AI has become the philosopher’s stone that Big Tech wants to achieve, so all resources are considered necessary to achieve the goal. In this sense, amd has announced the acquisition of the AI startup Noad.ai for enhance your AI open software capabilities and compete strongly with Nvidia.
In this way, it will be possible to have an experienced team that has already developed efficient software technology before in AI solutions for data center accelerators AMD Instinctprocessors AI Ryzenprocessors EPYC, SoC Versal and Radeon GPU to AMD.
The annexation of Noad.ai to AMD represents a fierce attempt to snatch Nvidia’s leadership of more than a decade in the AI chip market and the ecosystem of software developers. The AMD group that will annex Noad.ai is made up of about 1,500 software engineersthe forecast being the expansion of the team in 300 more hires for the remainder of the year and for 2024.
This agreement is in line with AMD’s AI growth strategy, focused on an open software ecosystem that reduces barriers to entry for customers through tools, libraries and developer models. This is how you can invest and create a unified collection of software to boost the chips the company manufactures, with the invaluable help of internal investments and external acquisitions.
It should be noted that Noad.ai has played a crucial role in maintaining the world’s most important AI repositories, such as the generation of SHARK code, Torch-MLIR and OpenXLA/IREE. They are AI solutions optimized for leading hyperscalers and startups.
The automation software capabilities, based on Nod.ai’s SHARK software compiler, greatly reduce the need for manual optimization and time to deploy high-performance AI models across a broad portfolio of data center, edge and client platforms powered by AMD CDNA, XDNA, RDNA and ‘Zen’ architectures.
AMD has a wide catalog of high-performance computing products. Your hardware is present in Frontier, the most powerful supercomputer in the world, and in many data centers intended for training AI models. However, today, its models fail to achieve the success achieved by Nvidia.
The economic keys of the agreement
So far, the economic terms of said agreement are unknown, although it should be noted that Noad.ai has raised some 36.5 million dollars in its last announced balance sheet. For its part, Nvidia has reached $1 trillion in market capitalization and has sold more AI chips than any other competitor.
Vamsi Boppana, senior vice president of the AI Group at AMD, assures that with Noad.ai we improve the ability to offer customers high-performance open software adapted to their hardware. Besides, open source compiler technology is improved and wearable AI solutions are enabled across its product portfolio.
Anush Elangovanco-founder and CEO of Noad.ai, assures that, once they are already present in the cloud and on a wide range of devices, this agreement will allow them to advance at the pace of an industry in constant evolution to solve problems together.
The latest developments from AMD
We recently learned that AMD has incorporated a tool to measure latency in its latest beta drivers. They are the first to integrate the possibility of activating the AMD Fluid Motion Frames (AFMF) or interpolated frames in games DirectX 11 and 12 through the control panel of the graphics card.
These drivers provided support for activating AFMF with the graphics AMD Radeon RX 7000 based on the RDNA 3 architecture, the subsequent update being to add support for the Radeon RX 6000. Through the overlay a metric called Sustem Lag which reports on the delay between an image being created by the game’s graphics engine and its subsequent rendering on the graphics card.
It will work with Radeon RX 7000 that are responsible for supporting the technique to reduce latency in another function that AMD classified as Anti-Lag+ and that will become the direct competition of NVIDIA Reflex. However, for the moment, it works on a small list of games.
