# LLM-Benchmark Konfiguration # Modell : gemma-4-E4B-it # Engine : vLLM # Run-ID : run-20260804-052133-c45fb3 # GPU : NVIDIA GeForce RTX 3090 Ti # CPU : AMD Ryzen 5 5600X 6-Core Processor # RAM : 30 GB bench@llm-benchmark:~$ /opt/vllm-gemma/venv/bin/python /opt/vllm-gemma/venv/bin/vllm serve google/gemma-4-E4B-it \ --served-model-name gemma-4-E4B-it \ --host 192.168.41.116 \ --port 8000 \ --dtype bfloat16 \ --max-model-len 8192 \ --gpu-memory-utilization 0.90 \ --enforce-eager

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.	gemma-4-E4B-it
Kontextlaenge?Maximale Anzahl Tokens (Eingabe + erzeugte Ausgabe zusammen), die das Modell pro Anfrage verarbeiten kann.	8192
Alias?Anzeigename des Modells nach aussen (served model name), unabhaengig vom Dateinamen.	gemma-4-E4B-it
Dtype?Zahlenformat der Modellgewichte bei der Berechnung (z.B. auto, float16, bfloat16). 'auto' waehlt automatisch das vom Modell empfohlene Format.	bfloat16
Kontext?Groesse des Kontextfensters in Token. 0 = der beim Training verwendete Kontext des Modells.	8192
GPU-Speicher?Anteil des GPU-Speichers (0 bis 1), den vLLM belegen darf. 0.92 = 92 %. Hoeher = mehr Platz fuer den KV-Cache (mehr/laengere parallele Anfragen), aber groesseres Risiko fuer 'Out of Memory'.	0.90
Enforce-Eager?Schaltet die optimierte Graph-Ausfuehrung (CUDA-/HIP-Graphs) AB und rechnet Schritt fuer Schritt. Startet schneller und spart etwas VRAM, ist im laufenden Betrieb aber meist langsamer als mit Graphs.	aktiv

All benchmarks of this model To leaderboard

Model comparison

gemma-4-E4B-it 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 1.758,6 tok/sNVIDIA GeForce RTX 5090 1.606,1 tok/sNVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition 1.423,3 tok/sNVIDIA GeForce RTX 5070 Ti 1.082,4 tok/s★ NVIDIA GeForce RTX 3090 Ti 983,7 tok/s this runNVIDIA GB10 (DGX Spark) 488,2 tok/sIntel Arc Pro B70 474,8 tok/s

CPUby processor

AMD Ryzen 9 9950X 16-Core Processor 1.758,6 tok/sAMD Ryzen 7 5800X3D 8-Core Processor 1.606,1 tok/sAMD Ryzen Threadripper PRO 9965WX 24-Cores 1.423,3 tok/sAMD Ryzen Threadripper PRO 5975WX 32-Cores 1.082,4 tok/sAMD Ryzen 9 8945HX with Radeon Graphics 983,7 tok/sNVIDIA Grace 488,2 tok/sAMD Ryzen 9 7945HX with Radeon Graphics 474,8 tok/s★ AMD Ryzen 5 5600X 6-Core Processor 346,3 tok/s this run

MBby mainboard

ASUSTeK COMPUTER INC. ProArt X870E-CREATOR WIFI 1.758,6 tok/sASUSTeK COMPUTER INC. ROG STRIX B550-A GAMING 1.606,1 tok/sASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE 1.423,3 tok/sASUSTeK COMPUTER INC. Pro WS WRX80E-SAGE SE WIFI 1.082,4 tok/sMeigao Innovation Technology (Shen Zhen) Co., Ltd DRFXL (MotherBoard Series) 983,7 tok/sASUSTeK COMPUTER INC. GX10 488,2 tok/sShenzhen Meigao Electronic Equipment Co.,Ltd DRFXI (MotherBoard Series) 474,8 tok/s★ ASUSTeK COMPUTER INC. PRIME A520M-K 346,3 tok/s this run

ENGby engine

llama.cpp 1.758,6 tok/s★ vLLM 1.022,8 tok/s this run

DRVby driver

unbekannt 1.758,6 tok/sNVIDIA 590.48.01 / CUDA 13.1 488,2 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 25 W

⚡ TDP 450 W

☁️ External LLM (API) show

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

Electricity0.30 EUR/kWh

Avg power (incl. idle)450 W estimated (TDP)GPU 450 W full load

Avg cost / hourEUR 0.14

Electricity / 1M tokensEUR 0.21

Token / kWh1.42M

Acquisition (system)EUR 1,149 partial priceGPU EUR 999 · PSU EUR 150

Electricity (2 years)–

TCO (2 years)EUR 3,514

Output tokens (2 years)11.19B

☁️ 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 (25 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

	gemma-4-E4B-itNVIDIA GeForce RTX 3090 Ti	gemma-4-E4B-itNVIDIA RTX PRO 6000 Blackwell Workstation Edition	gemma-4-E4B-itNVIDIA GeForce RTX 5090	gemma-4-E4B-it3x 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 GeForce RTX 3090 Ti · this run 177 tok/s NVIDIA RTX PRO 6000 Blackwell Workstation Edition 914 tok/s NVIDIA GeForce RTX 5090 893 tok/s 3× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition 812 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

Marcel Sommer

@marcelsommer