Modulr Challenges NVIDIA’s GPU Dominance with New Software Development

Startup Modulr is taking on tech giant NVIDIA by developing an alternative to NVIDIA’s flagship software, CUDA. This move signifies a shift in the GPU market competition, transitioning from hardware to software.

Recent reports suggest that Modulr is in crucial negotiations to secure an investment of $600 million (approximately 800 billion won) for this venture. Despite being a relatively new player in the market, Modulr has made significant strides, having launched its main product just three months ago.

Notably, Chris Lattner, known for his work on Apple’s ‘Swift’ programming language, and Tim Davis from Google Cloud’s ML application team are key figures at Modulr.

The ambitious goal of the startup is to provide a flexible alternative to NVIDIA’s GPU and CUDA software, aiming to reduce the industry’s reliance on NVIDIA. This initiative has already garnered substantial attention.

Currently, NVIDIA dominates the GPU market, holding an impressive share of over 80%. Their CUDA software, pivotal in parallel processing on GPUs, has been a significant contributor to this dominance.

While CUDA is offered for free, it is designed exclusively for NVIDIA chips, giving the company a strategic advantage. However, Modulr envisions a more inclusive software that supports chips from various manufacturers, including AMD, Intel, and Google. Their proposed software is expected to integrate seamlessly with popular AI frameworks like PyTorch and TensorFlow.

Industry analyst Holger Mueller commented on this development, stating, “Challenging the AI behemoth NVIDIA is no easy task, but the emergence of competitors like Modulr is a positive sign for the industry.”

Modulr’s challenge presents a new paradigm in the GPU market, heralding the future of an industry that seeks technological diversity and innovation. This move against NVIDIA’s monopolistic control indicates a direction where various companies collaborate to build a better technological ecosystem. The GPU market is now poised to transition from mere competition to a stage of collaboration and innovation.