Newsroom

NVLink connectivity solutions will further bolster Astera Labs’ Intelligent Connectivity Platform, expanding options for hyperscalers to deploy high-performance, scale-up networks

– SANTA CLARA, CA, U.S. – May 19, 2025 – Astera Labs, Inc. (Nasdaq: ALAB), a global leader in semiconductor-based connectivity solutions for AI and cloud infrastructure, today announced it is collaborating with NVIDIA to provide scale-up connectivity solutions for the new NVIDIA NVLink Fusion ecosystem.

“We have a long history of early engagement and industry firsts with NVIDIA, and we are excited to continue this collaboration to advance the NVLink Fusion ecosystem,” said Sanjay Gajendra, President and COO, Astera Labs. “To keep pace with rapidly growing AI models which have reached trillions of parameters, AI infrastructure providers must adopt scale-up platforms that deliver both high performance and high efficiency. By adding NVLink solutions to our Intelligent Connectivity Platform, we will enhance optionality and time-to-market for our hyperscaler customers and expand our market opportunity.” 

Next-generation AI workloads demand high bandwidth, low latency scale-up connectivity with native memory semantic support to synchronize hundreds of accelerators at rack-scale. Astera Labs’ NVLink solutions will expand its Intelligent Connectivity Platform, which seamlessly integrates PCIe®, CXL®, and Ethernet silicon and hardware solutions with its COSMOS (COnnectivity System Management and Optimization Software) suite for unparalleled data center visibility and optimized performance. This will empower hyperscalers to efficiently deploy heterogeneous compute-based AI infrastructure with NVIDIA GPUs and custom accelerators in their data centers.

“NVIDIA’s accelerated computing platform has been supported by Astera Labs for multiple generations, including HGX, MGX, and NVL72 with its PCIe connectivity solutions,” said Ashish Karandikar, Vice President of SoC Products, NVIDIA. “NVIDIA’s collaborations with industry leaders enable the NVLink Fusion ecosystem to deploy high-performance AI factories that can quickly scale and meet the demands of intensive workloads for model training and agentic AI inferencing.”

Astera Labs and NVIDIA’s deep collaboration extends over multiple generations of NVIDIA platforms, with Astera Labs’ Aries PCIe/CXL Smart DSP Retimers deployed in volume across NVIDIA Hopper and NVIDIA HGX platforms, and in hyperscalers’ customized NVIDIA rack-level solutions. Most recently, Astera Labs demonstrated the industry’s first end-to-end PCIe 6 interoperability with Scorpio P-Series Fabric Switches, Aries 6 Retimers and an NVIDIA Blackwell GPU at NVIDIA GTC 2025. Scorpio P-Series Fabric Switches have also been integrated with the NVIDIA MGX platform for PCIe 6-ready modular designs. Astera Labs looks forward to the addition of NVLink solutions to its scale-up connectivity portfolio, building on its tradition of delivering interoperable solutions that meet hyperscalers’ requirements for AI infrastructure.

Resources:

• Blog: Unlocking AI Performance: The Role of Scale-Up Interconnects in Accelerated Computing
• Press Release: Astera Labs Optimizes Connectivity for NVIDIA Blackwell-based MGX Platforms at Scale
• Video: Optimizing NVIDIA MGX Platforms with Scorpio Smart Fabric Switches
• Video: Industry’s First End-to-End PCIe 6 Interop with NVIDIA Blackwell