Overnight, NVIDIA's formidable rival AMD delivered a record-breaking performance in its data center revenue. However, the company's stock fell nearly 8% in after-hours trading, as market disappointment over the Q1 revenue guidance ($9.8 billion) cast a shadow. In the subsequent earnings call, AMD's management sought to reassure the market: the true AI volume ramp is scheduled for the second half of the year, and the path forward is now very clear. "We are very bullish when we look at the full year," emphasized Lisa Su, highlighting that the core focus for 2026 is the parallel advancement of two growth vectors in the data center: one is the continued strength of server CPUs, and the other is an anticipated "inflection point" for data center AI in the latter half of the year.
Behind the seemingly "less explosive" guidance lies an accelerating core business. CFO Jean Hu acknowledged on the call that the Q1 guidance was impacted by multiple factors, including a "significant double-digit decline" due to the gaming console cycle entering its seventh year and reduced one-time revenue in certain specific markets. However, stripping away these noise factors, the core data center business exhibited exceptional strength. AMD's primary engine is far from stalling. Lisa Su underscored a key structural highlight: "Even in the typically seasonally soft Q1, we expect both our data center GPU and server CPU businesses to achieve sequential growth." This signifies that while the PC and gaming segments are dragging on overall performance, the AI and server businesses are accelerating against the trend. Management explicitly stated that 2026 will demonstrate a "slow start, strong finish" pattern, with the real breakout point locked in for the second half.
The decisive battle in the second half hinges on the MI450 shipping on schedule and the OpenAI order being "on track." Investors are most concerned about whether AMD can continue to close the gap with NVIDIA in high-end AI chips. In response, Lisa Su provided a clear timeline, placing the company's bets squarely on the latter half of the year. Regarding the MI450 series, which competes with NVIDIA's Blackwell, Lisa Su confirmed: "The MI450 series is on track for launch and volume production in the second half of the year... revenue will begin in the third quarter and ramp significant volume in the fourth quarter as we move into 2027." Addressing a pointed question from a Morgan Stanley analyst about the progress of the OpenAI partnership—specifically, whether the deployment of "6 Gigawatts" would commence as planned—Lisa Su offered reassurance: "We obviously have a very strong relationship with OpenAI. We are planning to enter the ramp phase starting in the second half of this year and continuing into 2027. That is on track." She further revealed that beyond OpenAI, other customers are also actively seeking rapid deployment of the MI450, directly countering market concerns about AMD's reliance on a single major customer for large model training. Lisa Su emphasized, "I would also remind everyone that we have a very broad set of customers who are very excited about the MI450 series," adding that the company is "also actively engaged in discussions with a number of other customers on multi-year, scaled deployments starting with Helios and MI450."
Server CPU: Order Book Strengthens, "Especially in the Last 60 Days"; Venice Enjoys "Very High Customer Pull." Simultaneously, Lisa Su pointed out that Agentic AI is becoming a new engine for CPU demand. "The demand for high-performance CPUs is extremely high right now. This is credited to agentic workloads... when you have these AI processes or AI agents in the enterprise, they actually spin off a lot of traditional CPU tasks." Lisa Su repeatedly positioned server CPUs as another growth curve within the AI cycle: "CPUs are very important as AI continues to ramp." "We see the CPU order book continuing to strengthen over the last several quarters, especially in the last 60 days." She also highlighted an "anomalous strength" seasonally: "We are seeing server CPUs grow from Q4 to Q1, whereas typically that is a seasonally down period." Regarding new products, she mentioned the next-generation Venice in her remarks: "Customer pull for Venice is very high, and engagements are underway to support large-scale cloud deployments and broad OEM platform availability when Venice launches later this year."
AI Revenue Target "Tens of Billions of Dollars," Data Center Growth Over 60%, 2027 Technology Roadmap Targets 2nm. Looking further into the future, AMD confirmed for the first time the process technology details of its next-generation flagship chip, showcasing an aggressive catch-up posture in manufacturing technology. "The MI500 series utilizes our CDNA6 architecture, is built on an advanced 2-nanometer process technology, and features high-speed HBM4E memory. We are on track to introduce MI500 in 2027." Based on this roadmap, Lisa Su reiterated ambitious long-term targets: "We are well-positioned to grow data center segment revenue by more than 60% annually over the next three to five years and scale our AI business to tens of billions of dollars in annual revenue by 2027."
Supply and Delivery Risks: "Extensive Testing" Conducted, Claims "Will Not Be Supply Constrained." Addressing potential system integration bottlenecks at the rack-level delivery stage, Lisa Su stated, "The development of both the MI450 series and the Helios rack is progressing very well... we have done extensive testing—both at the rack level and at the chip level." She also mentioned that "customers have given us a lot of things to test, so we can do a lot of testing in parallel." On the question of whether supply might constrain the second-half ramp, her response was more direct: "We have planning at every component level... relative to our data center AI ramp, I don't believe we will be supply constrained."
China MI308: "Not Forecasting More" After Q1; MI325 License "Submitted." When asked about the sustainability of China-related revenue, Lisa Su indicated that the Q4 MI308 sales "were actually from licenses approved following our work with the government," and that "those orders actually came in very early in 2025." Regarding assumptions beyond the first quarter, she stated plainly: "We expect about $100 million of revenue in Q1. We are not forecasting additional revenue from China beyond that because it's a very dynamic situation." She added, "We have submitted license applications for MI325."
Expense Leverage and Investment Direction: 2026 OpEx Growth "Should Be Slower Than Revenue." Addressing rising expenses, Lisa Su expressed "very high conviction" in the company's roadmap but indicated that operating leverage should materialize in 2026: "We should absolutely see leverage... in our long-term model, OpEx growth should be slower than revenue growth, and we expect 2026 to be that way as well." The CFO added that 2025 investment priorities focused on the "data center AI hardware roadmap," "expansion of software capabilities," and the "acquisition of ZT Systems to add system-level solution capabilities." She noted that 2026 will see "continued aggressive investment" but "revenue growth is expected to outpace operating expense growth."
PC and Gaming: PC TAM Might See "Slight Decline"; Semi-Custom SoC to See "Significant Double-Digit Drop" in 2026. On the PC front, Lisa Su highlighted demand-side risks: "Based on everything we see today, we might see the PC TAM down slightly... given some inflationary pressures on commodity pricing, including memory." She stated the company's modeling for the full-year cadence is "second half slightly below seasonality relative to the first half." However, she simultaneously emphasized the market share narrative: "Even in a declining PC market, we believe we can grow our PC business... the focus is on the enterprise and continuing to grow in the higher-end, higher-price bands." Regarding the gaming segment, management expects the semi-custom business to be in the later stages of its cycle: the company stated that full-year 2026 semi-custom SoC revenue will "decrease by a significant double-digit percentage."
Comments