Arao AI
Back to search results

Qwen3-1.7B · FP16

Start a discussion
NVIDIA RTX 509032GBOllama v0.38L2 Cross-framework verified

This page aggregates 1 real-world runs of Qwen3-1.7B (FP16) on NVIDIA RTX 5090 with Ollama, contributed by 1 independent source platforms; metrics are averages of published measurements.

199.27 tok/s

Decode

Decode speed

1,057.11 tok/s

Prefill

Prefill speed

2.17 s

TTFT

Time to first token

8.7 GB

VRAM

VRAM usage

L2 Cross-framework verified

1

Measured runs

1

Independent sources

Other

Source platforms

Today

Last verified

Performance

  1. llama.cpp · Q5_K_M Decode 353.77 · Prefill 1906.09 ·
  2. llama.cpp · Q6_K Decode 345.13 · Prefill 1779.25 ·
  3. Ollama · FP16 (current)Decode 199.27 · Prefill 1057.11 ·
  4. Strata · Q2_K Decode 300.65 · Prefill 1243.6 ·
  5. Strata · Q4_K_M Decode 280.21 · Prefill 1416.33 ·

Core figures

Decode (avg)
199.27 tok/s
Prefill (avg)
1,057.11 tok/s
TTFT (avg)
2.17 s
VRAM (avg)
8.7 GB
MTP acceptance rate
—
TTFB
1.59 GB
Power draw
424.1 W

Configuration

Member-level fields are taken from the most recent run

Model
Qwen3-1.7B
Quantization
FP16
Framework
Ollama
Version
v0.38
Context length
2048 tokens
Batch size
256
GPU layers
20
Flash Attention
Off

Hardware

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

GPU
NVIDIA RTX 5090
Nominal VRAM
32 GB
Measured VRAM (avg)
8.7 GB
OS
macOS 15.0
Driver
560.94
CUDA
12.4
Power draw
424.1 W

Sources & evidence

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

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

    v0.38 · macOS 15.0 · CUDA 12.4 · 2048 ctx

    199.27 tok/s

    Decode

    1,057.11 tok/s

    Prefill

    2.17 s

    TTFT

    8.7 GB

    VRAM

    —

    MTP

    424.1 W

    Power draw

    L2 batch seed 2026-10-10: Ollama on hw_id=13 running model_id=55 at FP16