR 8,288
Output tokens (2 years)380.32M
☁️ 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 (35 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-Super-120B-A12B2x AMD Radeon RX 7900 XTX | Nemotron-3-Super-120B-A12BNVIDIA RTX PRO 6000 Blackwell Workstation Edition | Nemotron-3-Super-120B-A12B3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition | Nemotron-3-Super-120B-A12B3x 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) | – | – | – | – |
2× AMD Radeon RX 7900 XTX · this run 6 tok/sNVIDIA RTX PRO 6000 Blackwell Workstation Edition 348 tok/s3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition 238 tok/s3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition 238 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.