Startup emerges from stealth to transform AI computing
The company’s Blaize Graph Streaming Processor (GSP) architecture is claimed to be the first to enable concurrent execution of multiple neural networks and entire workflows on a single system, while supporting a diverse range of heterogeneous compute intensive workloads. The startup says its fully programmable solution brings new levels of flexibility for evolving AI models, workflows, and applications that run efficiently where needed.
“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,” said Dinakar Munagala, Co-founder and CEO, Blaize. “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.”
Unlike single-function ASICs designed for AI, the GSP is meant to be more general purpose to reach many markets, from automotive to the edge to the cloud. The Blaize GSP architecture together with the Blaize Picasso software development platform 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 minimized or eliminated. This approach yields the lowest possible latency, reduces memory requirements and reduces energy demand at the chip, board and system levels.
Blaize GSP is the first fully programmable processor architecture and software platform that is built from the ground up to be 100% graph-native. While there are many types of neural networks, all neural networks are graphs. With the inherent graph-native structure, developers can now build multiple neural networks and entire workflows on a single architecture that is applicable to many markets and use cases. End-to-end applications can be built 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. And AI Application developers can now build entire applications faster, optimize these for edge deployment constraints, and run them efficiently using automated data-streaming methods.
Blaize – www.Blaize.com
Related articles:
ThinCI seeks engineers: opens up two offices in UK
ThinCI raises $65 million in Series C
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
