Sean O'Loughlin

Sean O'Loughlin asked about the Active LED Cable (ALC) technology and Credo's relationship with Hyperloom, specifically whether a non-ramp in 2027/2028 would be due to a technology or business problem. He also inquired if the Zero-flap Optics (ZF Optics) platform targets a specific line or lane rate, or if it is agnostic across 50, 100, and 200 gig per lane applications.

Answer

President and CEO Bill Brennan explained that Credo's interest in micro-LED technology for ALCs began a couple of years ago, leading to the acquisition of Hyperloom to bring the technology and team in-house for predictable time to market. He views the current stage as an execution play, with the technology well-developed, and expressed confidence in product launch in fiscal 2027 and initial ramps in fiscal 2028, seeing no major customer acceptance barriers. For ZF Optics, Bill Brennan confirmed it can be leveraged across various lane rates, with the first product targeting 100 gig per lane (800 gig ports) and a path to 1.6T, but adaptable to 50 gig per lane if a customer requires it.

Sean O'Loughlin asked about the AEC opportunity in front-end networking and any product differences versus back-end solutions. He also sought clarification on whether a previously mentioned 'emerging hyperscaler' is now one of the three customers in volume production.

Answer

CEO William Brennan explained that front-end networks exist for both general compute and AI clusters, but AI is driving faster lane speeds (100-gig and 200-gig). He confirmed Credo's largest customer uses its AECs for both front-end and back-end connections. Brennan also clarified that the previously mentioned 'emerging' hyperscaler is now classified as a full hyperscaler due to its spending and is one of the three customers in volume production.

Sean O'Loughlin asked about the expected Q4 data center growth, specifically the 'meaningful contribution' from LPO, and how LPO deployments are envisioned to enter the market—whether project-specific and lumpy, or a more gradual integration. He also questioned the Q4 gross margin outlook, particularly the mix shift within IoT, and potential headwinds from silicon photonics/germanium foundry capacity constraints.

Answer

President and CEO Hong Hou projected approximately 10% sequential data center growth in Q4, primarily from FiberEdge, with LPO contributing a meaningful mid-single-digit percentage. He noted ACC's contribution would remain nominal in Q4. Executive Vice President and CFO Mark Lin clarified that semiconductor gross margins (SIP, data center, LoRa, PerSe) are accretive, while IoT systems and connectivity (cellular modules) have lower gross margins, driving the Q4 consolidated gross margin outlook. Hong Hou addressed foundry capacity, emphasizing long-standing partnerships and co-planning processes to ensure component availability.

Sean O'Loughlin asked about the anticipated impact of a recent competitor merger on MACOM's competitive landscape, and inquired if MACOM is observing or anticipating the effects of AI deployments and data center builds on telecom access and long-haul networks, particularly regarding increased bandwidth and two-way inference traffic.

Answer

President and CEO Steve Daly congratulated the merging companies, stating no direct impact on MACOM as they are not in the handset business and neither company is a customer or supplier. He speculated that potential fab closures could create opportunities for MACOM. Regarding AI's impact on telecom, Mr. Daly confirmed MACOM's engagement with RAN manufacturers and understanding of front-haul networks, viewing increased bandwidth as a long-term trend where MACOM's high-speed, high-frequency technology is well-positioned.

Sean O'Loughlin asked about the potential impact of a recent competitor merger on MACOM's competitive landscape, particularly outside the handset market. He also inquired if MACOM is observing or anticipating the effects of data center deployments and increased bandwidth on telecom access and long-haul networks.

Answer

Steve Daly, President and CEO, MACOM, stated no direct impact from the merger as MACOM is not in the handset market and the companies are not customers/suppliers, but noted potential opportunities from fab closures. Regarding telecom trends, Mr. Daly confirmed MACOM is working with RAN manufacturers and sees long-term potential for high-speed data movement, believing MACOM's high-speed, high-data rate, and high-frequency technologies position it well.

Speaking for Matt Ramsey, Sean O'Laughlin asked about the future of the traditional legacy video processor business and whether it could be a growth lever. He also inquired about the gross margin profile of new products on advanced nodes, considering potential foundry price increases.

Answer

CEO Fermi Wang stated that the company is no longer investing in new legacy video processors, so revenue is not expected to grow, but the rate of decline should slow significantly. CFO John Young addressed margins, stating they do not expect the profile to change significantly on new nodes. He reiterated the long-term target of 59-62%, which will be influenced more by the mix of business (like automotive) than by foundry costs.