Here are my initial thoughts on Astera Labs ALAB Q4 earnings report,
The revenue that the company landed at in Q4 was decent, but I had expected more from the revenue guidance for Q1 2026. The company had guided for 245 - 253M of revenue and analysts expected 250M.
Revenue landed at 271M which was +92% yoy and +17% qoq. Guidance for Q1 revenue was 286 - 297M which would be +86% yoy and +10% qoq at the top end of the guidance range. Normally these would be pretty good numbers to report but there were a couple reasons I was expecting more,
- Management had a press release on January 22 saying Scorpio X was now in production and I expected this to give a large boost to projected Q1 revenues
- A peer in the AI networking space Credo CRDO announced a phenomenal quarterly number, although they tempered their enthusiasm on future quarters
With the press release of Scorpio X, I am now understanding that Astera has three categories to classify sales of pre-production, production, and volume production. It seems like Scorpio X is in production now, and they mention the volume ramp happens in the second half of 2026. Another encouraging sign in this report was their Leo product line is growing and expecting a production ramp in the second half of 2026 as well.
On the competition side, Credo’s impressive pre-announced revenue led me to believe Astera would have an outsized Q1 guide. However, in contrast to Astera’s enthusiasm for the ramp in the 2nd half of 2026, Credo told investors to expect mid single digits sequential growth for those same quarters.
I believe the contrasting forecasts for these two companies is because their product lines do not overlap as much as one would expect. Both companies entered the AI networking space in the past few years with AEC cables using PCIe technology. The companies started off more as direct competitors. However, their expansion product lines do not have that much overlap. For example Credo has no plans to enter the switch market that Astera has with the Scorpio product line. Also, Astera’s Leo and Credo’s Weaver products for memory are for different use cases and address different memory issues. This means that the product ramp timelines for these companies are diverging a bit. This could explain why Astera mentions a bigger ramp in the second half of 2026, and Credo describes a leveling off of growth rate.
One big announcement this quarter was that Amazon is investing in Astera further. Prior to this report Amazon held 43M of Astera stock and now they are investing up to 466M in stock with a performance clause. It is a somewhat unusual arrangement, but if Astera delivers 6.5B of potential products, these stock warrants for Amazon will exercise.
Going back I looked up how much sales Astera Labs has done over the course of the company lifetime and the number was just 1.6B. Keep in mind, Astera was only getting 11M of revenue in Q2 of 2023 to show how fast this company has scaled up. This deal with Amazon shows they are looking to purchase over 4x the amount of products that Astera has sold in the lifetime of the company! The terms of the deal go until the year 2033 which also shows the long timeline that Amazon has envisioned with Astera.
We also learned that Amazon’s Tranium 4 chip will support UALink, which is the open standard that Astera is a lead promoting member of. There is also support for UALink being used with NVLink Fusion. Additionally, AMD announced their MI 500 series will also support UALink. This sounds like the open standard that Astera prefers is gaining hold.
Overall, I found the long term vision of Astera completely intact and I am looking to keep them as a top confidence position. The guidance they gave could have been stronger, but we expect Astera to significantly outperform that guidance.