- Delivering New Computing Paradigm Using Photonics
- Improving Workload Performance with Low-Latency, Energy-Efficient Connectivity Between xPUs and Memory
- Demonstrating Photowave Capabilities at Flash Memory Summit 2023
Lightelligence, the global leader in photonic computing and connectivity systems, today announced Photowave™, the first optical communications hardware designed for PCIe and Compute Express Link (CXL) connectivity, unleashing next-generation workload efficiency.
Photowave, an Optical Networking (oNET) transceiver leveraging the significant latency and energy efficiency of photonics technology, empowers data center managers to scale resources within or across server racks. The first public demonstration of Photowave will be at Flash Memory Summit today through Thursday, August 10, in Santa Clara, Calif.
With the unveiling of Photowave, Lightelligence delivers on its promise of a new computing paradigm using photonics. Today’s product release follows Lightelligence’s introduction of Hummingbird™, an Optical Network-on-Chip (oNOC) processor for domain-specific artificial intelligence (AI) workloads in June, and PACE, the world’s first integrated photonic computing system for Optical Multiply Accumulate (oMAC) in late 2021.
“We are more than excited to unveil the next piece of our product portfolio,” comments Yichen Shen, CEO of Lightelligence. “Photowave will soon set the standard for workload efficiency. It makes it possible to build scalable computing and accelerator pods with just the right amount of resources.”
Introducing Photowave
As compute and memory stranding becomes an increasingly costly issue for data center workloads, the industry-backed CXL standard enables the disaggregation of compute, accelerator, and memory resources to reduce stranding of resources while improving infrastructure efficiency. The current Ethernet-based infrastructure carries too much latency overhead to enable this disaggregation.
Photowave provides a solution by interconnecting remote devices together using CXL over lower-latency fiber optic cable, extending reach to enable memory pooling at pod scales and beyond. This facilitates scalable CXL fabrics in the composable data center.
The Photowave product line includes various form factors, including a standard PCIe card, OCP 3.0 SFF card, and an active optical cable to achieve successful deployment of CXL-based infrastructure enhancements. They can be used in server platforms, CXL switches, memory appliances, and xPUs.
Photowave enables CXL 2.0/PCIe Gen 5 connectivity over optics with support for x16, x8, x4 and x2 bifurcation modes of operation, allowing for a wide variety of deployment scenarios. In addition to high-speed electrical-to-optical conversion for the data signals, Photowave also supports sideband signals over optics making it possible to implement more efficient and reliable disaggregation architectures.
Availability and Pricing
Photowave is now available for customer trials and deployments. Contact Lightelligence at info@lightelligence.ai for inquiries on pricing and availability.
Lightelligence at Flash Memory Summit and Hot Chips
Lightelligence will demonstrate Photowave in collaboration with AMD, MemVerge and Samsung at Flash Memory Summit in Booth #915 from today through Thursday, August 10, at the Santa Clara Convention Center in Santa Clara, Calif. The demo will showcase the benefits of memory expansion using CXL for superior performance of a large language model (LLM) inference workload running on AMD Genoa server platforms with optical connectivity to Samsung memory expansion modules over Photowave PCIe cards. MemVerge’s memory machine software will be utilized for enhanced memory performance.
“As the insatiable demand for memory both in capacity and bandwidth continues, optical connectivity will play a critical role in making it easier for compute to have scalable access to memory,” says Charles Fan, CEO and co-founder of MemVerge. “Lightelligence’s Photowave CXL over optics technology and products, together with our Memory Machine software, will enable highly efficient usage of memory for next-generation AI/ML workloads.”
Ron Swartzentruber, Director of Engineering at Lightelligence, will present “Advantages of optical CXL for disaggregated compute architectures” August 9 at 3:30 p.m. P.D.T. in Ballroom G.
Lightelligence will demonstrate Hummingbird at Hot Chips 2023 at Stanford University in Palo Alto, Calif., August 27-29. Maurice Steinman, Lightelligence’s Vice President of Engineering, will present “Hummingbird Low-Latency Computing Engine” August 29 at 11:30 a.m. P.D.T.
Visit AITechPark for cutting-edge Tech Trends around AI, ML, Cybersecurity, along with AITech News, and timely updates from industry professionals!
