NVIDIA cuOpt uses mPDLP to handle decision optimization problems with over 100 million variables
Announced October 7, 2026
What happened
NVIDIA described how cuOpt's mPDLP technique can solve decision optimization problems with more than 100 million variables. It targets optimization problems that keep growing, such as supply chains and power grids with more distributed generation. The excerpt does not give the conditions for its performance comparisons.
Why it matters
GPU acceleration of operations research problems such as large-scale linear programming is reaching practical, production scale.
Sources
More about NVIDIA
-
Microsoft prices RTX Spark-based Surface Laptop Ultra from $2,599 and opens $5,999 Surface RTX Spark Dev Box preorders (report)
Specs and pricing are now public for the Surface Laptop Ultra, which uses NVIDIA's Arm-based RTX Spark chip. It starts at $2,599 with an...
-
NVIDIA shares lessons from teaching a robot to assemble GB300 tester trays
An NVIDIA developer blog post describes how the company trained a robot to assemble GB300 tester trays. It also summarizes lessons...
-
NVIDIA says fine-tuned Nemotron models reached 'gold-level' results at IOI and IMO 2026
In a Hugging Face blog post titled 'One Model Family, Two Gold-Level Results: Fine-Tuning Nemotron for IOI and IMO', NVIDIA said it...
-
Microsoft and NVIDIA unveil RTX Spark-based Surface devices; Windows Copilot gets local file access and OS-wide actions
Microsoft held a Windows and Surface event in San Francisco. On stage, NVIDIA CEO Jensen Huang and Microsoft CEO Satya Nadella said the...
-
NVIDIA releases AICR v1.0, a tool for validating GPU Kubernetes cluster configurations
NVIDIA introduced AICR v1.0 for GPU cluster configuration as an 'open, stable and verifiable' tool. GPU-accelerated Kubernetes clusters...