DSP, Micros & Memory


Speech processing SoC uses Microchip’s SuperFlash technology

28 February 2022 DSP, Micros & Memory

Computing-in-memory technology is poised to eliminate the massive data communications bottlenecks otherwise associated with performing artificial intelligence (AI) speech processing at the network’s edge, but requires an embedded memory solution that simultaneously performs neural network computation and stores weights.  

Microchip Technology, via its Silicon Storage Technology (SST) subsidiary, announced that its SuperFlash memBrain neuromorphic memory solution has solved this problem with the WITINMEM neural processing SoC (system-on-chip), the first in volume production that enables sub-mA systems to reduce speech noise and recognise hundreds of command words, in real time and immediately after power-up. 

Microchip has worked with WITINMEM to incorporate Microchip’s memBrain analog in-memory computing solution, based on SuperFlash technology, into WITINMEM’s ultra-low-power SoC. The SoC features computing-in-memory technology for neural networks processing including speech recognition, voiceprint recognition, deep speech noise reduction, scene detection and health status monitoring. WITINMEM, in turn, is working with multiple customers to bring products to market during 2022 based on this SoC. 

Microchip’s memBrain neuromorphic memory product is optimised to perform vector matrix multiplication (VMM) for neural networks. It enables processors used in battery-powered and deeply embedded edge devices to deliver the highest possible AI inference performance per Watt. This is accomplished by both storing the neural model weights as values in the memory array and using the memory array as the neural compute element. The result is 10 to 20 times lower power consumption than alternative approaches, along with lower overall processor bill of materials (BOM) costs because external DRAM and NOR are not required.  

Permanently storing neural models inside the memBrain solution’s processing element also supports instant-on functionality for real-time neural network processing. WITINMEM has leveraged SuperFlash technology’s floating gate cells’ non-volatility to power down its computing-in-memory macros during the idle state to further reduce leakage power in demanding IoT use cases.  


Credit(s)



Share this article:
Share via emailShare via LinkedInPrint this page

Further reading:

Microchip enhances TrustMANAGER platform
Altron Arrow DSP, Micros & Memory
Firmware over-the-air updates and remote cryptographic key management provide scalable solutions for addressing IoT security challenges.

Read more...
Adaptive optics’ power solution
Altron Arrow Opto-Electronics
Vicor power-dense adaptive optical modules enable colossal telescopes to look into the past for deep space discoveries.

Read more...
MCU for low-power, IoT applications
NuVision Electronics DSP, Micros & Memory
Silicon Labs recently announced the PG26, a general-purpose microcontroller with a dedicated matrix vector processor to enhance AI/ML hardware accelerator speeds.

Read more...
Wide input voltage buck-boost converter
Altron Arrow Power Electronics / Power Management
The MAX77859 from Analog Devices is a high-efficiency, high-performance buck-boost converter targeted for systems requiring a wide input voltage range of between 2,5 and 22 V.

Read more...
High-density power module for AI at the edge applications
Altron Arrow Power Electronics / Power Management
The MCPF1412 power module from Microchip has integrated I2C and PMBus interfaces for flexible configuration and monitoring.

Read more...
EEPROMs for industrial and military markets
Vepac Electronics DSP, Micros & Memory
Designed to ensure the data retention and the secure and safe boot of digital systems, the memory product line includes small and medium density EEPROMs from 16 kb to 1 Mb.

Read more...
When it comes to long-term reliability of RF amplifier ICs, focus first on die junction temperature
Altron Arrow Editor's Choice Telecoms, Datacoms, Wireless, IoT
When considering the long-term reliability of integrated circuits, a common misconception is that high package or die thermal resistance is problematic. However, high or low thermal resistance, by itself, tells an incomplete story.

Read more...
PLCnext – Open, IIoT-ready industrial platform
IOT Electronics DSP, Micros & Memory
PLCnext can be used alongside an existing PLC system, collecting control system data via EtherNet/IP, PROFINET, or MODBUS, and can push this information to a cloud instance.

Read more...
ICs vs modules: Understanding the technical trade-offs for IoT applications
NuVision Electronics Editor's Choice DSP, Micros & Memory
As the IoT continues to transform industries, design decisions around wireless connectivity components become increasingly complex with engineers often facing the dilemma of choosing between ICs and wireless modules for their IoT applications.

Read more...
Hardware quantum resistance to embedded controllers
Avnet Silica DSP, Micros & Memory
To help system architects meet evolving security demands, Microchip Technology has developed its MEC175xB embedded controllers with embedded immutable post-quantum cryptography support.

Read more...