Nvidia’s AI business is through the roof and the end is nowhere in sight | Pocket Gamer.biz
This morning Nvidia confirmedwhat many business analysts have speculated. Turns out that AI and Nvidia’s place at the core of making it happen is no flash in the pan.
Sales of Nvidia’s chips have spiked even more this quarter than at its famous, surprising leap at the last. In May the company’s share price valued the company at $1 trillion for the first time – placing it alongside the likes of Apple, Alphabet, Amazon and Microsoft – now the latest numbers show revenue of $13.5 billion (£10.6bn) for the three months to the end of June. And their next projection takes the figure even higher – they’re shooting for $16bn for the three months to the end of September and instigating a $25 billion buyback program to ensure that they benefit from their own success – literally putting their money where their mouth is.
And on the back of this good news they’re now projecting a revenue amount that’s an unprecedented 170% greater than where they were at last year.
So where did it all go right?
“Twenty-five years ago, I founded NVIDIA with Chris Malachowsky and Curtis Priem to solve the problem of 3D graphics for the PC,” explains the company’s founder and CEO Jensen Huang. ‘Our invention of the GPU in 1999 sparked the growth of the PC gaming market, redefined modern computer graphics, and revolutionized parallel computing.”
Nvidia built its company on graphics chips, first emerging on PC graphic accelerator cards and soon proudly touted as the power behind Microsoft’s original Xbox upon its launch back in 2001. Of course, the power of such chips has increased exponentially since their advent – Nvidia’s PC graphics chips are still arguably second to none – but so has their application for tasks other than plotting vertices and shading surfaces.
More recently graphics chips have been the ideal medium for use in crypto applications with larger and larger processor ‘farms’ featuring many thousands of greedy GPUs being put to work mining bitcoin. But this is time consuming power-hungry work that would always one day hit a wall.
Fortunately the GPU’s true golden age was yet to come.
Really intelligent
Rather than fall out of favour, the advent of artificial intelligence and interest in AI applications seems to have been heaven sent to save the day for Nvidia. Just as PC graphics acceleration becomes both equally high-end niche and built-in commonplace and crypto mining applications begin to no longer make commercial sense, so the desire for AI has swept up all of Nvidia’s available development bandwidth and production capacity.
And this is a line of work that – certainly taking at the rate of growth from 2021 to 2023 as evidence – might only ever be on the upswing…
The secret of the GPU is in its programmability and through that, its ability to carry out multiple calculations at once. While a CPU takes on a task and does it as quickly as possible – there is a single throughput pathway a process takes from inception to completion – GPUs are designed to carry out multiple, often very similar tasks simultaneously. So while CPUs – tasked with everything from drawing on the screen, to spinning a hard drive, to reading the keyboard – rely on ever greater clock speeds to get through it all, so GPUs benefit from comparatively simple but numerous tasks, giving them more threads and pathways to fly through to completion.
They were perfect for the heavy maths involved in graphics processing, sure, but thanks to their programmability – something pioneered at Nvidia – these multi-thread processors could easily be retasked elsewhere. And what better application than replicating the exact same, multiple thread, multiple pathway thought and logic that happens within the human brain? Or more accurately training algorithms with the rapid ingestion of huge data sets and as such enabling decision making that to all intents and purposes could be perceived as human. Thus enabling the kind of AI that was fantasy ten short years ago.
By default or by design these chips have now truly found their vocation.
Maximum power up
So what is the games industry going to do with all this power? Certainly there’s no shortage of takers and with investors to appease and column inches to fill the number of companies actively using AI to better their business shows no sign of abating:
The potential for AI in gaming is vast, touching nearly every aspect of the industryHoma’s CTO Michael Le Bail explains why they’re all in on AI
Room 8 Group announces new partnership with Zibra AI
Three ways where AI can add value to your business
AI startup Inworld raises another $50 million for a $500 million valuation
Using AI to personalise your games,
Even the claim that Big Tech is over… Small Tech is where it’s at
But what about the elephant in the room? Isn’t all this just an attempt to put honest, hardworking artists, writers and coders out of a job? Seems AI has got that covered too.
AI is essentially another creative tool in the toolbox, not a replacement
How to accelerate ASO with ChatGPT
Working AI into your art pipeline: expectations VS reality
Ubisoft clarify the use of AI-script writing Ghostwriter
Making mobile game ads using AI. It’s possible, but is it useful?
Why AI is pushing the games industry forward
These are companies who see AI as a tool for helping them to do what they do best. They’re not cutting costs or pulling plugs, instead they’re making more from what they have and becoming more relevant and more capable in a market that’s always on the hunt for new innovation and excitement.
Somewhere there will inevitably be large firms with large footprints, perhaps pondering the value of that 2019 expansion and wondering when that payday might finally come… But for every one of those there are hundreds that just got powered up and are looking for new talent to help them use it.
And at the end of the day, better games, made faster, by more companies for more players, who love them even more?
What’s not to like?
