Renesas Electronics’ ISL91301B PMIC designed in Google AI products

Author: EIS Release Date: Mar 5, 2020


Renesas’s ISL91301B power management IC (PMIC) has been designed into Google’s Coral products.

They include the Mini PCIe Accelerator, M.2 Accelerator A+E Key, M.2 Accelerator B+M Key, and system-on-module (SoM). Google Coral integrates into processes at any scale, helping designers employ artificial intelligence (AI) in several industries.ISL91301B

The ISL91301B PMIC is designed for low power edge AI processors. It has quad buck regulator outputs, each delivering up to 4A (total 16A) with 94% peak efficiency.

The PMIC incorporates Renesas’ R5 modulation technology for quick single-cycle transient response, tuned compensation, and fast switching frequency.

Its 70mm2 size means power supply designers can design a small AI solution with 2x2x1mm low profile inductors, small capacitors, and a few passive components.

The ISL91301B boasts low quiescent current, light load efficiency, high DC accuracy and fast dynamics, Renesas claims.

Google Coral technology is designed to enable advanced neural network processing for low-power devices.

It uses the Edge tensor processing unit (TPU) coprocessor, which can perform up to four trillion operations per second (TOPS), using 0.5 watts for each TOPS (2 TOPS per watt).

It can execute mobile vision models, such as MobileNet v2 at 400 FPS, in a power efficient manner, the company says.

Andrew Cowell, vice president, mobility infrastructure & IoT power business division at Renesas says “Our multiphase PMIC employs Renesas’ industry-leading R5 modulation technology for blazingly fast transient response, which allows Google’s designers to dynamically scale power to improve overall system performance.”

Kai Yick, Google hardware manager said, “Renesas’ Multiphase PMICs provide our engineering team with the high performance and design flexibility required to build Coral. Our collaboration with Renesas ensures our Edge TPU achieve the highest power efficiency to perform calculations offline and locally.”