Earlier this year, Renesas completed its acquisition of Dialog Semiconductor, which was meant to help Renesas move into the Internet of Things (IoT) market. Finally, this acquisition is starting to pay off, with Renesas releasing its first-ever FPGA (field-programmable gate array) last week.
An overview of an FPGA. Image used courtesy of Lattice Semiconductor
The new FPGA, which is aimed at consumer and IoT applications, claims to be an industry first to combine ultra lower power with low cost.
This article will take a look at the role of FPGAs in IoT applications and a deeper dive into Renesas’ new release.
FPGA for IoT
For low power and cost, many IoT devices employ a standard, single-core microcontroller unit (MCU), which, in many ways, is not the ideal processing unit for IoT.
When it comes to IoT devices, we generally expect the following workflow:
- Read an on-board sensor(s)
- Compare the sensor value to some predetermined threshold
- Send the data to the cloud for ML inference and logging and/or feed it to an edge ML model
- Actuate (if applicable) or wait for the next reading
The problem with using an MCU in this scheme is that they lack parallelism (i.e., they tend to be single-threaded). That is to say, the workflow above has to be done sequentially; only once the previous step is completed can the next one begin.
This single-thread waiting time is a hindrance for an IoT device, which is often trying to read multiple sensors at once and take real-time measurements. The device will have to wait for the workflow of one sensor to finish before it can check the next sensor.
Further, the device can only take the measurements as frequently as allowed by the time required to finish the rest of the workflow.
A general overview of MCUs vs. FPGAs. Image from FPGA Key
In general, IoT devices can benefit greatly from the parallelism that FPGAs can provide. With an FPGA scheme, the device can simultaneously read multiple sensors, perform a comparison, and do any communication necessary.
Further, FPGAs can offer benefits in terms of AI/ML (Artificial Intelligence/Machine Learning) computation over MCUs, meaning that IoT devices can also have the potential to bring computing to the edge through the use of an FPGA.
To this point, one of the major challenges with bringing FPGAs to IoT devices is that they are generally not cheap or very low power, both of which are essential aspects of IoT design.
Recently, to serve consumer and IoT markets, Renesas has released its first-ever FPGA family. This new family, dubbed ForgeFPGA, aims to address the need for a low-cost, low-power programmable logic device.
Renesas’ ForgeFPGA. Image used courtesy of Renesas
The devices in the family will begin at sizes smaller than 5000 logic gates, with the first two offerings being 1K and 2K Look Up Tables.
Concerning power, ForgeFPGA will offer a standby current as low as 20 uA, which Renesas claims represent half the power consumption of competing FPGAs on the market.
In terms of cost, Renesas has stated that, at volume, the price for a ForgeFPGA device will be well under $0.50.
Further, the new family will hope to sidestep FPGA development struggles with free development software to accommodate both new and experienced FPGA developers. This accommodation will include a “macrocell mode” that uses a schematic capture-based development flow and an “HDL” mode that provides a familiar Verilog environment.
A Future With FPGAs
Looking at IoT, it seems that the ability to have parallel computing elements could be highly beneficial. To this end, FPGAs are a great option, so long as they are low power and cheap enough to meet the specs of an IoT device.
With its new ForgeFPGA family, Renesas is trying to give designers the best of all worlds. With this breakout, it will be interesting to see where Renesas goes next with this technology or further leverages its boost from Dialog Semiconductor.
For lighting, electrical, signage, and technology solutions that allow you to do more call Sverige Energy today at +4(670) 4122522.