EdgeCortix has made a strategic shift from selling AI intellectual property (IP) to selling its own edge-AI inference chips for line-powered systems. The new die, dubbed Sakura, started as a test chip, but the company says customer interest convinced it to offer the chip as a product. The chips come mounted on one of two cards: a dual-M.2 (M-key) card and a low-profile PCIe card. Based in Japan, EdgeCortix has supplemented its $13,5 million in total funding with revenue from FPGAs and ASICs that employ its soft IP. It targets perception – vision, lidar, and related technologies – are aimed at transportation, augmented/virtual reality, industry, smart cities, and drones.
Sakura, revealed first at the recent Linley Spring Processor Conference, implements the company’s dynamic neural accelerator (DNA) engine, adding on-chip SRAM, two LPDDR4X ports, and I/O. The chip has no host CPU, so it operates under the control of an external host. Sakura has a maximum performance of 40 TOPS; on ResNet-50, it achieves 0,4 ms latency at 4,7 W, yielding 533 inferences per second per watt (IPS/W).
Sakura combines six engines for standard 3D convolutions, six engines for 2D depth-wise convolutions, and three vector engines for activations and other miscellaneous operations. Standard convolutions combine all three colour channels, whereas depth-wise convolutions separate the channels, combining results later to reduce computing operations. To boost utilisation, the Mera tools can convert between standard and depth-wise convolutions to keep all engines occupied. Together, the engines are capable of 25 000 MACs per clock cycle.
Within a vector engine, dedicated blocks implement activation functions. Each block is configurable for a range of variants in a family of activation functions. Although this approach yields better performance, it’s less flexible than a fully programmable engine. A 20 MB on-chip SRAM can hold weights and activations or act as a scratchpad. The runtime software reconfigures the network-on-a-chip (NoC) on the fly, reassigning engines as part of EdgeCortix’s focus on achieving high hardware utilisation.
RFiber Solutions has added this group to its line card.
Wi-Fi 6/BLE module enables rapid development Altron Arrow
Telecoms, Datacoms, Wireless, IoT
Telit Cinterion has announced the WE310K6, a fully integrated, low-power module featuring dual-band, dual-stream Wi-Fi 6, and dual-mode Bluetooth/BLE.
Read more...Low phase noise amplifier RFiber Solutions
Telecoms, Datacoms, Wireless, IoT
The MAAL-011158 from Macom is an easy-to-use low-phase noise amplifier that provides 12 dB of gain in a 32-lead QFN plastic package.
Read more...Full sensor to cloud solution CST Electronics
Telecoms, Datacoms, Wireless, IoT
NeoCortec has demonstrated the seamless and rapid development of full sensor-to-cloud solutions using NeoMesh Click boards from MikroE and the IoTConnect cloud solution from Avnet.
Read more...Long-range Wi-Fi HaLow module TRX Electronics
Telecoms, Datacoms, Wireless, IoT
One of Mouser’s newest products is the Morse Micro MM6108-MF08651-US Wi-Fi HaLow Module, which adheres to the IEEE 802.11ah standard.
Read more...Quectel launches 3GPP NTN comms module Quectel Wireless Solutions
Telecoms, Datacoms, Wireless, IoT
Quectel Wireless Solutions has announced the Quectel BG95-S5 3GPP non-terrestrial network (NTN) satellite communication module.
Read more...SIMCom’s A7673x series Otto Wireless Solutions
Telecoms, Datacoms, Wireless, IoT
SIMCom recently released the A7673X series, a Cat.1 bis module based on the ASR1606 chipset, that supports wireless communication modes of LTE-FDD, with a maximum downlink rate of 10 Mbps and a maximum uplink rate of 5 Mbps.
Read more...Multi-channel cable assembly RFiber Solutions
Interconnection
Withwave’s Multi-Channel SMPM Cable Assemblies (WMCM Series) provide a wide range of multiple coax connectors and flexible cable assemblies, with a choice of 26,5, 40 and 50 GHz configurations.
Read more...Long-range connectivity module Avnet Silica
Telecoms, Datacoms, Wireless, IoT
Digi XBee XR 868 RF Modules support the deployment of long-range connectivity applications, and support point-to-point and mesh networking protocols.
printer friendly version