Every year-end brings a season of reflection and introspection from the year gone by, and we have the opportunity to take the learnings that came from it, framing them into a canvas for the year to come. Against a global background that often created back-and-forth macroeconomic uncertainties, Astera Labs steadily and surely forged its path forward. If I could characterize the year with one word, the word I would choose would be “growth.” Growth along organic and inorganic dimensions, growth through products and solutions being launched and reaching production status, growth through application-level work and partnering deeply with our ecosystem, and finally growth through the most tangible metrics of employee numbers, expansion of global headquarters, and revenue.
From strengthening our Intelligent Connectivity Platform to deepening partnerships across the AI rack, we pushed the boundaries of what’s possible for next-generation AI infrastructure. These achievements were recognized by the Global Semiconductor Alliance and our industry peers when Astera Labs was named the Most Respected Public Semiconductor company. And this momentum was further validated as our co-founders – Jitendra Mohan, Sanjay Gajendra, and Casey Morrison – won the EY Entrepreneur Of the Year National Award! They will represent Astera Labs on the global stage this coming May as finalists for the global award.
Strategic Expansion into Photonics
A defining milestone for us this year was the acquisition of aiXscale Photonics, which significantly expanded our capabilities in optical connectivity. This strategic move strengthens our ability to deliver scalable, high-performance interconnect solutions that meet the growing demands of rack-scale AI architecture, while accelerating our roadmap for optical and electrical convergence.
Advancing Product Innovation at Rack-Scale
We continued to raise the bar for purpose-built connectivity with key product milestones.
- Ramping Production of our PCIe 6 Portfolio – Our purpose-built PCIe 6 connectivity portfolio ramped production to fast-track deployment of modern AI platforms at scale for our AI and cloud server customers.
- Introducing Aries 6 Smart Gearbox – We expanded our Aries portfolio with the industry’s first purpose-built PCIe 6 gearbox solution that intelligently bridges the performance gap between the latest PCIe 6 devices and the existing PCIe 5 ecosystem. Customers can deploy mixed-generation systems while ensuring full utilization of PCIe 6 high-speed lanes.
- Enabling the Industry’s First CXL Deployment with Leo – In 2025, we took an important step forward for CXL adoption! Our Leo CXL Smart Memory Controllers smashed the memory wall on Microsoft Azure M-Series VMs. Microsoft’s private preview of Azure M-Series VMs represents the industry’s first cloud service provider deployment of CXL-attached memory, addressing the growing demands of memory-intensive workloads.
- Expanding our Connectivity Portfolio with Custom Solutions – With new custom offerings, we are directly addressing hyperscalers’ needs for custom connectivity solutions tailored to their unique architectural requirements and workload-specific performance goals.
- Delivering Rack-Level Management with COSMOS Developer Kit – We released our COSMOS Developer Kit which allows our customers to test, debug, and manage connectivity infrastructure at scale. With user-friendly test and debug capabilities, our customers can accelerate their validation workflows.
- Expanding our Cloud-Scale Interop Lab with Scorpio – We enhanced our interoperability testing infrastructure by integrating Scorpio Smart Fabric Switches, further enabling ecosystem validation at scale.
Expanding Our Ecosystem Leadership
We continued to play a leading role in shaping open, interoperable AI infrastructure by actively collaborating across the industry and advancing open standards-based innovation and custom solutions.
- Drove Rack-Scale AI Ecosystem Momentum –At OCP Global Summit, we showcased our comprehensive collaborations with ecosystem partners and the breadth of our solutions built with PCIe, UALink, Ethernet, CXL, and OpenBMC standards.
- Advanced UALink Ecosystem Leadership – Along with our silicon, system, and hyperscaler partners, we continue to advance the UALink ecosystem as an open, scale-up interconnect for next-generation AI infrastructure. Our leadership was amplified through 15+ speaking engagements around the world.
- Joined the NVIDIA NVLink Fusion Ecosystem – We are now collaborating with hyperscaler partners to develop custom solutions designed to support NVLink connectivity.
- Joined Arm® Total Design – We are collaborating with Arm to combine our Intelligent Connectivity Platform with Arm Neoverse CSS to enable chiplet solutions that meet the growing demand for custom AI infrastructure.
Investing in Global Growth
To support our expanding operations and our growing customer base, we made significant investments in our global infrastructure – crowned by the move to new headquarters in the heart of Silicon Valley.
- New Global Headquarters in San Jose, CA – Designed to accommodate up to 900 employees, our new HQ features expanded state-of-the-art R&D labs, collaborative workspaces, and an executive briefing center (EBC).
- Expanded Global Footprint – We also added new locations in Irvine, CA, Ho Chi Minh City, Vietnam, Seattle, WA, and Singapore to complement our existing R&D centers. These new locations enable us to tap into diverse engineering talent, deepen regional customer engagement, and scale innovation globally.
Developing the Next Generation of Leaders
In 2025, we welcomed more than 130 interns across our global locations, underscoring our commitment to cultivating future engineers and early-career professionals. Our internship program provides hands-on experience across hardware, software, systems, and business functions, giving participants meaningful ownership to solve real-world challenges at the forefront of AI infrastructure.
Strengthening Ties and Giving Back to Our Communities
Giving back locally is a core tenet of our culture. This year, our team members around the world supported their communities through hands-on volunteer initiatives.
Our teams partnered with organizations including Step30 International, RAFT (Resource Area For Teaching), Happy Hollow, and BeautifySJ to support STEM education, environmental conservation, and community art initiatives. In addition, we also volunteered with the Family Giving Tree Backpack Drive and Holiday Wish Drives, donating hundreds of backpacks, fulfilling hundreds of wishes for students and families in need.
Looking Ahead
As we begin 2026, we remain focused on leading the AI Infrastructure 2.0 era, advancing the industry with purpose-built connectivity that unlocks higher performance, efficiency, and scale. We are grateful to our customers, partners, and teams for their trust and collaboration throughout 2025, and we’re energized to build on this momentum to drive even greater innovation and impact together.