EDACafe Industry Predictions for 2022 – Menta

Yoan Dupret

Yoan is the Managing Director and CTO of Menta – a leader in embedded FPGA IP cores for chips and smart sensors. Prior to his position at Menta, he held various managerial and technical positions at DelfMEMS, Samsung, CSR, Infineon and Altis Semiconductor. Yoan holds a PhD from Supelec (France) and an Engineering degree (MSEE) from ESEO (France).

‘Enabling Adaptive and Resilient Compute at The Edge and Extreme Edge’

Two years on, the world Is still experiencing the aftershocks of Covid-19 – but despite the human tragedy it continues to be, without doubt science and technology helped us cope with it better. Imagine how much worse it would have been 30 years ago.

At the same time, east-west geopolitical tensions are growing – with technology as its core battlefield. So, both at a human level and political level, there is heightened awareness of science and technology as a key differentiator for nations and enabler for a better life.

Just look at the chip shortage we are experiencing that has an impact on everyday life – even affecting the availability of new cars.

Of course, such a situation presents both risks and opportunities for the semiconductor industry, but also a duty to prepare the future. The most important lesson of the time we are living in, is that we must be prepared for uncertainties and rapid changes as an industry – whatever the cause. We need to ensure that our industry becomes more resilient.

Last year, we looked at “Disruptive Adaptive Compute at The Edge”. We explained how the adaptive compute that is prevalent at Cloud and Data centers will need to move to Edge Compute – which means on one hand low power adaptive accelerators and on the other hand low cost, fast time to market chips production through the use of chiplets. We will explore this year, how Menta is Enabling the Adaptive and Resilient Compute at The Edge but also at the Extreme Edge.

Before we delve into it though, if the world’s chips or chiplets already had our eFPGA technology embedded in them, it would have mitigated the impact of the chip shortages and would have allowed the industry to re-purpose some of its inventory in an adaptive and resilient way for the most pressing needs – there is a very important lesson to learn from this saga and puts an additional meaning on the word ‘Adaptive’ by bringing in the concept of resilience.

According to Wikipedia, Edge computing is a distributed computing paradigm that brings computation and data storage closer to the sources of data. This definition is sufficiently broad to cover devices anywhere from local servers or 5G base stations to a smart locks or VR headsets. We will call the later kind Extreme Edge – which means battery powered devices.

Edge computing, and moreover Extreme Edge computing comes with huge constraints on power consumption, size and, most of all, cost. It also comes today with the requirements of adaptive compute acceleration.

The most energy efficient way for a device to adapt to future algorithmic changes (or to re-purpose an existing chip) is to allow an adaptation at hardware level. This is where embedded FPGA IPs come at play. Embedded FPGAs bring all the benefits of FPGAs – as the hardware can be recoded to perform any digital function, post-production and an infinite number of times – but with all the ASIC/SoC benefits in terms of power consumption and high-volume cost. This is what makes it resilient.

To accelerate the adoption of eFPGA IPs and therefore facilitate adaptive and resilient computing at the edge and extreme edge, Menta has revolutionized the eFPGA IPs by announcing in December the availability of its eFPGA IP as a soft IP. That means that the IP can be provided on any Foundry and technology node, fitted to the product requirements in terms of power, performances and area and completely handled by our customers and partners.

Menta has also built a self-contained easy to use software, that can be integrated as an API into customers SDK to further facilitate the usage of the eFPGA IP adaptive acceleration capabilities by anyone who has the commercial device embedded the eFPGA IP in hands.

We are on a mission to facilitate Adaptive and Resilient Compute at The Edge and Extreme Edge and allow our customers to navigate through uncertainties and rapid changes – be it for algorithmic changes or the need to re-purpose existing inventory because of shortages.

About Author:

Yoan Dupret

Yoan is the Managing Director and CTO of Menta – a leader in embedded FPGA IP cores for chips and smart sensors. Prior to his position at Menta, he held various managerial and technical positions at DelfMEMS, Samsung, CSR, Infineon and Altis Semiconductor. Yoan holds a PhD from Supelec (France) and an Engineering degree (MSEE) from ESEO (France).

Category: Predictions

This entry was posted on Tuesday, January 25th, 2022 at 6:02 am. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.