AMD to acquire industry-leading software technology in a bid against rival chipmaker NVIDIA. The experienced team at Nod.ai, a major contributor to the world’s most important AI repositories, is set to bolster the company’s open AI software capabilities. What does the Nod.ai acquisition mean for the AI industry?
What is Nod.ai?
Nod AI is an AI software startup based in Santa Clara, CA, specializing in open-source AI solutions. The end result of the firms research and development are contributions to open-source projects Torch-MLIR, IREE, and SHARK software – “Portable, Performant ML”. SHARK is built on “LLVM, MLIR, OpenXLA’s IREE and nod.ai’s tuning”. The former two share much in common, with developers using IREE to translate neural network source code into an executable program, which can then run on graphics cards or other AI-optimized chips. It’s clear why AMD’s growth strategy includes such a major contributor to the open software ecosystem.
This early-stage Artificial Intelligence and Machine Learning systems company is set to give AMD a vital edge with regards to “open source technologies for future A.I systems using advanced compiler based approaches (instead of legacy handwritten kernels).”
Essential AI Tools
Content Guardian – AI Content Checker – One-click, Eight Checks
Jasper AI
WordAI
Copy.ai
Writesonic
Who owns Nod.ai?
Post-acquisition, Nod,ai will be owned by AMD. The founding team currently stands as following:
- Anush Elangovan: “Founder and CEO worked on the first Google ARM Chromebook and was part of the Agnilux team that became the foundation of the ChromeOS team at Google.”
- Harsh Menon: “Co-founder and CTO was an early employee at Zee.Aero / KittyHawk and worked on flying cars. He has an MS in Aerospace Engineering and Computer Science from Stanford University.”
Nod.ai acquired by AMD
In a bid to bolster the software capabilities of Radeon GPUs, chipmaker AMD have enlisted the expertise of a team of engineers in California. The two firms will collaborate (with the smaller as a subsidiary) on the deployment of AI solutions for AMD’s CPUs (Zen), GPUs (RDNA), and FPGAs (XDNA), harnessing Shark compiler-based automation software and AI models.
Vamsi Boppana is senior vice president of the Artificial Intelligence Group at AMD.
“We have been growing rapidly, with plans for next year as well,” says Boppana, when asked about recent AI acquisitions. With a team of over 1,500 engineers, and growing, the vision to become #1 is clear. “The acquisition of Nod.ai is expected to significantly enhance our ability to provide AI customers with open software that allows them to easily deploy highly performant AI models tuned for AMD hardware,” continues the AMD SVP. “The addition of the talented Nod.ai team accelerates our ability to advance open-source compiler technology and enable portable, high-performance AI solutions across the AMD product portfolio. Nod.ai’s technologies are already widely deployed in the cloud, at the edge, and across a broad range of endpoint devices today.”
Does AMD have AI?
The very fact that people are asking this question shows how the chip manufacturer is falling behind. In 2023, everyone needs AI – whether that’s for genuine reasons or just to secure investment, because lord knows investors wont shut up about it.
Yes, AMD does have AI solutions, including AMD Ryzen AI hardware and AMD Instinct AI-acceleration servers. This includes Ryzen AI processors, EPYC processors, Versal SoCs and Radeon GPUs, such as MI200 & MI100 series accelerators.
Accelerate your discovery on the AMD ROCm™ open software platform with AMD Instinct™ accelerator-powered servers.
AMD
It’s only fair to note AMD’s historical contributions to large language models (LLM) and generative adversarial networks (GAN).
AMD powers some of the world’s fastest supercomputers, including EuroHPC’s LUMI supercomputer and Oak Ridge National Laboratory’s Frontier system – the first to break the exascale barrier. AMD exascale-class compute technologies are uniquely well suited to deliver the processing power needed by even the most complex Generative AI models.
AMD
The hardware manufacturer aims to make AI “pervasive”, spanning from “cloud to the edge to the endpoints”. It could be clearer what the firm means by that statement, but its 3 key areas of focus do help to clarify:
- Products: “Delivering a broad portfolio of high-performance and adaptive computing engines that make AI possible.”
- Architecture: “Building these products on cutting-edge compute memory dataflow architectures that accelerate AI processing.”
- Open ecosystem: “Enabling proven, open solutions with ecosystem partners to solve the world’s toughest challenges.”
Is AMD or NVIDIA better for AI?
As it stands, NVIDIA is better for AI.
Of course, the Advanced Micro Devices firm aims to change that – but in all honesty, that’s going to take a lot of effort. NVIDIA is not only months (and billions of dollars) ahead, but has contractually and technologically cemented itself into the infrastructure of every other big tech firm currently investing in the AI market.
Arguably, NVIDIA is years ahead if we consider that the hardware infrastructure of NVIDIA GPU’s has proven, quite serendipitously, to be ideal for AI acceleration in addition to their intended purpose. To be clear, AMD is not inexperienced, having contributed to one of the worlds most significant AI supercomputers in the Oak Ridge National Laboratory’s Frontier system – but in terms of consumer technology, the winner is clear.
NVIDIA have met demand head-on, unveiling the H200 “Grace Hopper” AI superchip on August 8th, 2023. By comparison, AMD have had no high-profile AI hardware announcements this year.
Don’t take my word for it, the proof is in the pudding, with Founder & CEO Jen-Hsun “Jensen” Huang leading NVIDIA to an all-time market capitalization high of $1.13 trillion USD. Earlier this month, NVIDIA shares prices hit approximately $493.55 USD, compared to roughly $136 USD just one year prior. Only 6 companies in the world are currently valued at $1 trillion USD or higher, as of August 2023, including Apple, Microsoft, Saudi Aramco, Alphabet, Amazon, and NVIDIA.
