Release Date: July 30, 2024
For the complete transcript of the earnings call, please refer to the full earnings call transcript.
Positive Points
- Arista Networks Inc (ANET, Financial) reported a non-GAAP earnings per share of $2.10, showcasing strong financial performance.
- The company achieved a non-GAAP gross margin of 65.4%, driven by outstanding manufacturing disciplines and cost reductions.
- International revenue contribution was 19%, with the Americas contributing 81%, indicating strong performance in key markets.
- Arista Networks Inc (ANET) launched a rich portfolio of 800 gig products, enhancing its competitive edge in the AI and networking space.
- The company celebrated its 10th anniversary on the New York Stock Exchange and now supports over 10,000 customers with 100 million ports deployed worldwide.
Negative Points
- Operating expenses for the quarter were $319.8 million, up from $265 million last quarter, indicating increased costs.
- R&D spending increased to $216.7 million, reflecting higher new product introduction costs and increased headcount.
- International revenues decreased to 18.7% of total revenue, down from 20.1% in the prior quarter, driven by weaker performance in the APJ region.
- The company faces supply chain constraints and inflationary pressures, which could impact future financial performance.
- Deferred revenue balance increased significantly, indicating potential delays in revenue recognition and customer acceptance.
Q & A Highlights
Q: As we head into next generation GPUs with Blackwell and VL. 36 72, there's been some discussion about whether the system is maybe less modular and service components, particularly for back in networking. Can you share your views on the vendor to vendor modularity of Blackwell, particularly as it relates to networking and how that might affect Arista's positioning over the next couple of years?
A: Sure, Michael. As GPUs get faster, the dependency on the network for higher throughput increases. Our timely introduction of 800 gig products will be crucial, especially for Blackwell. The density of GPUs and their connection methods, such as PCIE and UAL, will influence this. Arista will play strongly on the scale-out front, connecting Ethernet ports into scale-out switches, regardless of modularity.
Q: The deferred revenue balance is up significantly year-on-year. Can you help us understand how to think about that number and its relation to AI opportunities?
A: Deferred revenue ebbs and flows, particularly with new products and use cases. It's not unusual to see higher deferred revenue in one quarter or year and then a dip. The balance is a mix of customers and use cases, so it's not tied to any single factor.
Q: Last quarter, you mentioned four greater area trials. Can you provide an update on those trials?
A: All four trials are progressing well, moving from trials to pilots this year. We hope to see them fully adopted next year. Additionally, we have tens of smaller customers starting pilots and trials.
Q: You mentioned a $70 billion TAM. Can you explain what is included in that TAM and how it relates to the $750 million AI networking revenue target for next year?
A: The TAM includes data center, campus, wide area, and routing markets, along with software components like CloudVision and observability. The $750 million AI networking revenue target is a subset of this larger TAM, focusing on specific AI-related opportunities.
Q: How are investments in traditional infrastructure trending, given the focus on AI?
A: Last year, there was a shift from classic cloud to AI spending. However, we continue to see a refresh cycle with customers moving from 100 to 400 gig. While AI will grow faster, classic cloud remains an important aspect of our contributions.
Q: Regarding the updated 2024 guidance, is it reasonable to expect growth to accelerate in the second half of the year?
A: We see multiple scenarios for the second half and expect some acceleration in growth. We are confident about the second half and increasingly confident about 2025.
Q: The implied growth in Q4 seems lower than previous quarters. Is this due to conservatism or other factors?
A: Our gross margin guidance remains consistent, and the lower implied growth is due to the large numbers we are working with. We aim for double-digit growth and will have more visibility as we go into Q3.
Q: What are your expectations for product deferred revenue in Q3?
A: We don't guide product deferred revenue. It varies based on customer use cases and deployment timings. AI-related deployments tend to take longer, while classic cloud deployments are quicker.
Q: Can you discuss the opportunity for refreshing front-end networking to alleviate bandwidth concerns from the back-end?
A: The goal is to demonstrate wins in enterprise and non-cloud spaces. We see a lot of activity in migrating to higher bandwidths and simplifying network environments, which aligns with our $750 million AI networking target for next year.
Q: How is customer concentration evolving compared to last year?
A: We expect Microsoft and Meta to remain greater than 10% customers. No other customer is expected to approach 10% this year or next year. The concentration may vary based on their priorities and spending patterns.
For the complete transcript of the earnings call, please refer to the full earnings call transcript.
