Skip to main content
Radeon RX 9070
AMD · GPU · RDNA 4

Radeon RX 9070 — Benchmarks & Specs

16 GB VRAM220W TDP
Buy Radeon RX 9070 on Amazon →

*Price sourced from Amazon.com. Price and availability subject to change.

Bottom line: how fast is the Radeon RX 9070?

At 1440p (Ultra), the Radeon RX 9070 averages 447 fps in Doom Eternal, per PCGamesN. For local LLM inference it generates 711.0 tokens/sec running qwen3:8b at Q8_0 under llama.cpp, per LocalLLaMA. In 3DMark Solar Bay it scores 106,882 points, per Notebookcheck. Its 16 GB of VRAM is the binding constraint for local inference: that capacity fits 13-17B-parameter models at Q4 with room for long context.

Every figure above is a row in the tables below, and each row links out to the review or public benchmark database the number was taken from. SpecPicks aggregates published measurements; it does not report first-party benchmark runs.

The Radeon RX 9070 is a graphics card from the RDNA 4 family from AMD. Key on-paper specs include 16 GB of GDDR6 VRAM, 220W TDP. Data on this page draws on 8+ Amazon listings, 10 synthetic benchmark results, 12 community AI inference reports, 14 measured game frame-rate results, aggregated from public benchmark databases (TechPowerUp, PassMark, Geekbench, Cinebench) and the LocalLLaMA community. Read this page when shopping the Radeon RX 9070, comparing it against other graphics cards in your build, or sizing it for a specific workload (gaming at 1080p/1440p/4K, productivity benchmarks, or local LLM inference).

Gaming Performance (measured FPS)

Average and 1% low frame rates by game, resolution, and quality preset. Bars are scaled against the fastest result on this page.

Measured gaming frame rates for the Radeon RX 9070 by game, resolution, and quality preset. “1% low” is the frame-time floor that determines perceived smoothness. Each row links to its original review or benchmark database.
Game Resolution Settings Relative Avg FPS 1% low Source
Doom Eternal 1440p Ultra 447 fps — PCGamesN 2025-03-10
Doom Eternal 1440p Ultra Nightmare 447 fps — PCGamesN 2025-03-05
Doom Eternal 1440p Ultra Nightmare RT on 289 fps — PCGamesN 2025-03-05
Doom Eternal 1440p RT Ultra RT on 289 fps — PCGamesN 2025-03-10
Doom Eternal 4K Ultra 260 fps — PCGamesN 2025-03-10
Doom Eternal 4K Ultra Nightmare 260 fps — PCGamesN 2025-03-05
Resident Evil 4 1080p Prioritize Graphics 247 fps — Gamers Nexus 2025-03-07
Resident Evil 4 1440p Prioritize Graphics 172 fps — Gamers Nexus 2025-03-07
F1 24 4K Ultra High RT on FSR Quality 168 fps — PCGamesN 2025-03-10
Doom Eternal 4K RT Ultra RT on 163 fps 108 fps PCGamesN 2025-03-10
Doom Eternal 4K Ultra Nightmare RT on 163 fps 108 fps PCGamesN 2025-03-05
Resident Evil 4 1440p Max RT on FSR Quality 162 fps — Gamers Nexus 2025-03-05
Resident Evil 4 1440p RT Ultra RT on FSR Quality 162 fps — Gamers Nexus 2025-03-07
Resident Evil 4 1440p Max RT RT on FSR Quality 162 fps — Gamers Nexus 2025-03-06

AI Inference Performance

Tokens per second under each model + quantization. Higher = faster generation. Bars compare runs across the same model.

Local LLM inference throughput on the Radeon RX 9070, in generated tokens per second. Higher is better; each row links to the community report or benchmark database it came from.
Model Quantization Relative Tokens/sec VRAM used Source
qwen3:8b Q8_0 llama.cpp 711.0 tok/s — LocalLLaMA 2026-03-27
llama3.2:1b q4_K_M llama.cpp 163.0 tok/s — LocalScore.ai 2025-06-01
phi-2 — ONNX/DirectML 139.2 tok/s — StorageReview 2025-03-05
llama2:7b q4_0 llama.cpp 114.5 tok/s — KnightLi GPU Benchmark Scoreboard 2026-04-23
mistral:7b — ONNX/DirectML 101.3 tok/s — StorageReview 2025-03-05
llama2:7b q4_0 llama.cpp 99.4 tok/s — llama.cpp GitHub Discussion #15021 2025-04-01
llama2:7b q4_0 llama.cpp 99.4 tok/s — llama.cpp GitHub Discussion 2025-08-17
gpt-oss:20b — ollama 91.9 tok/s — GitHub/hirokuze 2026-04-01
gpt-oss:20b q4_K_M ollama 91.9 tok/s — GitHub hirokuze/local-llm-benchmark-rx9070xt 2025-06-01
llama3:8b — ONNX/DirectML 87.6 tok/s — StorageReview 2025-03-05
llama3.2:3b q3_K_L llama.cpp 81.0 tok/s — LocalScore.ai 2025-06-01
qwen3.5:9b q6_K llama.cpp 62.0 tok/s — digtvbg.com (Ivan Angelov) 2025-08-01

Synthetic Benchmarks

Higher is better. Bars are scaled within each benchmark family (multi-thread, single-thread, etc.) so you can compare like-with-like at a glance.

Synthetic benchmark scores for the Radeon RX 9070 — higher is better. Each row links to the public database the number was taken from.
Benchmark Relative Score Source
3DMark Solar Bay 106,882 points Notebookcheck 2025-03-06
3DMark Fire Strike 46,934 points Notebookcheck 2025-03-06
PassMark G3D Mark 25,397 pts PassMark 2026-04-20
3DMark Port Royal 15,760 points StorageReview 2025-03-05
3DMark Time Spy Extreme 11,702 points 3DMark.com 2025-04-01
3DMark Steel Nomad 6,276 points UL Benchmarks (3DMark official) 2025-04-01
3DMark Steel Nomad 6,150 points Notebookcheck 2025-03-06
3DMark Steel Nomad 5,992 points StorageReview 2025-03-05
3DMark Speed Way 5,860 points Notebookcheck 2025-03-06
3DMark Speed Way 5,791 points StorageReview 2025-03-05

Products Featuring the Radeon RX 9070

Full Specifications

rops128
tmus224
tdp w220
opengl4.6
vulkan1.4
directx12 Ultimate (12_2)
foundryTSMC
outputs1x HDMI 2.1b 3x DisplayPort 2.1a
vram gb16
gpu chipNavi 48
l2 cache8 MB
rt cores56
bandwidth644.6 GB/s
vram typeGDDR6
cuda cores3584
generationNavi IV (RX 9000)
process nm4
gpu variantNavi 48 XT (215-176600331)
spec sourceTechPowerUp GPU Database
architectureRDNA 4.0

Radeon RX 9070 — Frequently Asked Questions

What is the Radeon RX 9070 best used for?
Radeon RX 9070 is positioned as a 16 GB VRAM RDNA 4-family graphics card. Use it for 1440p/4K gaming and mid-tier AI workloads. See the synthetic + AI benchmark tables below for measured performance.
When was the Radeon RX 9070 released, and what was its launch MSRP?
Release year and launch MSRP aren't on file for Radeon RX 9070. Current pricing is visible on the linked Amazon product cards below.
Where do the benchmark numbers on this page come from?
Synthetic benchmarks are scraped from public databases (TechPowerUp, PassMark, Geekbench Browser, Cinebench leaderboards). AI inference numbers come from the LocalLLaMA community (Reddit threads, llama.cpp / Ollama discussion logs, and Phoronix when available). Every benchmark row carries an inline source citation — click through to verify the original number.
Can the Radeon RX 9070 run local LLMs?
Yes — Radeon RX 9070 has 12 AI inference benchmarks on file (see the AI Inference Performance section above for model + tokens-per-second numbers). 16 GB VRAM is enough for 13-17B parameter models at Q4 with comfortable context.
Where can I buy the Radeon RX 9070?
Amazon listings are linked at the top of this page (in the hero CTA) and in the "Products Featuring this Hardware" section below. SpecPicks earns a small affiliate commission on qualifying purchases.

Buying guides that rank the Radeon RX 9070's class

This page is the raw performance data. The guides below turn it into a ranked pick for a specific build.

Related GPUs

Editorial guides covering the Radeon RX 9070

In-depth SpecPicks reviews, build guides, and head-to-heads referencing this graphics card.

More guides & deep dives from the SpecPicks archive

Browse all articles & guides →

More reviews from the SpecPicks archive

Browse all reviews →

More buying guides from SpecPicks

Browse all buying guides →
Radeon RX 9070
Radeon RX 9070
$689.99
View on Amazon →