SpeedaStartup, based on Tel Aviv, designed to accelerate large data analytics and AI’s workloads, has increased the $ 44 million series B funding round, making it $ 114 million.
Series B Round led by its existing investors, including Walden Catalist Ventures, 83 North, Coach Dispotable Technologies, Patango First, and Viola Ventures, as well as strategic investors, including Intel’s CEO and Walden Catalist Ventures, and former CEO of Walden Catalist Ventures. O
APU architecture is focused on solving specific obstacles of analytics at computing levels, vice versa Graphics processing unit (GPU), which was initially designed for graphics, and later, according to Startup, was edited for AI and data related tasks.
“For decades, data analytics have relied on standard processing units, and recently, companies like NVIDIA have invested GPUs for the burden of analytics,” said Adi Jelon, CEO of Speeda. “But these are either general -purpose processors or processors designed for other workloads, not ground chips made for data analysis. Our APU is aimed for data processing and can change the same APU server rack, which can provide better performance.”
The speeda was founded by six founders in 2019, some of them were the first researchers in which the CGRA technology was developed with multi -threaded coarse -filled grains (CGRA) technology. The founders cooperated with ASIC design experts to solve a fundamental problem: were performing data analytics through ordinary purpose processors. If the workload is very complicated, they may need to tap in hundreds of servers. The founders believed that they could produce a single dedicated processor to complete the task using less energy.
“We have seen it as an opportunity for decades of research in the Silicon that the method of following industry data can be changed,” said Gelwan.
Its APU is currently the targets Apache spark work loadsAccording to the company’s CEO, its roadmap includes supporting every major data analytical platform.
“Our goal is to be a standard processor for data processing – as GPUA became a default for I training, we want you to be a default for data analysis in each database and analytical platform,” Jelwan told Tech Crunch.
Startup says several major companies are examining his APU, though he refused to name him. The official product launch is set up for data BRICS data and AI summit in the second week of June. Gilan said he would present his APU for the first time in the event.
Speeda has claimed a special case where its APU completed the pharmaceutical work burden in 19 minutes, which was significantly faster than 90 hours when using a non -special processing unit, which in turn improves 280X speed.
Startup said he had achieved several milestones since his last fundraising, which included finalizing his first APU design and manufacturing at the end of 2024.
“We have moved toward the concept by testing A field programable gate arrow (FPGA)And now we are proud to say that we have hardware we are currently launching. We already have a growing pipeline of enterprise consumer, which is eagerly awaiting this technology and we are ready to scale operations in our market.