Global leader in purpose-built connectivity solutions for rack-scale AI.
San Jose, California
Joined March 2019
- Tweets1.2K
- Following112
- Followers2.8K
- Likes25
Astera Labs is headed to @GlobalSemi's #WISH2026! Join us to celebrate women in #semiconductors and catch breakout and poster sessions, including some from fellow Asterans. If you're going, swing by and grab a coffee. We're sponsoring your caffeine break this year! ☕
Register now: gsaglobal.org/events/event/2…
📅 Nov. 10
📍 Santa Clara Convention Center
#AsteraLabs #WISH2026 #WomenInSemiconductors #GSA #Smiconductors #WomenInTech #womeninSTEM
How many #optical fibers fit along a chip package edge? That’s shoreline density. In his blog, Jeremy Witzens explores #NPO/#CPO connectors and why isothermal glass molding stands out for rack-scale #AI: dense connections + scalable manufacturing. Read: asteralabs.com/resources/blo…
Women at Astera Labs' Network hosted a speed networking event where interns and full-time employees connected over boba and music. The room was loud and lively from start to finish, and plenty of those conversations kept going long after the event ended. #AsteraLabs
Here at Astera Labs, we believe in uplifting each other and raising the bar together - one conversation at a time. Thank you to everyone who came out - we can't wait to see you at the next one! #networking #speednetworking #empowerment
Taipei, we're #hiring a Director of Hardware Design Engineering. Join #AsteraLabs to lead the team shaping OEM platforms and rack-scale AI systems across Asia. If you want to help build what's next for AI infrastructure, apply now: job-boards.greenhouse.io/ast… #semiconductor
Everyone needs inference capacity yesterday, and a closed rack design shouldn't slow you down. In his blog, Pravin Patel explains why flexibility is the biggest competitive advantage in #AIInference, and how Orion delivers it. qrco.de/bh4YoU.
#AI #GPU #CPU #neoclouds
One rack. Five components. Meet Orion, our compute-agnostic, rack-scale reference architecture. Watch to see the five modular designs, plus more coming soon. Then, visit our page to explore each component in depth: asteralabs.com/products/orio… #GPU #CPU #Neocloud #AsteraLabs
According to McKinsey & Company, the next AI cost bottleneck may not be the chip but the link. As inference overtakes training, every idle GPU drives up cost per token. Our take: gains will come from keeping accelerators, memory and switches moving. Read: mckinsey.com/featured-insigh…
Nothing out of reach for Aries 7. Chris Blackburn's new blog explores #PCIe 7's bandwidth and signal-budget challenge. Smart Retimers, Redrivers, Cable Modules and optical solutions cover links from inches to 50+ meters. Read now: asteralabs.com/resources/blo… #AsteraLabs #PCIe7
Booth B34 is packed up, but what a week at OCP Global Summit! We showcased Orion, Astera Labs' compute-agnostic, rack-scale reference architecture designed to simplify and accelerate AI infrastructure deployment. #OCPSummit26 #AIInfrastructure #AsteraLabs @OpenComputePrj
We also raised a glass at Wednesday's happy hour and lit up the Nasdaq Tower in Times Square to mark the week. Thank you to everyone who stopped by Booth B34, asked questions, and talked connectivity with us. Can't wait to see you next year! #AsteraLabs #OCPSummit26
Want to help build the silicon behind rack-scale #AI?Astera Labs is #hiring a Senior Manager, Digital Design in San Jose, California, to lead front-end digital design for high-performance PCIe, CXL, UALink, and Ethernet devices. Apply here: job-boards.greenhouse.io/ast…
Your first in-depth look at Orion is here. Mike Hendricks walks through how our compute-agnostic rack-scale architecture lets neoclouds and enterprises choose their own GPUs and CPUs, then scale inference without redesigning the rack. #AsteraLabs #OpenByDesign
See how Orion comes together as a complete rack or as individual building blocks, from the Compute Board and Link Board to the Fabric, Compute, and Memory Trays, all managed with COSMOS. Watch the full video here: asteralabs.com/videos/first-… #neocloud #gpu #cpu #aiinfrastructure #ai
Today, Astera Labs introduces Orion, our compute-agnostic rack-scale reference architecture. Orion lets #neoclouds and enterprises choose their own GPUs and CPUs, turning racks built on open standards into token-efficient #AI factories. Read more: asteralabs.com/news/astera-l…
Six production-grade reference designs cut integration complexity and testing cycles. A high-radix 320-lane #PCIe switch fabric scales from 8 to 160 GPUs per pod. COSMOS Rack Manager and COSMOS Fabric Manager deliver fleet-wide telemetry, diagnostics, and management.
Interested? Come see a powered Orion rack at the #AsteraLabs Booth B34 at OCP Global Summit, Oct. 12-15 in San Jose: asteralabs.com/about/events/… @OpenComputePrj #OCPSummit26