As artificial intelligence and data-driven applications continue to push the limits of computing power, data centers face an ever-growing demand for faster, more efficient, and energy-conscious data transfer solutions. In 2023 alone, data centers consumed approximately 4.4% of U.S. electricity, with estimates suggesting that number could rise to 12% by 2028. A startup based in Ottawa, Canada, Hyperlume, is on a mission to redefine chip-to-chip communication by addressing both speed and energy efficiency challenges.
Hyperlume has developed a groundbreaking microLED technology capable of transmitting data at unprecedented speeds while consuming significantly less power than conventional copper-based connections found in data centers. By focusing on reducing latency and increasing efficiency, the company aims to revolutionize how data is transferred across racks, ultimately improving computing performance and reducing operational costs for large-scale data operations.
The startup was founded in 2022 by electrical engineer and data transfer specialist Mohsen Asad and low-power circuit design expert Hossein Fariborzi. With their combined expertise, they saw a significant opportunity to tackle a growing issue in the tech industry: the inefficiency of traditional data transfer methods. Their innovative approach aims to eliminate bottlenecks in data flow, ensuring seamless communication between processing units, an essential need in an era where artificial intelligence models require near-instantaneous data exchange.
“I was working on microLEDs, I was working in data transfer, and this boom of AI and the requirements for sending information from chip to chip, power consumption—all things came together naturally,” said Asad. “We found a huge market opportunity.”
Latency, the time delay in data transmission, has long been a hurdle for high-performance computing. The exponential growth of AI-powered applications has further exposed the inefficiencies of conventional interconnect solutions, where delays can significantly slow down overall system performance. Hyperlume’s microLED technology is designed to counteract these inefficiencies, ensuring that chips within data centers can communicate almost instantaneously.
“If we can solve this latency issue practically, we make [chips] work faster together,” Asad explained. “When you have large language models, you need the chips to communicate with almost zero latency.”
While developing their solution, Asad and Fariborzi explored various existing interconnect technologies. Silicon-based connections were an option but proved too costly for mass deployment. Similarly, while laser-based solutions demonstrated efficiency, their high costs made them impractical. Hyperlume’s key innovation came from adapting microLED technology—widely used in display applications—for high-speed, low-energy data transmission.
Hyperlume’s technology is built around two core components: ultra-fast micro LEDs and a low-power application-specific integrated circuit (ASIC). The ASIC enables the microLEDs to efficiently transmit data while maintaining seamless communication with other chips, providing a high-speed, cost-effective alternative to fiber optic connections.
“The secret sauce is ultra-fast micro-LEDs and, on the other side, a low-power ASIC that drives everything and communicates with other chips,” said Asad.
The startup is currently working with several early adopters, mainly in North America, to fine-tune its technology and validate its real-world application. Hyperlume has received significant interest from hyperscalers—large cloud computing providers—as well as cable manufacturers and other industry players seeking to optimize their data transmission capabilities.
“The first stage for us is to work with those early adopters,” Asad explained. “As soon as the technology is proven and goes inside data centers with those early adopters, it’s going to give us a chance to scale to work with the rest of the market. The demand is there and is growing and growing every year.”
To accelerate its development and expansion, Hyperlume recently raised a $12.5 million seed funding round, led by BDC Capital’s Deep Tech Venture Fund and ArcTern Ventures, with participation from MUUS Climate Partners, Intel Capital, and SOSV. The funding will allow the company to expand its engineering team, further refine its technology, and prepare for broader market adoption.
Looking ahead, Hyperlume envisions its technology playing a crucial role in the next generation of AI and data processing infrastructure. Beyond optimizing chip-to-chip communication, the company aims to expand its bandwidth capabilities to meet the growing needs of high-performance computing applications. Asad sees Hyperlume evolving into a key AI connectivity solutions provider, offering cutting-edge interconnect technologies tailored for the increasing data demands of the future.
“Right now, we are focused on optical connections—to connect chips, to connect boards—but the way that we see the company growing is that it is going to be an AI connectivity solution provider,” Asad noted.
With global data consumption rising and AI-driven workloads requiring faster and more sustainable computing solutions, Hyperlume’s microLED-based interconnects have the potential to become a game-changer in the industry. By significantly reducing power consumption while improving data transfer speeds, the company is addressing one of the most pressing challenges facing modern data centers.
As hyperscalers, enterprises, and cloud providers seek innovative ways to optimize their computing performance, Hyperlume’s breakthrough technology could be a pivotal solution in reshaping the future of high-speed data communication. With demand for efficient data transfer solutions showing no signs of slowing down, Hyperlume is positioning itself at the forefront of the next evolution in AI-driven data infrastructure.