## febr 2021 [Memory Technologies Confront Edge AI’s Diverse ChallengesMemory Technologies Confront Edge AI’s Diverse Challenges](https://www.eetimes.eu/memory-technologies-confront-edge-ais-diverse-challenges/)

> ... multidimensional matrix multiplication, lends itself to analog compute techniques with an array of memory cells used to perform calculations. Using this technique, **Syntiant’s devices are designed for voice control of consumer electronic**s, and **Gyrfalcon’s devices have been designed into a smartphone, where they handle inference for camera effects**.

> The **Flex Logix chip will be used in edge AI inference** applications that require real-time operation, including analyzing streaming video with low latency. This includes A**DAS systems, analysis of security footage, medical imaging, and quality assurance/inspection applications**

> **Experiments have applied MRAM’s stochasticity capabilities to Gyrfalcon’s devices**, a technique whereby the precision of all the weights and activations is reduced to 1 bit. This is used to reduce compute and power requirements dramatically for far-edge applications. Tradeoffs with accuracy are likely, depending on how the network is retrained. In general, neural networks can be made to function reliably despite the reduced precision

> ... Andy Walker, product vice president at **Spin Memory. “We have found that such BNNs can still function with high levels of accuracy**

> **One application is as a so-called unified memory**, “where this emerging memory can act as both an embedded flash and SRAM replacement, saving area on the die and avoiding the static power dissipation inherent in SRAM,” said Walker

> Recent research by **Politecnico Milan using Weebit Nano’s silicon oxide (SiOx) ReRAM technology** showed promise for neuromorphic computing

> While the exact nature of memory systems for edge AI depends on the application, GDDR, HBM, and Optane are proving popular for data centers, while **LPDDR competes with on-chip SRAM for endpoint applications**

- When comparing GDDR6 with LPDDR, 

> ...the low-power DDR version that has been Nvidia’s approach for most non-data–center edge solutions from the Jetson AGX Xavier to Jetson Nano, **Ferro acknowledged that LPDDR is suited to low-cost AI inference at the edge or endpoint.**