🏆AMD MI300X Hackathon · lablab.ai

3-Agent AI Tumour Board
Powered by AMD MI300X

The only GPU with 192 GB unified memory — running GigaPath ViT-Giant + Llama 3.3 70B simultaneously, with multi-round agent debate grounded in NCCN 2024 guidelines. No model swapping. No OOM. Just results.

▶ Run a Demo Case — one click, no setup

Live VRAM Usage

Live
AMD MI300X 192 GB HBM3—
H100 SXM5 80 GB✗ OOM
OUT OF MEMORY — Cannot load both models

GigaPath (3.2) + Llama 70B (40) + Qwen-VL (15.4) + Llama 8B (16) + LoRA (1.8) + KV (25.9) = 102.3 GB > 80 GB limit

MI300X's 192 GB HBM3 unified memory enables permanent model loading — no swapping.
192 GB
HBM3 Unified Memory
3
AI Agents
≤3
Debate Rounds
~60s
Full Pipeline

How the Board Works

🔬
Pathologist
GigaPath ViT-Giant 1.1B

Analyses histopathology patches. Generates attention heatmaps. Runs MC Dropout uncertainty. Extracts 8 biomarker scores.

📚
Researcher
RAG + NCCN 2024

Retrieves evidence from NCCN guidelines. Challenges draft plan. Flags missing molecular tests and contraindicated regimens.

👨‍⚕️
Oncologist
Llama 3.3 70B

Synthesises pathology + evidence into a full management plan. Revises under critique. Finalises with consensus score ≥70/100.

🔬 Pathologist→📚 Researcher→👨‍⚕️ Oncologist (draft)→⚖️ Debate Loop (×3)→✅ Consensus ≥70→📊 Final Report
Upload Your Own Patches
Optional: submit 224x224 histology patches (PNG/JPG). Demo cases work best for judges.