Nvidia breaches new quantum computing milestone, using 90% fewer qubits — and it could now improve drug discovery, financial modeling and more
THE SO WHAT
Simulating 104 qubits on JUPITER with a 90% reduction in physical qubit needs keeps quantum “progress” firmly in the classical HPC domain for now. If you’re in pharma or finance, this is a signal to keep quantum on the horizon—but continue to prioritize GPU-heavy simulation and ML pipelines that work today.
READ THE SOURCE
MORE FROM THE WIRE
Deep & Emerging TechAmazon is spending over $1 billion on a new nuclear energy project to power its next data centers — but will it really live up to the test?
Hyperscalers are now underwriting generation, not just buying power—Amazon backing X-energy’s 960 MW plant is data center capex bleeding directly into nuclear. If you depend on cloud for AI workloads, start treating power availability and price as a first-order vendor risk, not a line item you assume stays flat.
Deep & Emerging TechPalmer Luckey Made a Genetically Modified Military Dog Startup
A $200M check for Warhound’s genetically tuned, armored working dogs shows defense is widening the aperture beyond drones and software into bio-physical hybrids. Dual-use founders should expect more DoD appetite for “unconventional” platforms—and more scrutiny on ethics, export controls, and long-term support obligations.
Deep & Emerging TechSK Hynix Scraps Solidigm US IPO After Shareholder Concerns
Pulling the Solidigm US IPO under shareholder and regulatory pressure shows how sensitive memory and storage assets have become in the capital markets. If your roadmap leans on NAND or SSD cost curves, factor in that corporate restructuring and listings may move slower than the demand story suggests.
Deep & Emerging TechUS Proposes Curbs on Investment in Chinese Biotech Firms
Biotech is getting pulled into the same capital-controls regime as semis—US restrictions on investing in Chinese firms will reshape who funds which parts of the drug pipeline. If your roadmap depends on China-based discovery, CROs, or JV structures, you need a parallel plan for capital and IP flows.