Edge AI Innovations #13

Hey, here are some of the note worthy news on Edge AI you won't want to miss:

1. Cognizant has announced the launch of Neuro® Edge which is a platform designed to empower businesses across industries to leverage artificial intelligence and generative AI at the edge. Neuro® Edge is designed to power the entire value chain of edge AI, from chips and devices to applications and business solution deployments. The platform is cloud agnostic, making it ideal for hybrid-and-multi-cloud environments, with computing power placed directly at the device.
2. Axelera AI, a company that designs hardware to accelerate artificial intelligence, has closed a successful $68 million Series B funding round. This funding will allow the company to expand into new markets and product offerings. Axelera AI's technology is based on Digital In-Memory Computing and RISC-V technology. Axelera AI's product is a hardware and software solution called the Metis™ AI Platform. It is designed to be used for edge AI inference applications including computer vision and generative AI. The platform is said to be high performance, cost-effective, and energy efficient. The Metis™ AI Platform is based on Axelera AI's Digital In-Memory Computing (DiMC) technology, which is a type of neuromorphic computing that uses specialized hardware to mimic the structure and function of the human brain. DiMC is said to be significantly faster and more energy efficient than traditional von Neumann computing architectures.
3. Mamba AI: A new AI architecture called Mamba is showing promise for edge computing applications. Developed by researchers, Mamba is a state space model (SSM) architecture that boasts linear computational growth with increasing context length, making it more efficient than traditional Transformer models. This efficiency translates to faster inference times and the ability to scale to larger models and longer context lengths, all while maintaining a relatively small model size. Mamba's design is particularly well-suited for GPUs, but researchers believe it can be adapted to other types of hardware accelerators commonly found in edge devices. This could significantly improve the performance and efficiency of AI applications on edge devices, which are often constrained by limited computational resources. While Mamba is still a relatively new architecture, its potential for edge computing is clear. With further research and development, Mamba could revolutionize the way AI is deployed on edge devices, enabling new and innovative applications across various industries.

This newsletter focuses on the technological advancements and innovations that can affect Edge AI engineers, hands-on developers and decision makers. There are more exciting Edge AI news on the horizon, stay tuned for the weekly reports.