TERADYNE, INC Q3 FY2024 earnings call

Management Statement and Operational Highlights

• Greg Smith summarized Q3 results, noting Semi Test business outperformed expectations with Memory and SoC driven by AI applications. Cloud AI drove Compute Revenue, particularly in networking devices. Robotics continued to outpace its peer group despite a difficult macro industrial environment.
• Sanjay Mehta provided the financial summary for Q3, with sales at the high end of guidance, non-GAAP EPS above guidance. He discussed Q4 outlook with sales expected between $710 million and $760 million, and gross margins estimated at 59.5% to 60.5%. He also revised TAM estimates for SoC and Memory, and discussed Robotics growth targeting 15%-20% above the industrial automation peer group.

Segment Performance

• Semi Test: Q3 sales were $543 million. SoC contributed $393 million, Memory $150 million. Memory TAM was revised to approximately $1.4 billion, up from the prior estimated range of $1.2 to $1.3 billion. The SoC TAM for the Compute market segment was increased by 200 million from $1.6 billion to $1.8 billion in 2024.
• Robotics: Revenue was approximately $89 million, flattish sequentially and up 3% year-over-year. UR contributed $73 million, MiR contributed $15 million.

Guidance

• Q4 sales expected to be between $710 million and $760 million, with gross margins estimated at 59.5% to 60.5%.
• Revised SoC TAM for Compute market segment to $1.8 billion in 2024, up from $1.6 billion. Memory TAM revised to $1.4 billion, above prior estimate range.
• Robotics growth targeted to be 15%-20% above the industrial automation peer group, with expected growth in 2024 of 5%-10% considering macroeconomic factors.

Risks

• Macro industrial environment has persisted for over two years, impacting Robotics business.
• Constraints in HBM supply and advanced packaging for VIP Compute devices.
• Seasonality impacts, with Q1 revenue expected to be roughly 5%-10% down quarter-over-quarter, affecting gross margins.

Question and Answer

Q: Mehdi Hosseini asked about System-Level Test opportunities and how escalating test time for AI Blackwell could open doors for Teradyne.

A: Greg Smith responded that SLT is a backstop to limit growth, and Teradyne's future is tied to vertically integrated producers and networking.

Q: Vivek Arya inquired about Robotics growth and how Teradyne can achieve 20%-30% growth.

A: Greg Smith discussed Robotics' efforts to broaden offerings and outgrow peers despite weak end markets.

Q: Tim Arcuri asked about green shoots in industrial automation, like Amazon's white paper on cobots.

A: Greg Smith noted excitement about AI applications in cobots for various industries but didn't comment on specific customer engagements.

Q: Samik Chatterjee asked about TAM growth in 2025 and seasonality in AI pieces.

A: Sanjay Mehta discussed seasonality in Q1 and visibility challenges, while Greg Smith mentioned incremental improvement in other market segments under a strong AI market.

Q: Brian Chin asked about industrial test TAM revision and VIP Compute TAM constraints.

A: Greg Smith explained better visibility into China's indigenous supplier TAM and constraints like HBM supply and advanced packaging for VIPs.

Q: Christopher Muse asked about Robotics growth drivers and OPEX leverage in 2025.

A: Greg Smith discussed seasonality in Robotics and Sanjay Mehta mentioned strategic planning and operating leverage considerations.

Q: Toshiya Hari asked about mobile SoC Test TAM and ownership of Robotics business.

A: Greg Smith provided thoughts on mobile SoC Test TAM through cycle and discussed Teradyne's unique value proposition in robotics.

Q: Krish Sankar asked about Semi Test OPEX increase and HBM market share.

A: Greg Smith stated OPEX increase is cyclical and related to supporting VIP customers, and discussed HBM market share opportunities.

Q: Gus Richard asked about complexity growth in mobile processors and capacity utilization.

A: Greg Smith discussed shifts in modem supply and utilization trends, noting increased utilization but unable to provide exact quantified numbers.