Release Date: August 06, 2024
For the complete transcript of the earnings call, please refer to the full earnings call transcript.
Positive Points
- Astera Labs Inc (ALAB, Financial) reported record quarterly revenue of $76.9 million, marking an 18% increase from the previous quarter and a 619% increase from the same period in 2023.
- The company achieved a strong non-GAAP operating margin of 24.4% and generated positive operating cash flows of $29.8 million.
- Astera Labs Inc (ALAB) is benefiting from new product cycles across multiple technologies, contributing to growth through higher average silicon content per AI platform.
- The company expanded its Cloud-Scale Interop Lab to Taiwan and opened a new R&D center in India, indicating strategic growth and investment in infrastructure.
- Astera Labs Inc (ALAB) is seeing increasing content on next-generation AI platforms, particularly with Nvidia's Blackwell GPU architecture, which presents strong growth opportunities.
Negative Points
- The company's non-GAAP gross margin slightly declined to 78% from 78.2% in the previous quarter, with an expected further decline to approximately 75% in Q3 due to a product mix shift.
- Astera Labs Inc (ALAB) faces competitive dynamics within the PCI Gen 5 retimers market, with potential encroachment from other qualified solutions in China and the US.
- The commercialization of PCI Gen 6, which doubles connectivity bandwidth, presents new challenges for link reach, reliability, and latency.
- The ramp of the Leo CXL products is dependent on the availability of CXL-capable CPUs, with significant production ramps not expected until 2025.
- The company's growth forecast for the September quarter is based on multiple production ramps, which may present risks if these ramps do not materialize as expected.
Q & A Highlights
Q: Given the rapid rise of AI ASIC accelerators, how confident is the team in driving continued growth over the next several quarters?
A: Jitendra Mohan, CEO, expressed confidence in the company's growth prospects, citing the early stages of AI adoption and increasing CapEx targets from hyperscalers. Astera Labs has design wins across various platforms, including internally developed AI accelerators, which are ramping up alongside third-party accelerators. The company is optimistic about outpacing market growth due to rising content per GPU.
Q: What is the outlook for PCIe retimer penetration into general-purpose servers, especially with the introduction of Gen 5 PCIe SSDs?
A: Sanjay Gajendra, President and COO, noted that while growth in general compute has been slower than expected, it is anticipated to pick up in 2025 as AMD and Intel release new CPUs. Astera Labs has design wins for Aries retimers and Taurus Ethernet modules in these platforms, indicating potential revenue growth from general compute in the future.
Q: Can you discuss the competitive dynamics within PCI Gen 5 retimers and the prospects for PCI Gen 6?
A: Sanjay Gajendra highlighted Astera Labs' advantage due to its deep integration with hyperscaler customers and the COSMOS software suite. The company has several design wins for PCIe Gen 6, particularly with Nvidia's Blackwell GPUs, and is shipping pre-production volumes to support initial ramps. Astera Labs is well-positioned to transition from Gen 5 to Gen 6 as platforms evolve.
Q: What is driving the September quarter growth, and how does the mix impact gross margins?
A: Michael Tate, CFO, explained that growth is driven by Taurus ramping into 400-gig Ethernet systems, Aries modules, and increased content in AI servers. The mix shift towards hardware solutions, including Taurus and Aries modules, is expected to impact gross margins, which are projected to be approximately 75%.
Q: How does the company view the potential for PCIe over optical to become the majority of back-end connectivity?
A: Jitendra Mohan discussed the evolving AI system architectures and the role of PCIe over optical in extending connectivity reach. Astera Labs has demonstrated robust PCIe links over optical interconnects, providing tools for hyperscalers to craft solutions for data center-scale AI deployments. The company sees significant opportunities as these architectures evolve.
For the complete transcript of the earnings call, please refer to the full earnings call transcript.
