A new BioNeMo recipe from NVIDIA combines grouped expert kernels, fused MXFP8 8-bit math and kernel fusion to train mixture-of-experts biological foundation models. In NVIDIA's own benchmark on eight B200 GPUs, it reached up to 2.21x the throughput of the Hugging Face baseline. Here is what was measured, what was not, and why it matters for labs without hyperscale budgets.
A shoebox-sized NASA CubeSat built around a single non-toxic fuel tank feeding both a powerful chemical engine and gentle electrospray thrusters has cleared leak, thermal vacuum and spin tests at Marshall Space Flight Center, ahead of a launch scheduled no earlier than October 1, 2026.
The top $300,000 prize went to Chinyere Ukeje of Philadelphia for the Adaptive Nourishment Infrastructure (ANI), a modular ecosystem of agriculture, fermentation, fungi cultivation and closed-loop nutrient recycling designed to produce half a crew's food away from Earth. Here is why feeding a Mars crew is one of the harder engineering problems, and what the five winners actually delivered.
OpenAI's Daybreak program will give Ukrainian teams AI tools to find and fix software vulnerabilities, working with the Ministry of Digital Transformation. What the arrangement includes, what its European precedents show, and what it does not yet prove.
The far-infrared telescope will survey the cold dust where planets, stars and water form, a band JWST and radio telescopes leave uncovered. Selection funds preliminary design; a confirmation review still stands between PRIMA and construction.
OpenAI's September 22, 2026 announcement details a rebuilt caching system for GPT-6 that stores more of the context an agent already processed, discounts reused input tokens by up to 90 percent within a 30-minute window, and gives developers new dashboards for diagnosing cache misses. Here is what that means for the cost and speed of long AI sessions.
NVIDIA's developer blog says DLSS 5's 3D-Guided Neural Rendering treats the game engine's frame as a fixed foundation and adds lighting and material detail with a deterministic, one-frame-in, one-frame-out model. It is shipping today in NBA 2K27 on GeForce RTX 50 Series GPUs. Here is what the technology does, what NVIDIA claims, and what remains unverified.
Minting is scheduled to stop in October, cross-chain capacity will narrow and Noble routes will pause in January. The timetable provides an orderly exit period, but later manual redemption remains conditional.
The September 10 investigation asks what leaders knew and how testing resumed, while public reports confirm the breach without settling internal responsibility.
Heat-driven modules passed an inflation wave around a wired ring and moved a laboratory crawler 25 millimeters, while slow cycles and thermal failures bound the result.
The Solana client’s test release replaces two silent-degradation paths with hard errors, refines disk controls and gives data plugins more context. It is not a stable release or network-wide deployment.
Remote entanglement carried error syndromes between two ion-trap modules, enabling detection and active correction. The small experiment remains well below fault-tolerant performance.
The bank moved its proprietary dollar token between related entities on Stellar while evaluating minting, redemption, freezing and clawback. The transaction was live, but its amount, identifier and broader availability remain undisclosed.
