NVIDIA (NVDA) Company Report

Trainium3 delivers 2.52 petaflops of compute and supports 144 GB HBM3e memory with 4.9 TB/s bandwidth for complex AI workloads.
Trainium3 UltraServers scale to 144 chips per server and can cluster up to one million chips for massive AI training clusters.
AWS will offer AI Factories services, enabling clients to deploy dedicated AI infrastructure within their own data centers.
Amazon announced Trainium4 chips integrating Nvidia’s NVLink Fusion technology for high-speed rack-scale interconnects, boosting AI performance and reducing deployment risks.

2 days ago

Bloomberg

Nvidia nears finalization of $100B OpenAI deal

NVDA

New Projects/Investments

Nvidia is negotiating a $100 billion investment deal with OpenAI, with terms yet to be finalized.
The agreement would deploy at least 10 gigawatts of Nvidia systems to OpenAI, enough to power over 8 million U.S. homes.
This potential deal would be in addition to Nvidia's existing $500 billion in chip bookings through 2026, as confirmed by CFO Colette Kress.
Nvidia shares rose approximately 2.6% following the announcement, reflecting investor confidence in the company's AI investments.

2 days ago

Fortune

NVIDIA discusses AI-driven data center growth and product roadmap

NVDA

Product Launch

New Projects/Investments

NVIDIA forecasts $3–4 trillion data center infrastructure overhaul by 2030, driven by CPU-to-GPU transition and AI/agentic AI workloads.
Introduced Grace Blackwell rack-scale systems with seven-chip co-design, maintaining competitive lead via full-stack hardware and CUDA software; Vera Rubin GPU taped out for 2H26 with significant performance uplift.
Confirms ability to maintain mid-70s % gross margins into next year through improved yields, cycle times, and cost management despite rising HBM costs.
Inventory and purchase commitments rose by $25 billion, supporting anticipated revenue growth; current $500 billion Blackwell and Vera Rubin orders for 2025–26, with additional potential from direct OpenAI and Anthropic deals.
Capital allocation prioritizes supply capacity funding, shareholder returns (buybacks, dividends), and strategic ecosystem investments, with selective M&A.

2 days ago

Transcript

NVIDIA outlines AI compute demand and product roadmap at UBS conference

NVDA

Guidance Update

Product Launch

New Projects/Investments

NVIDIA expects a $3 trillion–$4 trillion data center infrastructure market by 2030 driven by the transition from CPUs to GPUs for accelerated computing.
The company highlighted its Grace Blackwell rack-scale systems (200, Ultra and 300 series) and emphasized that its full-stack CUDA software ecosystem sustains its competitive lead.
NVIDIA detailed a $500 billion planned deployment of Blackwell/Vera Rubin through 2026 and noted a 10 GW OpenAI LOI (~$400 billion)—with definitive terms still under negotiation and current fulfillment via CSP partners.
CFO Colette Kress reaffirmed mid-70s gross margins for next year as Vera Rubin ramps and cited a $25 billion increase in inventory and purchase commitments as indicative of strong near-term revenue growth.

2 days ago

Transcript

NVIDIA discusses AI infrastructure outlook at UBS conference

NVDA

Guidance Update

Product Launch

CFO Colette Kress outlined a multi-phase shift from CPU to GPU accelerated computing, forecasting $3 – $4 trillion of data-center infrastructure spending moving to AI by the end of the decade.
NVIDIA has released Grace Blackwell 200/Ultra/300 series for full rack-scale deployment and maintains its performance lead through a tightly integrated hardware/software stack.
Over 50 percent of NVIDIA’s revenue comes from cloud service providers; the company is negotiating a 10 GW LOI with OpenAI (≈ $400 billion) and continues collaborations with Anthropic via Microsoft’s CSPs.
The company achieved and expects to sustain mid-70 percent gross margins, driven by improved cycle times, yields, and cost management ahead of Vera Rubin’s second-half ramp.
Inventory and purchase commitments grew by $25 billion, reflecting robust demand; supply-demand balance is managed daily through purchase orders and capital planning.

2 days ago

Transcript

NVIDIA and Synopsys announce strategic engineering partnership

NVDA

New Projects/Investments

NVIDIA and Synopsys have formed a multi-year strategic partnership to integrate Synopsys’ engineering software with NVIDIA’s accelerated computing, AI technology and Omniverse digital twins to transform system-level design, EDA and CAE workflows.
The collaboration targets 10x to over 1,000x speed-ups in workloads such as computational lithography, logic and circuit simulation, fluid dynamics and AI-based physics emulation.
NVIDIA will commit $2 billion in equity to Synopsys as a demonstration of long-term commitment; both companies are allocating engineering teams to co-develop CUDA-accelerated and AI-infused tools.
The non-exclusive partnership leverages Synopsys’ expanded customer reach (post-Ansys acquisition) and NVIDIA’s presence across all major clouds and OEM systems, aiming to expand the market from chip design to a multi-trillion-dollar product-design industry.

3 days ago

Transcript

NVIDIA and Synopsys announce strategic engineering partnership

NVDA

New Projects/Investments

NVIDIA and Synopsys unveiled a multi-year, non-exclusive partnership to integrate Synopsys’ EDA/IP tools with NVIDIA’s accelerated computing, AI technologies (CUDA, physical AI) and Omniverse digital twins to transform system-level design workflows.
The collaboration aims to deliver 10× to over 1,000× speed-ups in core engineering simulations—ranging from computational lithography and logic simulation to fluid dynamics and AI-driven physics emulation—so that weeks-long tasks can complete in hours.
NVIDIA will make a $2 billion equity investment in Synopsys to demonstrate commitment and provide optionality for joint R&D, without exclusivity on GPU procurement.
The partnership leverages NVIDIA’s global GPU footprint—across clouds, OEMs, on-prem and edge—to enable Synopsys solutions to run wherever customers require accelerated compute.
By extending from semiconductor design to sectors such as automotive, industrial, aerospace and drug discovery, the alliance targets a multi-trillion dollar total addressable market beyond traditional chip R&D.

3 days ago

Transcript

NVIDIA and Synopsys announce strategic partnership

NVDA

New Projects/Investments

NVIDIA and Synopsys entered a non-exclusive, multi-year strategic partnership to integrate Synopsys’ engineering software with NVIDIA’s CUDA-accelerated computing, AI technology, and Omniverse digital twins.
NVIDIA will invest $2 billion in Synopsys to jointly accelerate EDA, SDA, CAE, computer-aided drug discovery, and digital twin workflows.
The collaboration targets 10×–1,000× speed-ups in core engineering workloads—shrinking simulations from weeks to hours and enabling system-level, real-time digital twins.
It expands the total addressable market from the several‐hundred‐billion-dollar semiconductor sector to a multi-trillion-dollar global R&D industry, with key tool accelerations prioritized by 2026.

3 days ago

Transcript

Nvidia price target lifted to $250 by Morgan Stanley

NVDA

Guidance Update

Morgan Stanley maintains an Overweight rating on Nvidia while raising its price target from $235 to $250, reflecting confidence in the company’s AI-driven growth.
The consensus from 56 analysts shows an average price target of $250.57 (high of $432.78, low of $138.00), implying a 39.65% upside from the current $179.43 share price.
Morgan Stanley deferred raising its revenue forecasts despite Nvidia’s management projecting $500 billion in revenue over five quarters, choosing to validate the outlook through independent checks first.
Industry observers stress Nvidia’s AI dominance, with Wedbush’s Dan Ives remarking, “it’s Nvidia’s world, everyone else is paying rent,” underscoring the company’s leading market position.

3 days ago

CNBC

SoftBank sells Nvidia stake to fund OpenAI investments

NVDA

New Projects/Investments

SoftBank founder Masayoshi Son emotionally sold the company’s entire 32.1 million Nvidia shares in October to raise capital for AI investments, saying he “was crying to sell Nvidia shares”.
The sale generated $5.8 billion, used to support SoftBank’s $22.5 billion financing commitment to OpenAI due by year-end.
The firm reported its strongest quarter since summer 2022, with profits more than doubling to 2.5 trillion yen ($16.6 billion) primarily from paper gains on its OpenAI stake.
As part of its deepening AI strategy, SoftBank is acquiring Ampere Computing and collaborating with Hon Hai on a large-scale Stargate data center project.

3 days ago

Bloomberg

Earnings Transcripts

Quarterly earnings call transcripts for NVIDIA.