Revolutionizing with Menta’s eFPGA
Menta eFPGA IPs allow efficient implementation of Neural Network algorithms, e.g. for ADAS vision applications - thanks to large bandwidth and very low latency. The deterministic nature of the eFPGA IPs also makes them great candidates for situation awareness and decision-making algorithms.
Motor controls require simple but fast processing, often based on filters and Fourier transforms. These require deterministic systems that can be updated to take into account changes in environment, aging of components, etc. Menta eFPGA IPs are deterministic by nature and offer the optional DSP that is targeting, but not limited to, fast, and low-power FFTs.
Today, raw data from sensors is filtered before being sent to the central processing units due to bandwidth limitations, mainly limited by the amount of available electrical cables within the vehicle. Menta eFGPA IPs allow pre-processing of all raw data from sensors thanks to their highly parallel processing capability. Redundancy checking can also be added in the same way.
Security of electronics systems is of the utmost importance - especially in life-critical systems such as vehicles on the road. Running encryption / decryption algorithms on Menta eFPGA IPs allows these algorithms to be updated in the field - increasing the lifetime of the electronic systems embedded in the vehicle.
Menta eFPGA IPs allow automakers to reduce the number of Electronic Control Units in the automotive OEM supply chain – this drives strong simplification of the supply chain and hence is an important cost reduction.
Future complex automotive systems will soon make use of chips that are designed in technology nodes down to 7 or even 5nm. Using Menta eFPGA IPs helps reduce the risks of a re-spin and therefore saving mask cost and de-risking the time to market.
