NVIDIA Ising and quantum push

A quantum computer stores information in fragile quantum states, and tiny hardware drift can knock calculations off course. NVIDIA said on April 14 it is releasing an open model family called Ising to automate that tuning and speed up error correction. (investor.nvidia.com) NVIDIA said Ising starts with two jobs that now consume large amounts of lab time: calibration, which is the repeated retuning of a quantum processor, and decoding, which is the real-time cleanup step used in quantum error correction. The company said its calibration model can cut tuning from days to hours, while its decoding models run up to 2.5 times faster and 3 times more accurately than traditional approaches. (nvidia.com) The company described Ising as open source and said it is shipping pre-trained models, data, and retraining tools through its quantum software stack. NVIDIA’s product page says the models are meant for hybrid systems that pair quantum hardware with graphics processing units, or GPUs, to handle the classical computing work around the quantum chip. (nvidia.com) The Ising name comes from a physics model used to describe many interacting spins, a simplified way to represent systems with many moving parts. NVIDIA is using that label for software that tries to tame another messy system: quantum machines that need constant measurement, adjustment, and correction to stay usable. (youtube.com) NVIDIA has been building that hybrid pitch for more than a year. In May 2025 it said Japan’s ABCI-Q would use NVIDIA hardware for a quantum-focused supercomputer, and in November 2025 it introduced NVQLink as an interconnect for linking quantum processors to accelerated computing systems. (investor.nvidia.com 1) (investor.nvidia.com 2) The other half of the quantum story sits outside the lab, in encryption. A large enough quantum computer could break parts of today’s public-key cryptography, and the National Institute of Standards and Technology approved its first three post-quantum cryptography standards, Federal Information Processing Standards 203, 204, and 205, in August 2024. (csrc.nist.gov) Those standards cover a key-establishment method called Module-Lattice-Based Key-Encapsulation Mechanism, or ML-KEM, and two digital signature schemes, Module-Lattice-Based Digital Signature Algorithm, or ML-DSA, and Stateless Hash-Based Digital Signature Algorithm, or SLH-DSA. NIST says those three standards “can and should be put into use now” and will provide the foundation for most deployments. (csrc.nist.gov 1) (csrc.nist.gov 2) (csrc.nist.gov 3) (csrc.nist.gov 4) Google added a date to that pressure on March 25, 2026, saying it is setting a 2029 timeline for its own post-quantum cryptography migration. Google said the target reflects migration needs in light of progress in quantum hardware, error correction, and factoring estimates. (blog.google) NIST’s draft transition report says organizations need to identify quantum-vulnerable standards and move products and services toward quantum-resistant replacements. That turns “quantum readiness” into inventory work, software updates, certificate changes, and procurement deadlines rather than a distant research project. (csrc.nist.gov) NVIDIA’s announcement and the cryptography deadlines are aimed at different problems, but they land on the same date for a reason. One side is trying to make quantum computers more practical; the other is trying to make sure existing systems are protected before those machines become powerful enough to matter. (investor.nvidia.com) (blog.google)