26 June 2019. A start-up enterprise is developing computer chips designed to run artificial intelligence, or A.I., applications but with an architecture it says can avoid memory logjams. NeuroBlade, a two year-old company in Hod-Hasharon, Israel is also raising $23 million in its first venture funding round.
The company aims to solve a problem with A.I. hardware used in applications such as image and speech recognition, video analysis, and autonomous driving, where the large volumes of data and intensive processing can degrade and slow the system’s performance. NeuroBlade’s technology, says the company, is designed to operate more efficiently than other A.I. chip sets, working with a corresponding software stack to make better use of a system’s resources.
According to NeuroBlade, its chips and software improve the operating connections between logic and memory and thus re-balance these elements in the system. As a result, says the company, memory bottlenecks are reduced, while retaining access to a system’s high-density memories and maintaining the accuracy of algorithms being run.
NeuroBlade was founded in 2017 by software and algorithm designers Elad Sity and Eliad Hillel, who worked previously at the Israeli solar technology company SolarEdge. Both Sity and Hillel are engineering graduates of Tel Aviv University and the technical arm of Israel’s intelligence service. The company received $4.5 million in seed funds from StageOne Ventures and Grove Ventures at the outset and operated in incubation mode until now. Sity is NeuroBlade’s CEO, while Hillel is the company’s chief technology officer.
NeuroBlade’s first venture funding round is raising $23 million led by Marius Nacht,, a co-founder of security software company Check Point Software Technologies, with participation from seed-round investors StageOne Ventures and Grove Ventures. Also taking part in the financing is Intel Capital, the venture funding arm of semiconductor maker Intel. The company plans to use the funds to ramp-up R&D and staffing to bring its first-generation A.I. chip to market.
In a statement e-mailed to Science & Enterprise, Tal Slobodkin, a partner at StageOne Ventures notes, “The chip developed by NeuroBlade will allow running A.I. algorithms on par with the performance of market-leading chips but at a smaller size or with significantly better performance at similar throughput and size. NeuroBlade’s technology targets servers and is also relevant to end devices, such as security cameras, laptops, cars, multimedia, and more.”
More from Science & Enterprise:
- Sanofi, Alphabet Partner on A.I., Analytics
- A.I. Harnessed to Diagnose Children’s Gut Diseases
- Infographic – Microsoft, IBM Lead A.I. Patent Race
- A.I., Robotics Studied for Military Trauma System
- MIT, Air Force to Open A.I. Accelerator Lab
* * *
You must be logged in to post a comment.