#cuda

[ follow ]
#nvidia #gpus #artificial-intelligence #ai #ieee-medal-of-honor #gpu #risc-v #gpu-programming
#nvidia
Venture
fromBusiness Insider
4 months ago

If this startup wins, Nvidia's software reign over AI chips could end

Modular is building software to challenge Nvidia's CUDA and enable AI workloads to run across diverse hardware, reducing vendor lock-in.
Artificial intelligence
fromTheregister
4 months ago

Nvidia eyes design, simulation biz with $2B Synopsys stake

Nvidia is investing $2 billion in Synopsys to accelerate GPU-powered simulation, AI tools, and CUDA-X support across semiconductor design, manufacturing, and industrial digital twins.
Artificial intelligence
fromComputerworld
1 week ago

Nvidia's Stephen Jones on the toolkit powering GPUs: 'A wild ride'

Nvidia's CUDA toolkit is foundational for AI advancements and is driving innovations in quantum computing, robotics, and autonomous vehicles.
more#nvidia
Artificial intelligence
fromIT Pro
4 months ago

Nvidia buys $2bn worth of Synopsys in "huge deal"

Nvidia bought $2 billion of Synopsys stock and formed a multi-year partnership to accelerate chip-design workflows using GPU-accelerated computing, CUDA, AI physics, and Omniverse integration.
Startup companies
fromBusiness Insider
5 months ago

This startup created a tool to break Nvidia's chip lock-in. Read the pitch deck Spectral Compute used to raise $6 million.

Spectral Compute raised $6 million to build SCALE, a framework that enables CUDA-written applications to run on non-Nvidia GPUs, initially supporting some AMD architectures.
Software development
fromWIRED
6 months ago

A Former Apple Luminary Sets Out to Create the Ultimate GPU Software

Modular provides portable GPU optimization software that can outperform vendor tools, but faces adoption, competition from Nvidia/AMD, and customer concerns over added costs.
#gpu-programming
more#gpu-programming
Artificial intelligence
fromHackernoon
3 years ago

This New Language Might Kill NVIDIA's GPU Monopoly | HackerNoon

The shift from CUDA to heterogeneous computing signals a revolution in software design and deployment, unlocking new potentials in performance and efficiency.
Java
fromInfoQ
10 months ago

Bringing GPU-Level Performance to Enterprise Java: A Practical Guide to CUDA Integration

Java can integrate with CUDA to achieve significant performance gains in workloads.
JNI allows developers to connect Java with native CUDA for high-performance tasks.
Understanding concurrency vs multithreading is essential for utilizing CUDA effectively.
GPU computing offers benefits beyond AI, applicable to secure data processing and more.
[ Load more ]