Retym Raises $75M to Revolutionize AI Data Center Connectivity with Advanced DSP Chips

Spark, Mayfield, and Kleiner Perkins pour another $75M into AI data center chipmaker Retym | TechCrunch

Retym, a U.S. chipmaker with Israeli roots, has raised a fresh $75 million series D led by James Kuklinski of Spark Capital. Existing investors -- Mayfield's Navin Chaddha and Kleiner Perkins' Mamoon Hamid -- also participated, bringing its total raised to $180 million, it says. Retym is another startup benefiting from AI's glow-up. It's chips don't process AI workloads directly -- it's not a Nvidia GPU competitor. It's working on a new "programmable coherent digital signal processing DSP" chip that allows the innards of data centers to communicate faster both internally and with external sources. Data centers needed this kind of tech anyway, but the sudden rise of AI has put overwhelming pressure on them to be faster, more efficient, and handle bigger workloads. The company was founded in 2021 but has been quiet about itself until Monday when it announced this Series D round. Its CTO co-founder Roni El-Bahar published his first blog post on Monday, saying he founded the startup to bring competition to the DSP market that has historically been "controlled by a handful of large semiconductor companies." He was largely referring to Marvell Technology, which currently dominates the DSP industry and has partnerships with Nvidia, Juniper Networks, and many others. Retym -- pronounced "re-time" -- is using TSMC's state-of-the-art 5 nanometer fab for its first chip, which is being tested now, the company told Reuters. Retym did not immediately respond to a request for comment.

Chip startup Retym raises $75 million to build chip that will AI connect data centers

SAN FRANCISCO, March 31 (Reuters) - Chip startup Retym has raised $75 million this year as part of $180 million in total funding the company is using to pursue networking chips for artificial intelligence computing in data centers, the company said on Monday. Retym makes chips that perform digital signal processing (DSP) that helps move information speedily between large data centers, a capability that has become increasingly important with the surge of interest in AI. Creating the underlying AI models that power the likes of ChatGPT requires thousands of chips to be strung together with networking equipment. At the moment the market for the DSP chips Retym is working on is dominated by Marvell Technology (MRVL.O), opens new tab. The chip Retym - pronounced re-time - is working on will help solve a bottleneck that has developed in data centers, chief executive Sachin Gandhi said. Because large numbers of chips must work together on AI computing tasks the connectivity between them is increasingly important. "We are focusing on building coherent DSP chips for the next generation deployment of AI infrastructure and cloud," Gandhi said. Retym's first chip will be designed to move data around from a range of 10 kilometers and 120 kilometers, but will be optimized for 30 kilometers to 40 kilometers. The DSP chip Retym is building uses a modulation technique to ensure data transported is not corrupted. "They took the approach of solving the harder problem with longer distances," said Navin Chaddha, managing partner at Mayfield, a venture fund that has invested in Retym. Retym is using Taiwan Semiconductor Manufacturing Co.'s (2330.TW), opens new tab five nanometer manufacturing process for the first chip, which its engineers are currently testing and validating samples. The Series D funding round was led by Spark Capital. Retym launched four years ago and plans to bring its first product to market this year. Max A. Cherney in San Francisco. Editing by Gerry Doyle Our Standards: The Thomson Reuters Trust Principles., opens new tab Suggested Topics:Technology Max A. Cherney Thomson Reuters Max A. Cherney is a correspondent for Reuters based in San Francisco, where he reports on the semiconductor industry and artificial intelligence. He joined Reuters in 2023 and has previously worked for Barron's magazine and its sister publication, MarketWatch. Cherney graduated from Trent University with a degree in history.

AI datacenter infrastructure company Retym launches alongside $75M raise - SiliconANGLE

AI datacenter infrastructure company Retym launches alongside $75M raise Retym Inc., a technology company specializing in semiconductors for cloud and artificial intelligence data center connectivity, today announced its launch, accompanied by $75 million in a Series D funding led by Spark Capital. Existing investors Kleiner Perkins, Mayfield and Fidelity Investments also participated in this round, bringing the total funding for the startup to more tha $180 million. Pronounced "re-time," Retym focuses on developing programmable coherent digital signal processing or DSP solutions for cloud data centers and AI infrastructure. This semiconductor technology is used for enabling high-speed data transmissions between data centers and within data centers themselves. Co-founder and Chief Executive Sachin Gandhi told SiliconANGLE in an interview that as AI workloads continue to grow in size and complexity, so do the needs for data connectivity speeds and efficiency. Today's data centers contain thousands or millions of graphics processing units separated into clusters and racks, all connected via a backend network. They in turn connect to optical modules, which are commonly used to connect longer distances between buildings or data centers. Retym's DSP semiconductor technology is built into these optical modules, which are needed to allow signals to cross longer distances. But they also provide faster speeds, which makes them ideal for within the data center as well. "What is really happening is coherent technology is moving," Gandhi said. "It's between data centers, but it's also moving inside the data center as speed will increase. We are building the core and DSPs will target this next AI infrastructure and cloud connectivity evolution." Gandhi explained that the boundaries between "inside-the-data center" and "data center interconnect" are beginning to blur as AI model training and inference needs continue to advance. According to market research firm Dell'Oro Group, the global spending on data center compute and networking is expected to exceed $1 trillion by the end of the decade, which will position optical networking and DSP solutions at the top of the list for AI cloud infrastructure. "There is always a bottleneck when it comes to data," Gandhi said. "To avoid the bottleneck, what we do is spend a lot on infrastructure. We increase our speeds from 400 gigabits to 800 gigabits, and from 800 gigabits to 1.6 terabits. We used to see this speed increase happen every four years. Now it's happening every two years." Gandhi said he believes Retym is launching at the right time to capture the AI infrastructure market's requirement for speed and efficiency to stay competitive both inside and between data centers as this pressure increases. "Coherent optics and the DSPs that drive them are becoming increasingly crucial for AI-driven data centers as data volumes and performance requirements continue to rise," said Vlad Kozlov, founder and CEO of the research firm LightCounting. "The ongoing evolution of this market highlights the need for innovative and efficient solutions." With the new fundraise, Gandhi said Retym will move toward finalizing the production of its coherent DSPs and bring them to market before the end of the year.

California Chip Startup Retym Raises $75M to Build Chips for AI Data Centres

With this, the company officially launched publicly, out of stealth mode. Retym, a chip startup with roots in the US and Israel, has raised $75 million in a Series D funding round led by Spark Capital. This brings the company's total funding to $180 million, the company announced on Monday. With this, the company officially launched publicly, out of stealth mode, on Tuesday. The funding round also saw participation from existing investors Kleiner Perkins, Mayfield, and Fidelity Investments. Retym plans to use this funding to develop programmable coherent digital signal processing (DSP) chips for AI and cloud infrastructure. "We are focusing on building coherent DSP chips for the next-generation deployment of AI infrastructure and cloud," said Sachin Gandhi, Retym's co-founder and CEO. The company's DSP chips are designed to enhance data transmission within and between data centres, addressing the growing demand for faster and more efficient communication as AI workloads increase. Retym's first chip is being manufactured using TSMC's five-nanometer process and is expected to be launched later this year. Retym aims to challenge Marvell Technology's dominance in the DSP market. The Series D funding marks a significant step forward for Retym as it charts a multi-generation product roadmap. James Kuklinski, general partner at Spark Capital, has joined Retym's board of directors following the investment. The funding will support scaling to production and continued product development advancements. Retym's technology is crucial for AI data centres, where thousands of chips must work together efficiently to handle large AI models like ChatGPT. The company was founded in 2021 and has been operating quietly until now.

Chip Startup Retym Raises $75 Million to Build Chip That Will AI Connect Data Centers

SAN FRANCISCO (Reuters) - Chip startup Retym has raised $75 million this year as part of $180 million in total funding the company is using to pursue networking chips for artificial intelligence computing in data centers, the company said on Monday. Retym makes chips that perform digital signal processing (DSP) that helps move information speedily between large data centers, a capability that has become increasingly important with the surge of interest in AI. Creating the underlying AI models that power the likes of ChatGPT requires thousands of chips to be strung together with networking equipment. At the moment the market for the DSP chips Retym is working on is dominated by Marvell Technology. The chip Retym - pronounced re-time - is working on will help solve a bottleneck that has developed in data centers, chief executive Sachin Gandhi said. Because large numbers of chips must work together on AI computing tasks the connectivity between them is increasingly important. "We are focusing on building coherent DSP chips for the next generation deployment of AI infrastructure and cloud," Gandhi said. Retym's first chip will be designed to move data around from a range of 10 kilometers and 120 kilometers, but will be optimized for 30 kilometers to 40 kilometers. The DSP chip Retym is building uses a modulation technique to ensure data transported is not corrupted. "They took the approach of solving the harder problem with longer distances," said Navin Chaddha, managing partner at Mayfield, a venture fund that has invested in Retym. Retym is using Taiwan Semiconductor Manufacturing Co.'s five nanometer manufacturing process for the first chip, which its engineers are currently testing and validating samples. The Series D funding round was led by Spark Capital. Retym launched four years ago and plans to bring its first product to market this year. (Max A. Cherney in San Francisco. Editing by Gerry Doyle)

Chip startup Retym raises $75 million to build chip that will AI connect data centers

Retym makes chips that perform digital signal processing (DSP) that helps move information speedily between large data centers, a capability that has become increasingly important with the surge of interest in AI.Chip startup Retym has raised $75 million this year as part of $180 million in total funding the company is using to pursue networking chips for artificial intelligence computing in data centers, the company said on Monday. Retym makes chips that perform digital signal processing (DSP) that helps move information speedily between large data centers, a capability that has become increasingly important with the surge of interest in AI. Creating the underlying AI models that power the likes of ChatGPT requires thousands of chips to be strung together with networking equipment. At the moment the market for the DSP chips Retym is working on is dominated by Marvell Technology. The chip Retym - pronounced re-time - is working on will help solve a bottleneck that has developed in data centers, chief executive Sachin Gandhi said. Because large numbers of chips must work together on AI computing tasks the connectivity between them is increasingly important. "We are focusing on building coherent DSP chips for the next generation deployment of AI infrastructure and cloud," Gandhi said. Retym's first chip will be designed to move data around from a range of 10 kilometers and 120 kilometers, but will be optimized for 30 kilometers to 40 kilometers. The DSP chip Retym is building uses a modulation technique to ensure data transported is not corrupted. "They took the approach of solving the harder problem with longer distances," said Navin Chaddha, managing partner at Mayfield, a venture fund that has invested in Retym. Retym is using Taiwan Semiconductor Manufacturing Co.'s five nanometer manufacturing process for the first chip, which its engineers are currently testing and validating samples. The Series D funding round was led by Spark Capital. Retym launched four years ago and plans to bring its first product to market this year.