Making Kynisys: how we’re building the future of AI

Share on linkedin
Share on twitter
Share on facebook
Share on reddit
Share on digg
Share on email

I joined Imagination Technologies as the executive entrepreneur in residence in June 2019. The company is world-renowned for its GPU IP and more recently neural network accelerators and ray tracing and I knew I could develop a new idea to build on that IP.

This is the sort of tech which requires half a lifetime of specialist knowledge to really get a deep understanding of it. Shortly after joining I felt a bit like I had landed on a more advanced alien planet. Everyone seemed smarter than me.

After the initial learning-shock, I started to see how edge AI and inference on the edge is most likely going to be the future. The benefits of edge AI are about to make the world smarter: it provides speed, efficient data handling and storage, lower systemic costs and no real-time live connection.

Yep, I was sold.

Although edge technology is going to change the world, it is not the most straightforward to deploy. Senior developers and product specialists described to me their current workflow and pain points and I left horrified. Today’s workflow and solutions are far too complicated, perhaps even unknowable. It became clear to me that if you’re serious about deployment on the edge it would require an AI engineering team as big and expert as Imagination Technologies.

As these conversations progressed and the multiple product workshops we ran, I could see a value proposition emerging: An AI marketplace for edge compute that offered toolsets and all the building blocks that an edge AI system integrator needs. An easy-to-use platform to create, test and deploy edge AI in less than a day. Putting the power in the hands of those who have to solve “edge problems”.

Kynisys slide

AI on the edge is about solving a problem in a constrained problem space, whether that is cost, device size or low power, and understanding data at the point of seeing it. This allows us to use the intelligence provided by AI to seamlessly integrate into the world we live, responding instantly and reliably while maintaining the integrity and privacy of all of our data.

I always wanted to build a platform that developers would love, that gives them real power, and that makes things better and easier. I want to make it easier for start-up CEOs to be bold and take risks so that they can build those awesome ML-based businesses in AIoT or Robotics without huge Capex. I want to help the global players get faster and be more product focussed and market-led. Fewer pains and more gains.

So Kynisys is born. We’re already signing up our first alpha test partners. We’re launching competitions for talented developers in academia and in elite development groups to build out solutions on our platform. Our message is: build the future with us. Get moving. Get kinetic. Get Kynisys.

Please leave a comment below

Comment policy: We love comments and appreciate the time that readers spend to share ideas and give feedback. However, all comments are manually moderated and those deemed to be spam or solely promotional will be deleted. We respect your privacy and will not publish your personal details.

Blog Contact

If you have any enquiries regarding any of our blog posts, please contact:

United Kingdom
Tel: +44 (0)1923 260 511

Search by Tag

Search by Author

Related blog articles

ai edge

Reduced Operation Set Computing (ROSC) for Flexible, Future-Proof, High-Performance Inference

Designers of neural network accelerator (NNA) IP have a Herculean task on their hands: making sure that their product is sufficiently general to apply to a very wide range of current and future applications, whilst guaranteeing high performance. In the mobile, automotive, data centre and embedded spaces targeted by Imagination’s cutting-edge IMG Series4 NNAs, there even more stringent constraints on bandwidth, area and power consumption. The engineers at Imagination have found innovative ways to address these daunting challenges and deliver ultra-high-performance and future-proof IP.

Read More »
a series gpu performance concept

Taking the HyperLane with IMG A-Series

Inside every A-Series, from the smallest to the very largest are eight individual hardware control lanes, each isolated in memory, enabling different tasks to be submitted to the GPU simultaneously, for fully secure GPU multitasking. We call it HyperLane Technology.

Read More »
Samsung Ballie CES 2020 woman

CES 2020: The robots are coming…

Coming away from CES this year one thought filled my head: the robots are coming! Having walked around the floor of this enormous show I saw enough to convince me that the 1950s vision of the robot in your house ready to do the chores isn’t quite as far-fetched as it once was. That’s not to say we’ll have C3PO or Kryten available to order next year of course; things will start a little smaller and rounder.

Read More »


Sign up to receive the latest news and product updates from Imagination straight to your inbox.