Despite recent corrections in the global semiconductor sector, Citigroup's semiconductor analysts remain bullish on AI infrastructure-related chip companies, particularly Advanced Micro Devices (AMD.US). AMD has been labeled the "king of the hill," surpassing rivals NVIDIA (NVDA.US) and Broadcom (AVGO.US).
At AMD's recent analyst day, CEO Lisa Su projected a highly optimistic outlook for AI computing chips, forecasting robust growth over the next five years. Su outlined AMD's financial targets, aiming for a "double-digit" share in the data center AI chip market, with annual revenue expected to reach $100 billion (up from $16 billion currently) within five years. By 2030, profits are anticipated to triple.
Su also predicted the total addressable market (TAM) for AI data centers—encompassing AI CPUs, accelerators, and high-performance networking products—to exceed $1 trillion by 2030, far surpassing this year's $200 billion estimate, implying a compound annual growth rate (CAGR) exceeding 40%. Additionally, AMD expects earnings per share (EPS) to rise to $20 within three to five years.
AMD stands as one of the few companies capable of competing with NVIDIA's dominant AI GPU offerings. Wall Street institutions are optimistic about AMD's potential to erode NVIDIA's near-monopoly in the AI GPU market, driving faster growth. Recent deals, including a 6GW AI GPU supply agreement with OpenAI and innovative equity-AI compute contracts, have further bolstered confidence in AMD's fundamentals and valuation.
In Q3, AMD's revenue surged 36% YoY to $9.25 billion, beating Wall Street estimates of $8.7 billion, driven largely by its data center segment. This division provides competitive GPUs for AI workloads and high-performance CPUs for data centers. Major partnerships with OpenAI, Oracle, and the U.S. Department of Energy reflect growing demand for AMD's MI-series AI GPUs, which compete directly with NVIDIA's Blackwell GPUs in AI model training and inference.
Unlike NVIDIA, AMD is also a leading supplier of GPUs and CPUs for PCs and data center servers. Intel (INTC.US) recently highlighted strong demand for AI-enabled laptops and server CPUs, but AMD continues to gain market share from its long-time rival.
Rising prices for high-performance memory products (DRAM/NAND), OpenAI's $1 trillion AI infrastructure deals, and bullish forecasts from TSMC, Samsung, and SK Hynix have reinforced the long-term bullish narrative for AI infrastructure, including GPUs, ASICs, HBM, and cooling systems.
Citigroup analyst Christopher Danely reaffirmed a "Buy" rating for AMD with a $260 target, citing strong EPS growth potential. While Broadcom remains a core AI holding, concerns linger over its AI ASIC business clarity. NVIDIA, despite leading the AI infrastructure boom, is losing some investor favor due to moderating EPS growth expectations.
Intel's investor sentiment has improved on optimism around its foundry business and server CPU recovery, though profitability in chip manufacturing remains uncertain. Meanwhile, memory leader Micron (MU.US) faces skepticism about returning to peak margins, and wafer fab equipment spending may exceed $120 billion, signaling slower growth.
In the analog semiconductor space, Analog Devices (ADI.US) stands out as the sole bright spot amid sluggish demand and inventory challenges. Texas Instruments (TXN.US), however, is a prime short target due to downward EPS revisions.
ADI, a leader in analog and mixed-signal chips, serves industries like EVs, aerospace, and industrial automation, with clients including Boeing and Siemens.