News: Optoelectronics
25 February 2022
HP and Ayar Labs to co-develop silicon photonics-based optical I/O for HPC and AI
Texas-based Hewlett Packard Enterprise and chip-to-chip optical connectivity firm Ayar Labs of Santa Clara, CA, USA have announced a multi-year strategic collaboration to develop silicon photonics-based optical I/O technology, supporting future requirements for high-performance computing (HPC) and artificial intelligence (AI) solutions.
The two organizations also announced that HPE’s venture arm Hewlett Packard Pathfinder has made a strategic investment in Ayar Labs to accelerate the development and adoption of joint future technologies. As part of the collaboration, HPE and Ayar Labs will partner on photonics research and commercial development, building a joint ecosystem of solution providers, and customer engagements.
The collaboration will focus on Ayar Labs’ development of high-speed, high-density, low-power optical-based interconnects to target future generations of HPE Slingshot, the industry’s only high-performance Ethernet fabric specifically designed for HPC and AI solutions. By combining these technologies, the teams are positioned to design next-generation high-performance networking solutions and novel disaggregated system architectures that are critical for increased flexibility, efficiency, performance and throughput to support data-intensive demands of future workloads.
“We continue to invest in and develop our HPC and AI technologies to further boost application performance for our customers and help them unlock greater value from their research, engineering and business initiatives,” says Justin Hotard, senior VP & general manager, HPC and AI, at HPE. “By partnering with Ayar Labs, we will advance innovation for the HPC and AI market, and leverage their expertise in optical I/O in future generations of HPE Slingshot to deliver unprecedented bandwidth and speed, at lower levels of power and latency, to meet requirements for growing demands in scale and performance.”
Addressing future of HPC and AI application performance with optical interconnect
As the industry enters the exascale era (representing an increase in performance and advanced capabilities for HPC and AI), electrical-based networking offerings will eventually reach bandwidth limits, creating challenges in latency and overall application performance. To prepare for and address these future challenges, HPE and Ayar Labs plan to develop capabilities that leverage silicon photonics-based optical I/O technology to integrate with HPE Slingshot, a purpose-built HPC and AI interconnect designed for next-generation technologies, such as the upcoming US Department of Energy exascale supercomputers.
Optical I/O uniquely changes the performance and power trajectories of system designs by enabling compute, memory and networking ASICs to communicate with dramatically increased bandwidth, at a lower latency, and at a fraction of the power of existing electrical I/O solutions. The technology is also foundational to enabling emerging heterogeneous compute systems, disaggregated, pooled designs, and unified memory architectures that are critical to accelerating future innovation.
“HPE is the leader in advanced systems architectures for HPC, AI, and networking,” says Ayar Labs’ CEO Charles Wuischpard. “This collaboration will accelerate the introduction of a whole new class of system architectures that overcome the existing limitations of traditional interconnects, ushering in a new era for efficient scalable high-performance computing,” he believes.
“Ayar Labs represents an exciting investment opportunity for Hewlett Packard Pathfinder as we closely support HPE’s HPC and AI initiatives,” comments Paul Glaser, VP & head of Hewlett Packard Pathfinder. “Ayar Labs’ optical I/O technology is highly differentiated and critical to the evolution of high-performance computing architectures.”
Sivers to develop CW-WDM MSA-compliant DFB laser arrays supporting Ayar’s optical I/Os
Ayar demos ultra-dense optical interconnects on GlobalFoundries’ silicon photonics process
