Arao AI
Back to search results

Qwen3.5-4B · IQ3_S

Start a discussion
NVIDIA RTX 508016GBStrata v0.45L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3.5-4B (IQ3_S) on NVIDIA RTX 5080 with Strata, contributed by 1 independent source platforms; metrics are averages of published measurements.

356.3 tok/s

Decode

Decode speed

1,901.07 tok/s

Prefill

Prefill speed

2.31 s

TTFT

Time to first token

7.64 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

Today

Last verified

Performance

  1. llama.cpp · IQ1_S Decode 350.61 · Prefill 1829.73 ·
  2. llama.cpp · IQ2_XS Decode 312.6 · Prefill 1217.96 ·
  3. Strata · IQ3_S (current)Decode 356.3 · Prefill 1901.07 ·
  4. llama.cpp · IQ4_XS Decode 133.14 · Prefill 731.5 ·
  5. Strata · Q3_K_M Decode 231.73 · Prefill 785.76 ·

Core figures

Decode (avg)
356.3 tok/s
Prefill (avg)
1,901.07 tok/s
TTFT (avg)
2.31 s
VRAM (avg)
7.64 GB
MTP acceptance rate
0.75
TTFB
2.85 GB
Power draw
225 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3.5-4B
Quantization
IQ3_S
Framework
Strata
Version
v0.45
Context length
16384 tokens
Batch size
256
GPU layers
12
Flash Attention
Off

Hardware

Nominal and measured figures are shown side by side; whether it runs is the reader's call

GPU
NVIDIA RTX 5080
Nominal VRAM
16 GB
Measured VRAM (avg)
7.64 GB
OS
Ubuntu 22.04
Driver
560.94
CUDA
12.8
Power draw
225 W

Sources & evidence

1 measured records in total, each traceable to its original source

  1. L2 Cross-framework verifiedOtherOriginal link Verified on 2026-10-18

    v0.45 · Ubuntu 22.04 · CUDA 12.8 · 16384 ctx

    356.3 tok/s

    Decode

    1,901.07 tok/s

    Prefill

    2.31 s

    TTFT

    7.64 GB

    VRAM

    0.75

    MTP

    225 W

    Power draw

    L2 batch seed 2026-10-18: Strata on hw_id=14 running model_id=48 at IQ3_S