Radeon Vega 3

Everyday Tasks

Office

Evaluation - Difficult

Compared with GeForce GT 1030, a baseline GPU for everyday tasks such as Office work and web browsing, it delivers -63% performance.

Light Gaming

ValorantMinecraftRobloxUNDERTALELeague of Legends

Evaluation - Difficult

Compared with GeForce GTX 1650, a baseline GPU for light games such as Valorant, Minecraft, Roblox, UNDERTALE, and League of Legends at FHD / 120 FPS, it delivers -89% performance.

Mid-range Gaming

Apex LegendsOverwatchFortniteDead by Daylight

Evaluation - Difficult

Compared with GeForce RTX 3060 8GB, a baseline GPU for mid-range games such as Apex Legends, Overwatch, Fortnite, and Dead by Daylight at FHD / 120 FPS, it delivers -94% performance.

As an integrated GPU without dedicated VRAM, available capacity depends on the host machine's system memory.

Heavy Gaming

Battlefield 6Cyberpunk 2077Escape from TarkovMonster Hunter Wilds

Evaluation - Difficult

Compared with GeForce RTX 3070, a baseline GPU for heavy games such as Battlefield 6, Cyberpunk 2077, Escape from Tarkov, and Monster Hunter Wilds at FHD / 120 FPS, it delivers -96% performance.

As an integrated GPU without dedicated VRAM, available capacity depends on the host machine's system memory.

1080p Video Editing

Premiere ProDaVinci ResolveFilmora

Evaluation - Difficult

Compared with GeForce GTX 1650, a baseline GPU for 1080p video editing in tools like Premiere Pro, DaVinci Resolve, and Filmora, it delivers -89% performance.

4K Video Editing

Premiere ProDaVinci Resolve StudioAfter Effects

Evaluation - Difficult

Compared with GeForce RTX 3060 12GB, a baseline GPU for 4K video editing in Premiere Pro, DaVinci Resolve Studio, and After Effects, it delivers -95% performance.

As an integrated GPU without dedicated VRAM, available capacity depends on the host machine's system memory.

FHD Image Editing

PhotoshopLightroomFigmaCLIP STUDIO PAINT

Evaluation - Difficult

Compared with GeForce GTX 1050 Ti, a baseline GPU for FHD image editing in Photoshop, Lightroom, Figma, and Clip Studio Paint, it delivers -86% performance.

3D CG Creation

BlenderMaya3ds Max

Evaluation - Difficult

Compared with GeForce RTX 2060, a baseline GPU for 3D CG work (modeling and viewport preview) in tools like Blender, Maya, and 3ds Max, it delivers -94% performance.

As an integrated GPU without dedicated VRAM, available capacity depends on the host machine's system memory.

Game Development

UnityUnreal Engine 5Godot Engine

Evaluation - Difficult

Compared with GeForce RTX 3060 12GB, a baseline GPU for game development (editor work and in-editor preview) in Unity, Unreal Engine 5, and Godot Engine, it delivers -95% performance.

As an integrated GPU without dedicated VRAM, available capacity depends on the host machine's system memory.

CAD

AutoCADFusion 360

Evaluation - Difficult

Compared with GeForce GTX 1650, a baseline GPU for CAD work in tools like AutoCAD and Fusion 360, it delivers -89% performance.

Local LLM

GemmaQwenLlamaDeepSeek

Evaluation - Not applicable

24 GB+Top tier
16 GB+Mainstream tier
8 GB+Entry tier
Under 8 GB×Below entry

As an integrated GPU sharing system memory, it is outside the scope of this quick check.

As a rule of thumb, (quantized) 8B models such as Gemma, Qwen, Llama, and DeepSeek need 8 GB of VRAM, 22B models need 16 GB, and 32B models need 24 GB. At the same VRAM capacity, newer generations also run faster. The field moves fast, so this is a quick check of where Radeon Vega 3's VRAM stands.

AI Image Generation

KreaFluxPonyIllustrious

Evaluation - Not applicable

24 GB+Top tier
16 GB+Mainstream tier
8 GB+Entry tier
Under 8 GB×Below entry

As an integrated GPU sharing system memory, it is outside the scope of this quick check.

As a rule of thumb, (quantized) 2–6B models such as Krea, Flux, SDXL, Pony, Anima, and Illustrious need 8 GB of VRAM, 12B models need 16 GB, and 20B models need 24 GB. At the same VRAM capacity, newer generations also run faster. The field moves fast, so this is a quick check of where Radeon Vega 3's VRAM stands.

Note: many tools in this field assume CUDA, so non-NVIDIA GPUs face extra hurdles to get running.

AI Video Generation

WanLTXHunyuan

Evaluation - Not applicable

24 GB+Comfortable tier
16 GB+Practical tier
8 GB+Trial tier
Under 8 GB×Below entry

As an integrated GPU sharing system memory, it is outside the scope of this quick check.

As a rule of thumb, (quantized) models such as Wan, LTX, and Hunyuan need at least 16 GB of VRAM for practical use. At the same VRAM capacity, newer generations also run faster. The field moves fast, so this is a quick check of where Radeon Vega 3's VRAM stands.

Note: many tools in this field assume CUDA, so non-NVIDIA GPUs face extra hurdles to get running.