cd /news/generative-ai/title-sdxl-performance-on-low-vram-g… · home topics generative-ai article
[ARTICLE · art-105893] src=discuss.huggingface.co ↗ pub= topic=generative-ai verified=true sentiment=↑ positive

Title: SDXL Performance on Low‑VRAM GPU

Celestium Engine, an AI image generation tool, has added features to improve SDXL performance on low-VRAM GPUs, including dynamic FP16/FP32 switching in the VAE, autonomous VRAM cleanup, automatic CPU fallback, and the ability to run SDXL at 768x768 on 4GB GPUs with upscaling to 1536px using BuffaloCore, using around 3800 MB of VRAM.

read1 min views2 publishedAug 21, 2026

Continuing the discussion from SDXL Performance on Low VRAM: Hi, sharing the latest upgrades added to the Celestium Engine:

• Dynamic FP16 ↔ FP32 switching in the VAE

The engine now automatically switches between FP16 and FP32 inside the VAE whenever higher detail stability is required.

This reduces artifacts and preserves image quality even on low‑VRAM hardware.

• Immediate and autonomous VRAM cleanup

Every stage of the pipeline performs a full VRAM flush.

Zero fragmentation, zero accumulation, zero memory leaks.

VRAM stays stable even after long multi‑generation sessions.

• Automatic Hardware Check

Celestium now detects when the GPU is unavailable or when VRAM is insufficient.

In those cases it automatically switches to CPU fallback, supported by system RAM, without interrupting the generation.

• Stable CPU fallback (consistent style & lighting)

When the GPU can’t continue, the engine preserves the same style, lighting, and visual coherence.

No aesthetic shift between GPU → CPU execution.

• SDXL starts at 768×768 even on 4GB GPUs

The SDXL pipeline initializes at 768×768 with full stability and no crashes.

With BuffaloCore, upscaling reaches 1536px even on 4GB GPUs, with real VRAM usage around 3800 MB.

• Optimized BuffaloCore

Dynamic tensor resizing prevents overflow and fragmentation.

Fully functional even on borderline hardware.

── more in #generative-ai 4 stories · sorted by recency
── more on @celestium engine 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
Live at https://your-agent.zahid.host
Get free account → Pricing
from €0/mo · no card required
LIVE [news/title-sdxl-performan…] indexed:0 read:1min 2026-08-21 ·