Contributed byMario AlkaQwen (Alibaba) Qwen3.8-Flash-Next
Performance benchmark · measured on 08.11.2026 14:12
Benchmark-ID
run-20261108-133117-6ed91dTimebench 3 - Kombi (Prefill + Generation)MoE180BRuntime: strataQuantisierung: IQ2_XS
ℹ
Public note: Referenzmessung VOR Installation von gpu-unlock (zagitta/gpu-unlock 2.0.0), RTX 3090 24 GB, Treiber 595.99.02, Power-Limit 370 W, Kernel 7.0.0-38. Strata mit Qwen3.8-Flash-Next in GSQ-RCO IQ2_XS (63,35 GiB), Kontext 4096, 8 Streams, Expert-Cache 14,67 GiB im VRAM. Bei 10 Anfragen warten 2 auf einen freien Stream (Engine-Obergrenze 8), der Summenwert liegt daher ueber dem echten Durchsatz. Das Modell laeuft zudem deutlich unter dem empfohlenen Speicher (31 statt 48 GB), die Experten werden von der SSD nachgeladen. Teil eines Vorher-Nachher-Vergleichs zur Tensor-Core-Entsperrung auf GeForce-Karten.
Generation184,18tok/s
Prefill3.927,18tok/s
Time to First Token11.977,00ms
Total duration153,87s
Concurrency10parallel
Ranking in the field
937of 1499 systems
Performance benchmark · Primary metric: Generation-Speed (tok/s) · 10× concurrent
This run is better than 38 % of all comparable systems.
Generation
184,2 tok/s
-59 % vs Ø 448,7
Prefill
3.927,2 tok/s
-43 % vs Ø 6.886,2
Time to First Token
11.977 ms
-73 % vs Ø 43.960
Distribution in the field0 – 3.484 tok/s
Wie schlägt sich dieser Benchmark mit anderen Modellen?
NVIDIA-Nemotron-3.5-Lightning-30B-A3BNVIDIA GeForce RTX 3090 · run-20261026-180644-063abd
3.484,3 tok/s
NVIDIA-Nemotron-3.5-Lightning-30B-A3BNVIDIA GeForce RTX 3090 · run-20261026-183943-bbe82c
2.867,2 tok/s
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
gemma-4-E2B-itNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20261023-065752-48370d
2.412,3 tok/s
NVIDIA-Nemotron-3.5-Lightning-30B-A3BNVIDIA GeForce RTX 3090 Ti · run-20261021-093815-dce558
2.263,7 tok/s
NVIDIA-Nemotron-3-Nano-4BNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20260727-181607-4cca42
2.182,7 tok/s
NVIDIA-Nemotron-3-Nano-4BNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20261023-065758-96fe6a
2.144,3 tok/s
gemma-4-E2B-it3× 3x 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
gpt-oss-20bNVIDIA RTX PRO 6000 Blackwell Workstation Edition · run-20261023-065808-332d1f
1.981,1 tok/s
NVIDIA-Nemotron-3-Nano-4B3× 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260728-194136-0aed86
1.937,7 tok/s
NVIDIA-Nemotron-3-Nano-4B3× 3x NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition · run-20260728-194136-8db0fa
1.911,6 tok/s
Qwen3.8-Flash-Next this runNVIDIA GeForce RTX 3090 · run-20261108-133117-6ed91d
184,2 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 Max-Q Workstation Editionrun-20261012-175110-ea7bf0
1.564,6 tok/s
NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20261012-182514-eada92
1.555,3 tok/s
NVIDIA RTX A6000run-20261021-093816-82f62e
1.171,8 tok/s
NVIDIA GeForce RTX 3090run-20261013-092430-0e9347
1.132,7 tok/s
NVIDIA GeForce RTX 3090run-20261013-165111-37735f
638,6 tok/s
NVIDIA RTX PRO 6000 Blackwell Workstation Editionrun-20261023-180245-c8d477
584,1 tok/s
NVIDIA DGX SPARK (GB10)run-20261013-151008-fdcb3e
492,8 tok/s
2× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20261023-180239-e453cd
424,2 tok/s
NVIDIA DGX SPARK (GB10)run-20261013-165111-6f74d0
420,2 tok/s
NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20261023-183307-d048bb
418,8 tok/s
2× NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Editionrun-20261023-180238-0a23c1
411,3 tok/s
NVIDIA GeForce RTX 5090run-20261101-162845-2ebe13
346,0 tok/s
NVIDIA RTX PRO 6000 Blackwell Workstation Editionrun-20261026-133956-44316f
344,4 tok/s
NVIDIA RTX A6000run-20261021-093816-b52dc0
242,4 tok/s
Hardware
GPU: NVIDIA GeForce RTX 3090 · 24 GB VRAM
CPU: AMD Ryzen 5 5600X 6-Core Processor
RAM: 31 GB
Mainboard: ASUSTeK COMPUTER INC. ROG STRIX B550-A GAMING
Setup
Runtime: strata
Quantization: IQ2_XS
Model: Qwen3.8-Flash-Next
Configuration
# LLM-Benchmark Konfiguration
# Modell : Qwen3.8-Flash-Next
# Engine : strata
# Run-ID : run-20261108-133117-6ed91d
# GPU : NVIDIA GeForce RTX 3090
# CPU : AMD Ryzen 5 5600X 6-Core Processor
# RAM : 31 GB
bench@llm-benchmark:~$ strata \
--model Qwen3.8-Flash-Next-GSQ-RCO-IQ2_XS
| 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. | strata |
| Kontextlaenge?Maximale Anzahl Tokens (Eingabe + erzeugte Ausgabe zusammen), die das Modell pro Anfrage verarbeiten kann. | 4096 |
| execution_type | local |
Anzeige
Model comparison
Qwen3.8-Flash-Next 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 Max-Q Workstation Edition 1.564,6 tok/sNVIDIA RTX A6000 1.171,8 tok/s★ NVIDIA GeForce RTX 3090 1.132,7 tok/s this runNVIDIA RTX PRO 6000 Blackwell Workstation Edition 584,1 tok/sNVIDIA DGX SPARK (GB10) 492,8 tok/sNVIDIA GeForce RTX 5090 346,0 tok/sNVIDIA Tesla V100 PCIe 32GB 196,2 tok/sAMD Radeon RX 7900 XTX 18,7 tok/s
CPUby processor
AMD Ryzen 9 7945HX with Radeon Graphics 1.564,6 tok/sAMD Ryzen Threadripper PRO 3945WX 12-Cores 1.171,8 tok/s★ AMD Ryzen 5 5600X 6-Core Processor 1.132,7 tok/s this runAMD Ryzen 9 9950X 16-Core Processor 584,1 tok/sNVIDIA Grace 492,8 tok/sAMD Ryzen Threadripper PRO 9965WX 24-Cores 424,2 tok/sAMD Ryzen 7 5800X3D 8-Core Processor 346,0 tok/sAMD Ryzen 3 PRO 3200GE w/ Radeon Vega Graphics 196,2 tok/sAMD EPYC 7203P 8-Core Processor 79,2 tok/sIntel(R) Xeon(R) CPU E5-4657L v2 @ 2.40GHz 63,2 tok/sAMD Ryzen Threadripper PRO 3955WX 16-Cores 18,7 tok/s
MBby mainboard
Shenzhen Meigao Electronic Equipment Co.,Ltd DRFXI (MotherBoard Series) 1.564,6 tok/sunbekannt 1.171,8 tok/s★ ASUSTeK COMPUTER INC. ROG STRIX B550-A GAMING 1.132,7 tok/s this runASUSTeK COMPUTER INC. ProArt X870E-CREATOR WIFI 584,1 tok/sASUSTeK COMPUTER INC. GX10 492,8 tok/sASUSTeK COMPUTER INC. Pro WS WRX90E-SAGE SE 424,2 tok/sASUSTeK COMPUTER INC. ROG STRIX X570-F GAMING 196,2 tok/sASRockRack ROMED8-2T/BCM 79,2 tok/sDell Inc. PowerEdge R820 63,2 tok/sASUSTeK COMPUTER INC. Pro WS WRX80E-SAGE SE WIFI 35,4 tok/s
ENGby engine
DRVby driver
unbekannt 1.564,6 tok/sNVIDIA 590.48.01 / CUDA 13.1 492,8 tok/s
💰 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
12 W
⚡ TDP
362 W
☁️ External LLM (API)
Electricity0.30 EUR/kWh
Avg power (incl. idle)362 W estimated (TDP)GPU 350 + Board 12 W full load
Avg cost / hourEUR 0.11
Electricity / 1M tokensEUR 0.16
Token / kWh1.83M
Acquisition (system)EUR 1,518 full priceGPU EUR 850 · CPU EUR 121 · Board EUR 149 · RAM EUR 248 · PSU EUR 150
Electricity (2 years)–
TCO (2 years)EUR 3,421
Output tokens (2 years)11.62B
☁️ 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 (12 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
| Qwen3.8-Flash-NextNVIDIA GeForce RTX 3090 | Qwen3.8-Flash-NextNVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition | Qwen3.8-Flash-NextNVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition | Qwen3.8-Flash-NextNVIDIA RTX A6000 | |
|---|---|---|---|---|
| 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 · this run 184 tok/sNVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition 1,565 tok/sNVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition 1,555 tok/sNVIDIA RTX A6000 1,172 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.