In response to the reporter's first question asking if power consumption was the biggest reason behind the launch of the newly announced sensor processing SoC 'EOS platform,' Director Itsu Wang (QuickLogic Asia Sales & Marketing) answered without hesitation that it was.

QuickLogic's EOS platform is another name for S3, the successor to the existing sensor hub product S2 (Arcticlink). By putting the name 'Platform' front and center, it clearly demonstrated its determination to completely differentiate itself from its own existing products as well as the products of other companies.

Why Multicore?

Director Itswang emphasized that the foundation of this differentiation lies in QuickLogic's strategy of possessing software platform technology along with hardware platform technology in the sensor field.

This is also why the EOS platform, a multi-core SoC, can have three dedicated processing engines.

The core of the EOS platform is, first, the ARM Cortex-M4F microcontroller, the Micro DSP Flexible Fusion Engine, and finally, the Sensor Manager capable of managing sensors.

The EOS platform is a type of hybrid car

The FFE and sensor manager handled a large amount of algorithm processing, minimizing the duty cycle for the floating-point microcontroller, which allowed for a significant reduction in overall power consumption.

The reason why algorithms tailored to specific characteristics must exist to reduce power consumption is related to recent trends in sensor processing platforms. This is because increasingly complex algorithms are being added to features such as pedometers, user activities (running, biking), user transport, sensor calibration, and user gestures in existing platforms.

▲ Sensor Herbs market growth rate

Director Itswang stated that recent wearables and smartphones require platforms suitable for advanced algorithms such as motion heart rate correction, voice triggers, and indoor navigation. Since processing these latest features naturally increases power consumption, 'multicore processing' is necessary.

If you compare the new EOS platform to a car, it is a hybrid car. With a battery, you can drive (energy-efficiently) without using gasoline, using gasoline instead of the battery when power is needed.

It is also characterized by the fact that while it can basically be used with three processing cores, core performance can be increased as needed by adding an FPGA. He explained that although another core can be created by adding a Fusion Engine to the FPGA, running just the M4 and FFE is sufficient because the FFE is a low-power architecture.

FPGAs can be conveniently reprogrammed when FFE memory is overflowing or when additional peripheral functions are needed. It was also added that, given this situation, there is a stark difference in power consumption compared to flash-based MCUs.

That is not all. A sound system has been added to the EOS platform as a hardware block. This addresses the issue of high system load associated with competitor products, which support PDM via software.

While existing MCUs typically only support I2S, the EOS platform incorporates PDM and PCM as hardware blocks. PDM microphones have the advantage of being significantly cheaper than I2S while consuming similar amounts of power. Also notable is the addition of a Low Power Sound Detector (LPSD).

In other words, by processing noise analysis as a hardware block, this also contributes to reducing power consumption. Accordingly, QuickLogic stated that the power consumption is less than 350µA, which is significantly lower than that of existing microcontroller-based solutions.

Targeting the smartphone and wearable markets

The EOS platform, which boasts such low power characteristics, targets smartphones and wearables.

Designers of mobile, wearable, and IoT devices can easily design next-generation sensor-centric applications such as pedestrian dead reckoning (PDR), indoor navigation, and motion-compensated heart rate measurement.

The focus of the EOS platform is power saving. This is the reason why the EOS platform was designed as a multi-core system. Sensor hub and smart sensor companies can be competitors, but the important thing is whether they can satisfy what the customer wants. QuickLogic is a company that not only sells sensor hubs but also provides algorithms.