What changed

AMD has released Life Science 26.08, a complete medical imaging stack built on ROCm 10.0, designed to run on AMD Instinct™ GPUs. This release integrates several components to accelerate the entire medical imaging workflow, from reading gigapixel images to interactive annotation.

What this means in practice

The release includes four key components: hipCIM 26.06.00, MONAI Core 1.6.0, MONAI Model Zoo 26.08, and MONAILabel 0.8.5. These components work together to provide GPU-accelerated whole-slide image I/O, validated model inference, and interactive annotation. The stack maintains compatibility with existing tools like cucim and MONAI, allowing developers to use the same APIs and checkpoints without code changes.

Limitations and availability

The stack is available for AMD Instinct™ GPUs and requires ROCm 10.0. The components have varying maturity levels: hipCIM 26.06.00 and MONAI Core 1.6.0 are generally available, while MONAI Model Zoo 26.08 and MONAILabel 0.8.5 are in early access. The release includes benchmark results showing significant performance improvements over CPU-based solutions, with hipCIM achieving up to 10× faster image reading on AMD Instinct MI355X. The MONAI Model Zoo bundles show validated inference results with Dice scores meeting upstream thresholds. MONAILabel 0.8.5 provides 131× faster inference on AMD Instinct GPUs compared to CPU, enabling faster annotation workflows.

Source facts for AMD Releases Full Medical Imaging Stack for Instinct GPUs: But a medical imaging workflow isn’t a single step. A slide or a scan has to be read off disk and decoded, fed through a model, and often handed to a clinician to review and refine. To serve this work well, AMD Instinct™ has to accelerate every step, not just one of them.
Original AINewsTrend explainer based on the cited source. This graphic summarizes source statements; it is not an independent benchmark.

Sources and further reading

  1. amd: AMD Life Science 26.08: A Complete Medical Imaging Stack on AMD Instinct™ GPUs