Wall Street is ending 2023 on a high note. The technology-heavy Nasdaq Composite gained about 43% in 2023. It may rise even higher in the coming months, driven mainly by positive economic indicators such as higher-than-expected corporate earnings and third-quarter gross domestic product (GDP) growth that surpassed most forecasts.
A potential economic recovery makes it easier for technology companies to access capital for disruptive innovations. Not surprisingly, being composed of many innovative-technology stocks, the Nasdaq Composite will benefit significantly in the case of an uptick in economic activity in 2024.
Artificial intelligence (AI) has been the hottest investment theme in 2023, and is poised to remain a major trend in 2024. Improvement in the economic landscape may drive up share prices of many AI stocks in the coming months. Hence, it makes sense to pick up small stakes in some fundamentally strong AI stocks with well-proven monetization models. Here's why Advanced Micro Devices (AMD 2.37%) and Super Micro Computer (SMCI 8.89%) fit the bill.
Advanced Micro Devices
The first name on my AI stock list is leading chip designer Advanced Micro Devices (AMD 2.37%) -- a stock that has gained almost 130% in 2023. This semiconductor giant needs no introduction, especially since the company has been frequently in the news for its advances in the field of graphic processing units (GPUs) and central processing units (CPUs).
AMD expects demand for AI infrastructure in cloud, enterprise, embedded, and personal computing use cases to grow rapidly in future years. The company has also estimated the target addressable market of the data-center accelerator market to grow annually at more than 70% and reach $400 billion by 2027.
AMD's recently launched Instinct™ MI300 family of accelerators, optimized for power-intensive AI and high-performance computing (HPC) workloads in data centers, seem well positioned to capitalize on this growing opportunity. Chief Executive Officer Lisa Su claimed that while MI300 chips are comparable to Nvidia's H100 chips in training large language models (LLMs), they are 1.4 times better in performance for inferencing (real-time running) models. Further, the company is also striving to ensure a sufficient supply of MI300 chips to meet the rising demand from cloud and enterprise segments. AMD expects MI300 revenue to surpass the $2 billion target in 2024.
Besides MI300 chips, AMD is also focusing on offering a broad range of other energy-efficient GPUs, CPUs, and adaptive computing solutions for training and inferencing LLMs. The company is also working to develop networking solutions that can orchestrate across multiple GPUs for AI applications.
AMD has further developed a robust, open, proven, and developer-friendly software ecosystem for GPU computations called the ROCm (Radeon Open Compute) software stack. This includes frameworks, libraries, compilers, drivers, developing tools, and APIs (application programming interfaces) that support multiple programming models. ROCm is further supporting the adoption of AMD's chips in the AI and high-performance computing (HPC) markets.
Besides strength in its AI business, AMD's core server CPU business and personal computing business are also showing signs of improvement. In the third quarter, the company's fourth-generation Epyc server processor revenue was up sequentially by more than 50%. The company also witnessed rising demand for its Genoa and Bergamo chips from hyperscale customers. Furthermore, the company reported a solid 50% sequential growth in its client segment in the third quarter.
Considering the future growth potential in the AI business and the strength of its core offerings, AMD can be a great stock to buy ahead of more Nasdaq gains.
Super Micro Computer
The second stock on my list is Super Micro Computer, a leading provider of high-end server and storage systems, which has emerged as a major beneficiary of the explosive growth in the AI market. The company reported a surge in demand for its LLM-optimized AI platforms, especially for Nvidia's HGX-H100 server building block integrated into its servers, in the first quarter of fiscal 2024 (ended Sept. 30).
SMCI is also gearing up to monetize its new AI-optimized server offerings based on Nvidia's GH200 Grace Hopper Superchip, AMD's MI250 and MI300 family of chips, and Intel's Gaudi 2 CPU. Being the first to market advanced servers with the latest silicon chips gives the company a solid edge against the competition.
Super Micro has been quite successful in differentiating its products from other mass-produced servers, mainly due to its "building block" or modular approach of assembling proprietary chips and networking technologies into servers. Unlike traditional servers, which require a significant overhaul to suit different applications and technology upgrades, SMCI's highly configurable servers enable customers to easily scale or replace components, at minimal cost. The company's liquid cooling technology also allows better energy utilization and thermal management of its datacenter server and storage solutions. Considering the high-power consumption and thermal challenges posed by AI-optimized servers, SMCI's energy efficiency solutions are helping bring down data-center costs. The company is now expecting 20% of the total data-center deployments to opt for its liquid cooling technology.
Besides the AI training market, Super Micro is also benefiting from rising demand for servers in telecommunication-optimized edge computing products and AI inferencing platforms.
Despite the many pros, Super Micro currently trades at only 2.2 times trailing-12-month sales -- quite reasonable considering that analysts on average expect the company's revenue to grow 51% year over year to $10.76 billion in fiscal 2024 (ending June 30, 2024). Hence, Super Micro seems to be a compelling pick now.
