FICO releases GPU-accelerated optimization software with NVIDIA partnership

FICO (NYSE: FICO) released FICO Xpress 9.8, an optimization software that integrates NVIDIA CUDA-X libraries to accelerate computational problem-solving. The software features a GPU-accelerated implementation of the hybrid gradient algorithm that can deliver up to 50x speedups for large optimization problems.

The hybrid gradient algorithm targets extremely large, dense problems with tens or hundreds of millions of non-zeros in the problem matrix. NVIDIA CUDA-X provides GPU-accelerated libraries and frameworks that deliver higher performance than CPU alternatives across data processing and high-performance computing applications.

"The GPU-accelerated hybrid gradient algorithm represents an important addition to Xpress's comprehensive suite of optimization algorithms, enabling organizations to solve large-scale problems faster," said Bill Waid, chief product and technology officer at FICO.

The GPU acceleration feature operates as a beta function in FICO Xpress 9.8 and supports any NVIDIA GPU with CUDA Compute Capability 7.5 or later. The software requires compatible hardware to achieve the reported performance improvements.

FICO Xpress 9.8 includes additional performance enhancements beyond GPU acceleration. The mixed-integer programming engine shows 14% faster overall performance compared to the previous version and 24% faster performance on models requiring at least 100 seconds. The global solver demonstrates 68% faster overall performance and 5.3x faster performance for models taking at least 100 seconds.

FICO offers a 60-day trial license for the FICO Xpress Optimization Suite. The company provides optimization software that enables businesses to deploy analytic models into applications powered by optimization engines for decision-making processes.

The announcement comes from a company press release statement detailing the software release and partnership with NVIDIA for GPU acceleration technology.

You May Also Be Interested In