Kneron’s KL720 which is a tiny AI chip outperforms Google’s Coral TPU by two times at half the power. Kneron is a startup from California. A new neural processing unit (NPU) for edge AI applications has been introduced by Kneron.
Sophisticated AI-enhanced applications are enabled by KL720 NPU that do not use the cloud. More compute horsepower onboard is provided by the chip than competitors and consumes less power. Data are sent to data centers for processing and interpreting are done by numerous cloud-enabled AI platforms, such as Apple’s Siri and Amazon’s Alexa. It takes time and energy to send data to the cloud. Along with privacy and security concerns.
Due to performance and power limitations, processing and interpreting data locally is not always possible. Enough performance at the edge of the network can be offered by the Kneron’s KL720 to enable local processing and interpreting. The Kneron KL720 is a tiny SoC which controls the system. It is a pack of 700MHz NPU along with a 500MHz Cadence DSP which acts as a co-processor. It also contains an Arm Cortex-M4 400MHz core for system control. The SoC boasts 128MB of LPDDR4 RAM.
Interfaces including MIPI-CSI-2, USB, SPI, SDIO, and UART among others are supported. Depending on the targeted application, the chip will come in a 9x9mm (companion mode) or 11x11mm (host mode) form factor. Up to 4K resolution images and up to 2K resolution videos can be processed by Kneron’s KL720. The chip also supports full natural language processing. The performance of around 1.5TOPS at 1.2W (average power consumption) is offered by Kneron’s KL720 AI. At the same time, Google’s Coral edge AI solution offers 4TOPS at 2W.
When compared by Google’s Coral edge, Kneron’s KL720 is four times more efficient as per the MobileNetV2 benchmark. It could because of the company’s software excellence. If switched to the KL720, a DJI drone powered by Intel’s Movidius chips would double its battery life. As per Albert Liu, founder, and CEO of Kneron, the power through the combination of unmatched energy-efficiency and Kneron’s industry-leading AI algorithms by Kneron’s KL720 will enable a new era for smart devices.
Even more, devices can be used to take advantage of the benefits of edge AI and protecting user privacy which is enabled by its low cost. Competitors can’t match this to an extent. The company offers the most comprehensive suite of AI chips and software for devices on the market combined with existing KL520. Within the coming year, expect Kneron’s KL720 to power actual devices as its sampling will start soon.
