AMD is hosting a Second Quarter 2023 Earnings Conference Call, and Mitch Haws, Head of Investor Relations, is introducing Dr. Lisa Su, the Chair and Chief Executive Officer, and Jean Hu, the Executive Vice President, Chief Financial Officer, and Treasurer. The call is being recorded and will be available via webcast. The company will refer primarily to non-GAAP financial measures and the non-GAAP to GAAP reconciliations will be available in the press release and slides. Dr. Lisa Su and Jean Hu will also attend various conferences in late August and early September, and the third quarter 2023 quiet time is expected to begin on September 15. The discussion contains forward-looking statements that involve risks and uncertainties.

In the second quarter, AMD saw revenue decline 18% year-over-year and remain flat sequentially, with Datacenter segment revenue down 11% year-over-year and up 2% sequentially. AI cluster engagements increased significantly, with the 4th Gen EPYC CPU adoption doubling sequentially. 30 new AMD instances were launched in the quarter, with Genoa instances from AWS, Alibaba, Microsoft, and Oracle offering up to 1.9 times more performance in enterprise and cloud applications, and 1.8 times more performance per watt than the competition.

AMD-powered cloud instances are set to increase by 30% to nearly 900 by the end of the year, driven largely by new Genoa deployments. Bergamo and Genoa-X were launched in the quarter, with Bergamo offering more than double the performance of competitive offerings. These platforms are set to be launched by Dell, HPE, Lenovo, Supermicro and other large server providers in the third quarter. Despite macroeconomic uncertainty, sales of EPYC processors for enterprise servers grew sequentially, with large companies such as Banco de Brasil, BNP Paribas, Petronas, Uber, and SAP adopting them. Siena, AMD's first EPYC processor optimized for leadership, Edge server, and teleco infrastructure is set to launch this quarter, with EPYC revenue set to grow by a double-digit percentage in the third quarter.

In the second quarter, AMD made strong progress in their AI strategy by expanding their portfolio of leadership GPUs, CPUs, and adaptive computing solutions, optimizing thousands of models for their platforms, and delivering a performance and feature update to their ROCm software stack. They also expanded their partnerships with Hugging Face, PyTorch, TensorFlow, Onyx, and OpenAI Triton to accelerate deployments of AMD-based AI solutions at scale. In networking, they saw increased adoption of Pensando DPUs from major cloud providers, and EPYC and Instinct processors continued to be the solutions of choice for the world's most powerful and efficient supercomputers.

MosaicML's Instinct MI250 and MI300X GPUs are designed to be the world's most advanced accelerators for generative AI and have a large memory footprint and fast memory bandwidth. In the Client segment, Ryzen 7000 series CPU sales have grown significantly due to the launches of new notebooks from major OEMs, such as HP and Lenovo. AMD expects its Client segment to grow in the second-half of the year with the launch of Ryzen 7040 mobile CPUs, which have a dedicated AI engine, and the increased adoption of Ryzen 7000 CPUs. AI is expected to be a significant PC demand driver as software providers incorporate generative AI into their offerings.

In the second quarter, AMD's Gaming segment declined 4% year-over-year and 10% sequentially due to lower gaming graphic sales, despite strong semi-custom SoC sales due to improved retail availability and new AAA games. The Embedded segment increased 16% year-over-year and 7% sequentially, driven by demand with industrial, vision and healthcare, automotive, and broadcast customers. AMD launched the Versal Premium VP1902 adaptive SoC, Spartan UltraScale+ FPGA family, and enhanced versions of their Vivado and Vitis software platforms, as well as new AMD powered security, storage and networking solutions from HPE, Fortinet and other vendors.

In the second-half of the year, AMD expects Embedded segment revenue to decline as customers reduce their inventory levels. However, the PC market is expected to grow seasonally and the datacenter market is mixed as AI deployments expand. AMD is investing in AI related R&D, ecosystem enablement, and go-to-market investments to capture a share of the emerging AI market, and is confident that they can deliver leadership, training, and inference solutions powered by their instinct EPYC, Ryzen AI, Versal, and Alveo platforms.

In the second quarter, Intel reported revenue of $5.4 billion and diluted earnings per share of $0.58. Revenue decreased year-over-year due to lower Client segment revenue, but was flat compared to the first quarter. Gross margin was 50%, operating expenses were $1.6 billion, and operating income was $1.1 billion. Datacenter segment revenue was down 11% year-over-year due to lower third generation EPYC processor sales, but grew sequentially due to strong sales of fourth generation EPYC processors.

AMD reported lower operating income due to lower revenue and increased R&D investment. The Client segment saw a 54% year-over-year decrease in revenue, but a 35% increase on a sequential basis. The Gaming segment experienced a 4% year-over-year decrease in revenue, while the Embedded segment saw a 16% year-over-year increase. AMD generated $379 million in cash from operations and $254 million in free cash flow, with inventory increasing by $332 million due to the ramp of Advanced Technology products.

In the third quarter of 2023, Intel expects revenue to increase by 2.5% year-over-year and 6.5% sequentially. Revenue for the Client segment is expected to increase, while the Gaming and Embedded segments are expected to decline. Non-GAAP gross margin is expected to be 51%, non-GAAP operating expenses to be $1.65 billion, non-GAAP effective tax rate to be 13%, and the diluted share count is expected to be 1.63 billion shares. Intel is well-positioned to take advantage of the shift in workloads and spending patterns with its CPUs and AI investments.

Lisa Su discussed the positive market dynamics in the datacenter industry, including strong demand for CPUs, MI250 accelerators, and MI300 accelerators. She also noted that cloud spend outside of AI is softer and enterprise is weaker, but there is still customer pull for the datacenter business. She expects a large ramp in the second-half of the year, with a growth of 50% or more compared to the first-half. The company is also aggressively trying to ramp up both the hardware and software side of the MI300 programs to support AI.

Lisa Su reports that there is strong customer interest in AMD's AI solutions, and that the progress they have made with the ROCm software stack has been significant. Customers have been able to test the MI250 and MI300 hardware across a wide range of workloads, and the feedback has been positive. Customers are accessing the hardware either in AMD's labs or in their own labs for sampling.

Jean Hu confirms that the Datacenter business will have a high single-digit growth year-over-year. She also states that there will be a strong double-digit sequential growth in earnings in Q3 and a continued strong ramp in Q4. Aaron Rakers then asks how AMD has managed the supply chain side with the ramp in mind, to which Jean Hu responds that their manufacturing partners have expanded their capacity significantly.

Lisa Su explains that AMD has been investing in their supply chain and have ample supply for an aggressive ramp in the fourth quarter and into 2024. She also states that the double-digit sequential growth in Datacenter in the third quarter is primarily from EPYC, which is a combination of Genoa and Bergamo, and that the fourth quarter will see a significant ramp in revenue due to the continued ramp of Zen 4 server CPUs.

Lisa Su and Jean are discussing the competitive landscape of the client business and the path back to profitability. Lisa is pleased with the growth seen in the second quarter and the strength of their product portfolio. She believes the business is competitive, but they feel good about it.

Lisa Su states that the enterprise business is very strategic to AMD and that they are underrepresented in this area. The team is driving share gains with global corporations and ramping up Genoa. AMD anticipates that the enterprise segment will contribute to the strong second-half growth profile of their Datacenter business. They have been optimizing their R&D footprint and investing in AI, Datacenter, and other strategic priorities to generate a higher return on investment.

Lisa Su and Harlan Sur discuss the potential for AMD to develop China-specific SKUs for their MI250 and MI300 platforms to meet the demand for accelerated compute in the region. They also discuss the potential for GPU accelerator sales to reach $500 million this year, making up about 7-8% of datacenter sales.

Lisa Su explains that AMD's investments in AI are broad, encompassing the datacenter, edge, and client. She then goes on to explain that AMD's value proposition in the commercial market is its niche and differentiation, and that the company is just beginning to sample its product, giving a realistic sense of how big it can be.

AMD believes they have a strong value proposition for both training and inference workloads in the datacenter, and have increased their resources and focus on optimizing at higher level models. They are working closely with AI-centric companies to ensure that there is room for multiple winners in the exciting AI space.

In Q4, Lisa Su estimates that the largest component of the $700 million datacenter revenue increase is from the MI300 ramp. Additionally, there is a significant contribution from EPYC processor ramps with Zen 4 portfolio. Going into 2024, customer interest in MI300X is high and there are multiple customers looking to deploy as soon as possible, resulting in some lumpiness in revenue over the next few quarters.

Lisa Su stated that El Capitan would be several hundred million of the $700 million. Jean Hu then discussed that the gross margin for Q3 would be 51%, with it expected to improve from that level in Q4 due to double-digit growth in the datacenter and client business, but with a headwind from the embedded business declining in both Q3 and Q4.

Lisa Su and Jean Hu answer a question about the embedded business. They note that the core markets have held up well, but communications have been weak. For the third quarter, they expect a double-digit decline sequentially. They also do not expect the fourth quarter to be much different. Jean Hu notes that inventory days are expected to come down as they ramp up production for the fourth quarter.

Lisa Su commented that the supply chain for the MI300 product line is tight, but AMD has made commitments for significant capacity across the entire supply chain. She mentioned that co-host, high bandwidth memory, and general capacity requirements are all components of this. She also stated that AMD's goal is to make this a significant growth driver and that they are responsible for providing the supply for the demand.

Lisa Su responded to Christopher Rolland's question by stating that they are not ready to reveal revenue numbers yet, but they believe the MI300 opportunity is a multibillion dollar one. She also mentioned that 2024 is an important year for them, and that they are seeing strong customer interest from both hyperscalers and AI focused companies. She said that the performance they have seen on MI250 for large language models is good, and the early results on MI300 are also strong, with MI300X being the biggest seller for AI applications.

Lisa Su answers a question from Chris Rolland about Siena Telco, a market owned by AMD's competitor. She explains that Siena is a niche that AMD hasn't previously been focused on, but it can be used for edge applications and lower end applications that need the performance of Zen 4. She also states that AMD has an opportunity to gain market share over the next couple of years. Chris Danely then asks if the MI250, 300, etc. revenue is mostly GPU only, what kind of impact it would have on AMD's gross margin. Lisa Su responds that both MI300A and MI300X will be part of the ramp, particularly in the fourth quarter.

This paragraph discusses the company's AI business and its expected effect on gross margins at the corporate level. Lisa Su mentions that the process for customers to go from interest to purchase order is not much different from a server ramp, but customers are more willing to move quickly when it comes to AI deployments. She also notes that this agility is helping the company learn and perfect the software in preparation for a significant ramp in the coming year.

Harsh Kumar asked Lisa Su if there is a theoretical limit to the server share that AMD can get and if there are any large vendors where AMD has significantly more than 30% share. Lisa responded that the most important thing for their customers is that they have a strong roadmap that customers can count on and trust. She also noted that AMD has been underrepresented in the enterprise, but they have leadership in the market and an opportunity to continue to gain share.

Lisa Su discussed the market for generative AI solutions and estimated that it will grow to $150 billion by 2027. She also noted that it is AMD's number one strategic priority and that they are working closely with customers to optimize the balance between CPU and GPU spending. The call concluded with John thanking everyone for joining.

This summary was generated with AI and may contain some inaccuracies.