ALAB revenue surges 154%! Another winner emerges in the AI chip interconnect race!
Nvidia and AMD provide the computing power, but when tens of thousands of GPUs are working simultaneously, who handles data transmission between the chips?
Astera Labs (ALAB) is a major player in this field.
Q3 2026 earnings report
Revenue: $585 million, up 154% year over year
Adjusted EPS: $1.32
GAAP gross margin: 72.3%
Adjusted operating margin: 44.7%
Highlights: While revenue grew rapidly, the company maintained a high level of profitability.
01】Why is ALAB growing so fast?
The competition among AI data centers is no longer just about GPU performance.
As AI clusters grow in scale, high-speed data transmission, signal integrity, and interconnect efficiency are becoming increasingly important.
ALAB focuses on PCIe, CXL, and high-speed connectivity solutions, benefiting from upgrades to AI server architectures.
02] What’s the difference between ALAB and CRDO?
Both companies benefit from demand for AI interconnects, but they have different areas of focus.
☛ALAB: PCIe/CXL interconnects, signal connectivity, and system-level solutions.
☛CRDO: High-speed SerDes, active electrical copper cables (AECs), and data center connectivity technologies.
Their businesses are not exactly the same, but both play important roles in AI computing infrastructure.
3] Where are the real risks?
154% growth is impressive, but investors should pay closer attention to:
① Can orders from major customers be sustained?
② Can high gross margins be maintained?
③ Will growth slow over the next few quarters?
④ Do current valuations already reflect overly optimistic expectations?
My view:
AI investment is spreading from GPUs across the entire data center supply chain.
The more intensive the computing, the more important interconnect technology becomes.
ALAB’s earnings report demonstrates this market’s growth potential, but high-growth companies also tend to face higher valuation expectations.
Companies worth watching include not only Nvidia and AMD, but also those addressing bottlenecks in AI infrastructure.
Which do you favor more, ALAB or CRDO?
