Random access
ARM and the impending AI evolution
Jensen Huang, the Founder and CEO of Nvidia, framed the recent announcement of the company’s purchase of ARM Limited as something that will make Nvidia the “premiere computing company for the age of artificial intelligence”. At US$40 billion dollars this transaction might seem like a hefty price for something that’ll help the Google Assistant to realise you were actually talking to your dog. We kid, ARM Limited’s processor manufacturing business alone is probably worth that price tag since you’ll likely have some of its cores sitting in the smartphone you use every day, but Huang is right, Nvidia stands to make its investment back in a matter of years thanks to the expansion of the AI market.
We like to cover the quaint progression of AI into fields like writing, medicine and surveillance, but AI is already the backbone of companies like Facebook. Machine learning is the only thing capable of delivering a customised selection of targeted advertising for each individual, a tool that allows Facebook to maintain the interest of its users and sell guaranteed impressions and conversions to the highest bidder.
For the uninitiated, Facebook’s annual revenue was US$70 billion dollars in 2019 and AI is the driving force behind the company’s 11 percent year-on-year growth, so Nvidia would be well placed for an investment like this, even if Facebook was its only client. When you consider that since 2016 it’s also been clear just how critical Facebook’s AI has been in reshaping democracy by influencing the hearts and minds of selected swing voters for profit, Nvidia’s positioning looks even smarter. So if Facebook and its AI tools are the modern kingmakers, we expect Nvidia backed programs like GPU-Accelerated AI shopping and politically funded election campaigns to pay off for the GPU manufacturer in a big way.
If that wasn’t enough AI is also a major area of investment from businesses like Amazon Web Services and the Google Cloud Platform, which host AI server farms large enough to accommodate other businesses that experiment with machine learning applications. Again this might seem quaint, but add up the integration of AI in everything from coronavirus vaccine development to self driving cars and it’s clear that the power of AI will underpin all of the next major tech breakthroughs... or at least the most significant ones.
There are some obvious synergies between ARM and Nvidia when it comes to AI at both the macro and micro scales. We joked about smartphone AI earlier, but dedicated AI processors have been baked into flagship phone SoCs for the last couple of years now, a feature that has has powered major software advancements in voice recognition, real time image processing and photography enhancements.
Looking bigger, ARM servers for AI have been gaining a lot of recognition in both edge computing and the IoT space due to their size, cost and efficiency benefits, a direction that Huang pointed to directly in a statement about the acquisition. “In the years ahead, trillions of computers running AI will create a new internet-of-things that is thousands of times larger than today’s internet-of-people.” Every smart device or AI driven surveillance camera will need hardware, so Nvidia will have a healthy market for its ARM devices in the future… even if smartphone-sales have already peaked.
“If Facebook and its AI tools are the modern kingmakers, we expect Nvidia backed programs like GPU-Accelerated AI shopping and politically funded election campaigns to pay off for the GPU manufacturer in a big way.”