>Business >​EdgeQ obtains half a billion in funding to combine 5G and AI compute on a singular chip

​EdgeQ obtains half a billion in funding to combine 5G and AI compute on a singular chip

AI, programmable, Open, Edge, and 5G – these are the pillars, the mix of technologies that forecasted to become vital for the future of Internet-of-things (IoT) in a variety of fields, as diverse as security and shipping/logistics. A method to combine all of these technologies onto a singular device is merely to agglomerate a bunch of off-the-shelf silicon chips and mesh them into one product. For instance, take up a wireless radio chip, add some computational capability, and so AI chip magic and presto, you have just made for yourself a sophisticated IoT device.  

There’s one issue, however, these devices typically have a myriad of limitations and restrictions. Most noteworthy is the fact that they have power constraints, which implies that they typically cannot integrate five chips on a single board, but instead require one singular chip that can leverage power in a very efficient fashion. Also, there are dimensional limitations, and constraints surround how simple a chip is to reprogram remotely. In summary, there’s an observation to be had that this new market – which is just beginning but forecasted to be dominant in the not-too-distant future, requires a focused chip for it. 

Enter EdgeQ. EdgeQ is a stealthy silicon startup, emerged from stealth today to demonstrate some of its attempts surrounding its new chip, and also to make the announcement regarding the closure of its 38.5$ million Series A round spearheaded by Mohammad Islam of Threshold Ventures (previously known as DFJ). The enterprise prior shut down a seed round spearheaded by Homan Yuen at Fusion Fund, taking the organization’s cumulative fundraising to $51 million. AME Cloud Ventures also took part in the round, and so did an unnamed strategic customer. 

What’s thrilling here in this early phase is the team. EdgeQ was started by Vinay Ravuri, an ex-Qualcomm executive who was employed on data and mobile centre projects at the chipmaking giant at the time of the corporate takeover attempt by Broadcom, way back in 2018. He exited the company that year and collaborated with a unit of other senior engineers and executives from Qualcomm, Broadcom and Intel to work on the next-gen of chip technologies for the edge. We felt that there was an avenue to concentrate outside of mobile phones, Ravuri stated, which is what Qualcomm had mad major investments in provided its de-facto dominance in 4G technologies. 

EdgeQ is still under development, and the organization isn’t willing to disclose the comprehensive details regarding its product as it carries out work on an ongoing basis. Regardless, Ravuri provided a wide overview of EdgeQ’s chip, in its current status and form. 

I love to leverage this analogy of like a Goldilocks, he stated. You know, this is too cold and this too hot and what you actually require is someplace in between. And that’s where EdgeQ slots in. 

This systems-on-a-chip product will possess 5G connectivity with the on-chip processing of a conventional artificial intelligence chip, which has undergone compression under a tight power envelope to reduce energy utilization. Even more critically, the chips will feature reprogrammability, facilitating alterations on the fly as the scenario surrounding a product changes. Inputting all of this into a singular chip needs a lot of integration and tons of architectural innovation, which is what EdgeQ has been capable of doing.  

This model of systems-on-a-chip is a lot like what you might’ve observed previous week at Apple’s grand event surrounding its new M1 chip. By bringing together a bunch of individual silicon systems on a singular chip, you can enhance capacities while minimizing power. 

What could EdgeQ’s chips ultimately do? Ravuri indicated a variety of applications. You could visualize a security camera possessing a chip such as this to coordinate with other cameras while also undertaking process of some of the data that it is observing through its lens. Robotic machines and drones in an enclosed environment could leverage wireless to sync their actions with other devices, undertake processing of their actions via compute and additionally calculate their locations to an extreme degree of accuracy. And, obviously, an entire bunch of additional applications are at least feasible, but we’ll get more information as the chip technology evolves and matures.  

Homan Yuen of Fusion Fund stated that as his background was in semiconductors, he was aware how tough these organizations were, and why venture capitalists didn’t like doing them. Upon meeting up with Vinay, the team, learning of their background and experience, he thought if there was one organization he was going to bet on doing chips, it’s bound to be them. 

It’s a proposition that gained a lot more traction with new investors coming to the dance, but obviously, product development has to persist, and these chips basically have to be produced and actually put in the market for consumers to buy. To put it in other words, it’s a nice landmark, but there’s a ton of more work to be undertaken. 

EdgeQ has offices in San Diego, Santa Clara, and also in Bangalore, India.  

Add Comment