Celestial AI raised $100 million to expand its photonic fabric interconnect technology platform for data centers.

Celestial AI, a developer of optical interconnect technology, has announced a successful series B funding round, raising $100 million for its Photonic Fabric technology platform. The round was led by IAG Capital Partners, Koch Disruptive Technologies (KDT), and Temasek's Xora Innovation fund. Additional participants included Samsung Catalyst, Smart Global Holdings (SGH), Porsche Automobil Holding SE, The Engine Fund, ImecXpand, M Ventures, and Tyche Partners. This brings Celestial AI's total funding to $165 million from seed through series B.

Advanced artificial intelligence models such as GPT-4 and recommendation engines require exponentially increasing memory capacity and bandwidth, creating what is known as the "memory-wall" challenge for cloud service providers and hyperscale data centers. The limitations of electrical interconnect—including restricted bandwidth, high latency, and high power consumption—hinder the growth of AI business models and advancements in the field. Celestial AI CEO Dave Lazovsky explained to VentureBeat that "the key problem going forward is memory capacity, bandwidth and data movement (chip-to-chip interconnectivity) for large language models (LLMs) and recommendation engine workloads."

To address these challenges, Celestial AI developed Photonic Fabric, an optical interconnect designed for disaggregated, exascale computing and memory clusters. The company's proprietary Optical Compute Interconnect (OCI) technology enables the disaggregation of scalable data center memory and accelerated computing. According to Lazovsky, "Our Photonic Fabric technology allows you to integrate photonics directly into your silicon die. A key advantage is that our solution allows you to deliver data at any point on the silicon die to the point of computing. Competitive solutions such as Co-Packaged Optics (CPO) cannot do this as they only deliver data to the edge of the die." Photonic Fabric delivers package bandwidth of 1.8 Tbps/mm² with nanosecond latencies—25 times greater bandwidth than CPO and with latency 10 times lower. The platform supports industry-standard protocols including CXL and JEDEC (HBM), and is compatible with interfaces like PCIe, UCIe, and proprietary interconnects.

Celestial AI targets enterprise computation for LLMs such as GPT-4 and PaLM, as well as deep learning recommendation models (DLRMs) ranging from 100 billion to 1 trillion-plus parameters. Lazovsky told VentureBeat that inefficient usage of server DRAM memory translates to "hundreds of millions (if not billions) of waste across hyperscalers and enterprises," but Photonic Fabric enables memory disaggregation and pooling to reduce memory spending and improve utilization. He explained that "by increasing the addressable memory capacity of each processor at high bandwidth, Photonic Fabric allows each processor to store and process larger chunks of data, reducing the number of processors needed. Providing fast chip-to-chip links allows the connected processor to process the model faster, increasing the throughput while reducing costs."

Broadcom is collaborating on the development of Photonic Fabric prototypes based on Celestial AI's designs, with prototypes expected to be ready for shipment to customers within the next 18 months. The funding will be used to accelerate productization and commercialization of the Photonic Fabric technology platform by expanding Celestial AI's engineering, sales, and technical marketing teams.