comment on this article

‘First’ AI processor that allows execution of multiple neural networks and entire workflows on single system

Blaize has unveiled what it says is the first true Graph-Native silicon architecture and software platform built to process neural networks and enable artificial intelligence (AI) applications with unprecedented efficiency.

The Blaize Graph Streaming Processor (GSP) supports a diverse range of heterogeneous compute intensive workloads, so can meet the demands and complexity of new computational workloads found in AI applications. Also unveiled today was the Blaize Picasso software development platform.

Dinakar Munagala, Blaize’s co-founder and CEO, commented, “Blaize was founded on a vision of a better way to compute the workloads of the future by rethinking the fundamental software and processor architecture. We see demand from customers across markets for new computing solutions that address the immediate unmet needs for technology built for the emerging age of AI, and solutions that overcome the limitations of power, complexity and cost of legacy computing.”

The Blaize GSP architecture and Blaize Picasso software development blend dynamic data flow methods and graph computing models with fully programmable proprietary SOCs.

This allows Blaize computing platforms to exploit the native graph structure inherent in neural network workloads all the way through runtime. The massive efficiency multiplier is delivered via a data streaming mechanism, where non-computational data movement is minimised or eliminated. This gives Blaize systems low latency, reduced memory requirements and reduced energy demand at the chip, board and system levels.

“The proliferation of AI across multiple industries and application areas is dependent upon robust, programmable, efficient, scalable, high-performance hardware, that extends AI processing from cloud datacenters through to the end device, server or appliance,” said Aditya Kaul, Research Director, Tractica.

“It’s becoming clear that traditional processing architectures will not be enough to meet the demands of this new emerging market, with new techniques like graph-based computing showing promise. Success will be defined by combining new computing approaches with modular hardware and a deployment-oriented software stack, all of which is part of the Blaize value proposition from day one.”

The company says developers can use these products to build end-to-end applications integrating non-neural network functions such as Image Signal Processing with neural network functions, all represented as graphs that are processed 10-100 times more efficiently than existing solutions.

Above: Blaize founders Val G. Cook, Chief Software Architect, Satyaki Koneru, CTO, Dinakar Munagala, CEO, and Ke Yin, Chief Scientist, VP Engineering (left to right).

Charlotte Hathway

Comment on this article

This material is protected by MA Business copyright See Terms and Conditions. One-off usage is permitted but bulk copying is not. For multiple copies contact the sales team.

What you think about this article:

Add your comments


Your comments/feedback may be edited prior to publishing. Not all entries will be published.
Please view our Terms and Conditions before leaving a comment.

Related Articles