Key Highlights
- BofA identifies NVIDIA, Broadcom, and AMD as premier AI semiconductor investments with targets of $300, $450, and $280 respectively
- French AI firm Mistral secured $830M in financing to acquire approximately 13,800 NVIDIA GB300 GPUs, representing roughly $575M in hardware revenue
- Broadcom secured OpenAI as chip design collaborator and disclosed $100B+ in AI semiconductor bookings through fiscal 2027
- Aletheia Capital projects AMD’s data center business will surge from $17B in 2025 to $77B by 2028
- Worldwide cloud infrastructure expenditure reached $110.9B in Q4 2025, marking a 29% annual increase
Bank of America has designated NVIDIA, Broadcom, and Advanced Micro Devices as premier investment opportunities within the AI computing landscape, establishing specific price objectives and identifying potential challenges for each semiconductor manufacturer.
These recommendations arrive amid accelerating global investment in AI infrastructure. Worldwide cloud infrastructure expenditure climbed to $110.9 billion during Q4 2025, representing a 29% year-over-year surge, based on data from research organization Omdia. Projections indicate another 27% expansion throughout 2026.
BofA established a $300 valuation objective for NVIDIA, calculated using 28 times its projected 2027 earnings. The investment bank highlighted NVIDIA’s commanding position in AI computing and networking infrastructure as primary rationale.
Cantor Fitzgerald similarly maintained its Overweight assessment and $300 target for NVIDIA after the chipmaker’s GTC conference.
French artificial intelligence firm Mistral AI contributed to bullish sentiment. The company secured $830 million through debt financing to construct a data center facility near Paris, equipped with 13,800 NVIDIA GB300 GPUs. Industry analysts estimate this procurement could generate approximately $575 million in semiconductor revenue for NVIDIA.
Space technology venture Starcloud obtained $170 million at a $1.1 billion valuation. The firm previously deployed an NVIDIA H100 GPU into orbital space and intends to launch a second satellite featuring a complete GPU cluster within the year.
On March 16, NVIDIA unveiled its Space-1 Vera Rubin computing module, engineered to analyze data while in orbit instead of transmitting unprocessed information to Earth. NVIDIA has not disclosed a delivery timeline.
Broadcom Secures Massive AI Semiconductor Pipeline
Bank of America assigned a $450 valuation target for Broadcom, utilizing 26 times its anticipated 2027 earnings. The firm emphasized consistent double-digit profit expansion and robust free cash generation.
Broadcom recently onboarded OpenAI as a chip engineering collaborator through a multiyear agreement to jointly create 10 gigawatts of specialized AI accelerator technology. The semiconductor company maintains similar partnerships with Google, Meta, Amazon, Microsoft, and Anthropic for customized silicon solutions.
Executive leadership disclosed exceeding $100 billion in AI chip commitments secured for fiscal 2027. AI semiconductor sales are anticipated to hit $10.7 billion in the upcoming quarter.
Broadcom additionally secured a five-year, $970 million agreement with the Defense Information Systems Agency and initiated commercial shipments of its Tomahawk 6 switch processor.
AMD Pursues Aggressive Data Center Expansion
Bank of America designated a $280 price objective for AMD, emphasizing AI expansion and increasing CPU market penetration.
Aletheia Capital confirmed a Buy recommendation with a $330 target. The research firm anticipates AMD’s server CPU business will expand at a 45% compound annual growth rate spanning 2025 through 2028.
Data center revenue is expected to climb from $17 billion in 2025 to $77 billion by 2028.
AMD and Celestica introduced the Helios rack-scale AI platform. AMD additionally finalized a multi-year licensing agreement with Adeia Inc., settling all pending legal disputes between both entities.
Lisa Su received appointment to President Trump’s Council of Advisors on Science and Technology.
AMD voiced concerns regarding its consumer and gaming divisions due to escalating memory component costs.
