- 06 June 2025
- Imagination Technologies
As artificial intelligence continues to reshape industries and redefine the future, one thing is clear: none of it is going to happen without the right hardware. At the heart of the AI revolution are GPUs (graphics processing units), whose ability to handle massively parallel workloads – like those that AI typically generates - is unmatched.
In a special AI episode of Tech Impact, our very own Chief Product Officer, James Chapman, sat down with Evan Kirstel to discuss what AI needs to reach its potential.
You can read the full transcript below.
Evan: We all know that AI is a hot topic right now but underpinning this is the need forever increasing amounts of hardware to run this AI. GPUs, or graphics processing units, have been one of the biggest beneficiaries of this demand. So, let's talk to Imagination Technologies, a British company that has been developing GPU solutions for a long time, and meet their Chief Product Officer, James Chapman.
Hi James. Welcome to this special AI episode of Tech Impact.
James: Hi Evan, it's great to be here. I'm very excited to be talking with you today about Imagination Technologies and how we're putting AI in multiple devices.
Evan: So, James, tell us a little bit about Imagination Technologies and what you guys do.
James: We're one of a handful of companies that make graphics processing units, otherwise known as GPUs. We're headquartered in the UK, and we're privileged to have more than 500 engineers worldwide making GPUs. Now, we actually put our GPU in the very first smartphone, and today you'll be using our GPUs for UIs (user interfaces), for gaming and for AI processing across a wide range of devices: PCs, phones, TVs, even cars.
Evan: AI is shaping our future, and there is a lot of noise in the press about the demand for computer chips to run AI. What's your take on what's happening in this space?
James: AI is all about parallel computing. Now, traditionally we've done computing in what we call a sequential manner. That's just doing things step-by-step, one at a time. Parallel computing first came about when we tried to drive graphics and images on screens. We needed to do things at the same time to colour all of those pixels in. Now it turns out that that very same parallel platform is exactly what we need to do AI. That's why it's become the platform of choice in the data centre.
Here at Imagination, we're bringing that parallel compute platform, that GPU we talked about earlier, to a wide range of devices, enabling AI everywhere.
Evan: You said there were only a handful of GPU players. What is it that makes Imagination’s GPU different to the other players?
James: It's all about doing these things at low power. So, if you think about where we want to put AI, many of those devices are run off batteries - from your smartphone, even your electric car! If you waste power doing AI, that's power you haven't got to do the other things you need to do. So low power is critical. It's also actually critical in the data centre, where some of these data centres need to come with almost a power station attached, they're using so much power. So low power is the thing that allows AI to move forwards and at Imagination we deliver AI with super low power.
Evan: There's a lot of custom AI hardware out there with everyone claiming to have a better way of doing AI. What do you make of this?
James: You're absolutely right, there is a lot of specialised hardware out there that does do a specific task very well. The problem with it is it's specialised, so it's not flexible and it's not programmable. And when we look at the AI world, it's incredibly fast moving. We live in very exciting times, Evan.
There's a lot of innovation. Every day something new turns up. The problem is you can't run that on yesterday's hardware. You need to run it on a platform that's forward-looking, flexible and programmable. That is a general-purpose parallel compute problem. That is the GPU. That is what Imagination builds and supplies.
Evan: Where is this going in the future?
James: Well, we're only at the start of the AI journey. It's going everywhere, isn't it? It's going to all devices, it's going to all fields, all different industries. Where it's going, though, is pervasive - AI everywhere. And that requires low power. It also actually requires you to do these things in a very safe fashion. All of those we deliver with Imaginations GPUs with our parallel processing platform. So, here at Imagination, we are very much about delivering AI everywhere in a low power, safe format that people can trust, and people can enjoy.
Evan: What impressive work, James, thanks for joining us on this special AI episode of Tech Impact.
James: It's been an absolute pleasure, Evan, thank you.
For more information on Imagination, and our efforts in GPU and AI, feel free to check out the following pages and resources:
