ZeroPoint raises $5.5M for its on-chip data compression technology – Business
ZeroPoint Technologies AB, a startup with hardware that enables processors to make more efficient use of DRAM, has secured $5.5 million in early-stage funding.
The company said in its announcement of the Series A raise today that Munich-based Matterwave Ventures was the lead investor. Industrifonden, Climentum Capital and Chalmers Ventures participated as well.
ZeroPoint is a spinoff from the Chalmers University of Technology in Gothenburg, Sweden. The startup was launched to commercialize research into data compression techniques that its co-founders, Per Stenström and Angelos Arelakis, carried out at the university. ZeroPoint says that its technology can boost data center servers’ power-efficiency and reduce the cost of operating them in the process.
The speed at which a server processor can fetch data from DRAM directly influences its performance. As a result, speeding up data retrieval times is a major priority for semiconductor designers. One way to accelerate the retrieval workflow is by compressing the files that a chip processes: the smaller the file, the quicker it can be pulled from memory.
Using compression to speed up chips in this manner was historically impractical. The reason is that the most advanced compression algorithms are relatively slow, which negates any processing speedups they might provide by optimizing DRAM usage. There are faster algorithms out there, but they’re significantly less adept at shrinking data.
ZeroPoint claims that it has solved this tradeoff. The company offers a tiny semiconductor module, the Ziptilion-BW, that it says is both adept at shrinking data and fast. The chip promises to reduce the memory footprint of datasets by up to 400% while running 1,000 times faster than some traditional compression methods.
The Ziptilion-BW’s performance is made possible by what the ZeroPoint describes as a set of entropy-based compression algorithms. According to the company, those algorithms use statistical methods to analyze a chip’s memory and determine what types of data it contains. The Ziptilion-BW then uses this information to find the most efficient way of compressing the data.
The chip combines ZeroPoint’s entropy-based algorithms with a number of other performance optimization methods. Notably, it stores information that was used recently in an onboard cache. Servers can retrieve information from this cache more quickly than from the DRAM where the data is usually stored, which speeds up calculations.
According to ZeroPoint, the Ziptilion-BW can be implemented as a semiconductor module measuring just 0.4 square millimeters. That makes it relatively simple for chipmakers to integrate the technology into their processors. ZeroPoint estimates that the Ziptilion-BW can boost data center servers’ performance per watt by 50% and reduce their total cost of ownership by 25% in the process.
“ZeroPoint Technologies’ customers include some of the biggest semiconductor companies in the world and our products are in demand by data center operators looking to overcome the mounting challenge of memory bottlenecks,” said ZeroPoint Chief Executive Officer Klas Moreau (pictured, center, with co-founders Stenström and Arelakis).
The company will use the proceeds from its newly announced funding round to launch additional hardware acceleration products. ZeroPoint also plans to scale its sales efforts, as well as hire more employees in the U.S. and Sweden to support the growth initiative.