ARM chips power a billion devices, from smartphones to the world's fastest Japanese supercomputer, Fugaku . NVIDIA's takeover of ARM has yet to be completed. While the company continues to expand its range of proprietary tools based on the ARM ecosystem. NVIDIA is bringing new capabilities to 13 million developers supporting systems based on this architecture.
New opportunities
NVIDIA AI is the industry standard for accelerating artificial intelligence learning.
AI has led to many advances, including intelligent voice assistants, Google DeepMind's AlphaGo, and the development of unmanned vehicles. NVIDIA deep learning frameworks can reduce training time from days to hours.
NVIDIA RAPIDS is a set of software libraries for working with data and analytics on GPUs.
The tools help you manage massive amounts of data using high-speed GPU computing, parallelizing data loads 50 times faster.
NVIDIA HPC SDK - Compilers, libraries, and software tools for high performance computing.
A set of tools is essential to improve developer productivity. They are used to create and optimize applications based on NVIDIA, x86-64, ARM or OpenPOWER GPUs. You can use pluggable libraries, C ++ 17 parallel algorithms, and OpenAcc directives to speed up your code on the GPU.
NVIDIA RTX - Graphics drivers that provide ray tracing and AI capabilities.
NVIDIA claims it is the only solution in the world with dedicated RT cores for ray tracing and tensor cores for processing AI algorithms for super speeds in PC gaming.
NVIDIA uses the ARM architecture in SoCJetson processors and BlueField data processors.
Resources will be donated to ARM, which will provide an opportunity to expand the offer to partners and customers in four main categories.
- HPC - high performance systems.
- Cloud / Data Center - cloud / data center.
- Edge AI / Robotics - Edge AI / Robotics.
- PC.
Cloud gaming
NVIDIA has partnered with ARM and OEM partners in a variety of areas, including cloud gaming.
For example, NVIDIA engineers are porting ARM code and making new tools to optimize encoding and streaming games to cloud servers. Working with its own GeForce NOW service, the company is well aware of the potential of this approach. Apart from gaming, ARM-based servers are used for machine learning, data storage, and other applications. The NVIDIA Container Toolkit is used to run Docker containers on ARM Kubernetes.
As a reminder, the deal between SoftBank and NVIDIA has been announcedSeptember 13, 2020. According to the agreement, the American company will become the new owner of the ARM processor developer. The deal will help NVIDIA outperform companies like AMD, Intel and other competitors. For the final conclusion of the transaction, the parties need to obtain permission from regulators in the UK, China, EU and USA.
