NVIDIA open-sources BioNeMo Inference Runtime, boosting protein structure prediction throughput 2.9x
AllThingsApx · x · 2026-09-11
NVIDIA has opened public beta of BioNeMo Inference Runtime, a PyTorch-native open-source library that accelerates biomolecular structure prediction on NVIDIA GPUs.
- Models accelerated: Boltz-2, OpenFold2, and Protenix v2 via dedicated kernels and CUDA Graphs, with Ray-driven GPU replicas for scaling.
- Benchmark: On 1,000 human dimer targets across 8xH100 GPUs, accelerated Boltz-2 folded 58.5K residues per GPU-hour vs 20.2K for a torch-compiled open-source baseline — a 2.90x residue-normalized throughput gain; energy to fold 1M targets drops from 35 MWh to 11 MWh.
- Three optimization layers: kernel selection, CUDA Graph module optimization, and Ray pipeline scaling.
- Partners include SandboxAQ, Xaira Therapeutics, Latent Labs, and the UW Institute for Protein Design.
More from Infra
- Anthropic's ~11 GW contracted compute estimated at ~42% NVIDIA GPUs, ~4.5 GW Google TPU — ns123abc · 2026-10-27
- EdgeRock-4B holds KV cache at 72 MiB across 256K context, 99.1% smaller than Qwen3.5-4B — alifcoder · 2026-10-27
- RTX 5080 price jumps from $1658 to $2318 in ten days on Amazon — Frizzy-MacDrizzle · 2026-10-27
- ComfyUI on 12GB VRAM: measured peaks show offloading turns 24s into 11 minutes — Impossible-Mail-4619 · 2026-10-27
- Neon rebuilds its docs search entirely on its own stack: Lakebase Search, Functions, AI Gateway — matei_zaharia · 2026-10-27
- PyTorch details 2.12x faster on-device streaming ASR on Arm SME2 via ExecuTorch — PyTorch · 2026-10-27