Nvidia's Key Differentiators Include Its Ability To Expand Its Addressable Market: Analyst

BofA Securities analyst Vivek Arya maintained a Buy rating on Nvidia Corp. NVDA with a price forecast of $160 on Monday.

Arya listened to Nvidia CEO’s Computex 2025 keynote and then had a chance to virtually catch up with management (CFO) after the event.

The analyst highlighted Nvidia’s expanding portfolio, TAM (Total Addressable Market) expansion with the new NVLINK Fusion, rack-scale in full production, RTX Pro enterprise server, and sovereign AI buildout.

Also Read: Nvidia Unveils AI Tools For Humanoid Robots, Launches Quantum-AI Supercomputer To Drive Next Tech Era

Nvidia Arya’s top sector pick. The company’s expanding portfolio implies a broad set of AI products across cloud, enterprise, humanoid robots, workstations, gaming, software libraries, quantum, and 6G, all underlying what Nvidia referred to as a “multi-trillion dollar AI factory” industry, per the analyst.

He noted that TAM expansion with the new NVLINK Fusion opened Nvidia’s proprietary interconnect and networking tech to third-party custom accelerators (ASICs) and CPUs.

Arya is highly confident in rack-scale GB200 NVL 72 Blackwell Systems, and the transition path to next-gen Blackwell Ultra GB300 systems will be easier from the third quarter.

The analyst said the RTX Pro enterprise server extends AI into the enterprise, causing more competition for legacy x86.

He also highlighted some benefits from last week’s Middle-East deals in the second half of 2025, though more benefits will be seen from 2026 onward (a 5-year project), given the time required to set up new infrastructure.

As per Arya, the next significant catalyst is the fiscal first-quarter earnings call, which could be contentious, related to recent restrictions on China H20 shipments.

As per the analyst, Nvidia’s key differentiators include its ability to expand its addressable market across multiple vectors of silicon, systems, software, and services, drive scale with global supply-chain partners, and solid balance sheet to make strategic investments in the ecosystem.

With NVLInk Fusion, Nvidia can provide its proprietary interconnect fabric and switch (Spectrum) to cloud service providers for use as chiplet in systems with (modified) third-party ASICs (from partners Marvell Technology, Inc. MRVL) or extend the integration of Nvidia GPU with other Arm Holdings ARM-based CPU (Qualcomm Inc. QCOM) Arya noted. Separately, Nvidia launched Grace CPU C1, a new single-socket CPU for power efficient edge, telco, and storage systems, competing against rival x86 products from Intel Corp. INTC and Advanced Micro Devices, Inc. AMD, he added.

Although Nvidia said it was incredibly challenging, Grace Blackwell is in full production. The analyst noted that his management discussion suggests a higher level of confidence in the rollout and robust demand. Unlike GTC, it did not provide any specific metrics on cloud orders.

Nvidia is on track to upgrade to Grace Blackwell GB300 in the third quarter, which has similar architecture as GB200 but with an upgraded Blackwell chip, 1.5x more inference likely due to its 1.5x more high-bandwidth memory and 2x networking, Arya said. The analyst highlighted that management made no specific comments on gross margin progression.

Last, no new Hopper or salvage of China H20 products were written down recently; only low-volume L20 is being sold, he said. He added that China is still important at $50 billion TAM per Nvidia.

Price Action: Nvidia stock is trading lower by 0.08% to $135.30 at last check Monday.

Read Next:

• Super Micro Rolls Out AI Servers With Nvidia’s Blackwell Chips To Cut Costs, Boost Performance

Image by Piotr Swat via Shutterstock