A decade ago, NVIDIA was a name mostly known to gamers. Today it is arguably the most important company in the artificial-intelligence economy: the supplier whose chips train nearly every major AI model, and one of the few stocks whose quarterly earnings can move the entire market. Its climb from a graphics-card maker into a multi-trillion-dollar bellwether has become the clearest single proxy for how much capital the world is pouring into AI, which is why traders far outside the semiconductor industry now watch it as closely as any index.
That now includes crypto traders. NVDA sits at the intersection of AI, semiconductors, and macro risk appetite, and on BloFin you can take a position on it as tokenized NVIDIA (NVDAX) without a traditional brokerage account, around the clock rather than only during US market hours. If you trade equities, crypto, or tokenized assets, understanding what actually sits behind the NVDA ticker, from where its revenue comes to what could threaten it, gives you a concrete edge when reading its price action. This guide breaks down NVIDIA's business, technology, financials, and relevance to traders on BloFin.
What is NVIDIA?
NVIDIA Corporation (ticker: NVDA) is a semiconductor and software company headquartered in Santa Clara, California. Founded in 1993, the company originally designed computer graphics processors for PC gaming. Over the past decade, those same GPUs turned out to be ideally suited for parallel workloads like neural network training, and NVIDIA pivoted hard into accelerated computing, deep learning, and artificial intelligence infrastructure. The company now drives artificial intelligence and data centers alongside PC gaming.
That pivot paid off in extraordinary fashion. NVIDIA's fiscal year 2025 revenue hit $130.5 billion, a 114% increase over the prior year, with data center sales accounting for $115.2 billion of the total (source: NVIDIA Q4 and FY2025 results). Its market capitalization reached $1 trillion in May 2023, crossed $4 trillion to become the first company to surpass that mark, and then exceeded $5 trillion on October 29, 2025. As of January 2025, NVIDIA's market cap stood at $3.66 trillion before resuming its climb. By mid-2026, market cap sits in the $4.7 to $4.9 trillion range, fueled by demand from cloud service providers, agentic AI systems, and hyperscale data centers.
For traders, NVDA is one of the most actively traded mega-cap stocks globally. Daily volumes are measured in hundreds of millions of shares, and the stock reacts sharply to product launches, earnings, and regulatory shifts. On BloFin, you can gain price exposure via tokenized NVDA (NVDAX); read our guide on what tokenized NVIDIA (NVDAX) is and what you actually own for a full breakdown of how that works.
Corporate identity: Origins, founders, and headquarters
NVIDIA was founded in April 1993 by Jensen Huang, Chris Malachowsky, and Curtis Priem. CEO Jensen Huang, who co-founded the company at age 30 after working at LSI Logic and AMD, has led NVIDIA for over three decades without interruption. That continuity is rare in tech and has shaped the company's willingness to make long-horizon bets on new computing paradigms.
The company is headquartered in Santa Clara, CA, in the center of Silicon Valley. NVIDIA employs tens of thousands of people across global R&D centers, from Austin to Taipei to Tel Aviv. The company does not own fabrication plants; it designs chips and outsources manufacturing to partners like TSMC. For a deeper look at that dependency, see the dedicated article on NVIDIA's supply chain and TSMC dependency.
The path from startup to trillion-dollar AI infrastructure company was not smooth. In 1996, NVIDIA laid off half its employees to survive after its first product failed to gain traction. That near-death experience shaped the company's culture of aggressive reinvention, from graphics to general-purpose GPU computing to full-stack AI platforms.
From graphics to AI: NVIDIA's core business model
NVIDIA designs chips, systems, and software platforms. It does not own fabs. Instead, it invests in intellectual property, reference hardware designs, and complete AI stacks that span silicon, networking, operating frameworks, and developer tools. NVIDIA has a full technology stack for artificial intelligence infrastructure, covering everything from chip design to cloud deployment software.
The company's original focus was computer graphics for PC gaming. GeForce GPUs became the standard for high-performance gaming in the late 1990s and 2000s. Over time, NVIDIA expanded into professional visualization (workstation graphics for designers, engineers, and architects), high-performance computing for scientific research, and eventually data center accelerated computing for AI workloads.
By fiscal year 2025, the revenue mix had shifted dramatically. The Compute & Networking segment generated roughly $116.2 billion, or 89% of total revenue. The graphics segment contributed $14.3 billion, or about 11%. Gaming, once the company's bread and butter, now represents a fraction of total sales. NVIDIA sells discrete GPUs like GeForce for gaming, NVIDIA RTX workstation cards for professional visualization, and data center accelerators like A100, H100, and Blackwell-class GPUs to hyperscalers, cloud computing providers, and enterprises. NVIDIA supplies data-center infrastructure combining GPUs and networking for AI at a scale no other chip designer matches.
Business segments: GPUs, Compute and Networking, and beyond
NVIDIA reports results across two following segments: a Graphics segment and a Compute & Networking segment.
The Graphics segment covers:
GeForce GPUs for gaming, including desktop and laptop discrete GPUs sold through add in board manufacturers and original equipment manufacturers
Enterprise workstation graphics cards (RTX / Quadro lineage) for professional visualization and rendering
Related multimedia software, including GeForce Experience, Studio drivers, and creator-focused tools for content production and computer graphics
The Compute & Networking segment is where most of the growth sits:
Data center GPUs (A100, H100, H200, Blackwell) that serve as compute engines for training large language models and running inference at scale
DGX and HGX computing and networking platforms, sold as turnkey AI systems to enterprises and cloud service providers
Networking gear inherited from the Mellanox acquisition: InfiniBand switches, Spectrum-X Ethernet, and NVLink fabric for linking thousands of GPUs
CUDA, cuDNN, and AI framework libraries that form the software layer of NVIDIA's compute networking stack
Automotive platforms and embedded SOCs (Jetson, DRIVE) also fall under Compute & Networking, generating $1.7 billion in fiscal 2025, up 55% year over year. For a granular breakdown of each segment's revenue and trajectory, see our article on NVIDIA's business segments explained.
GPU technology: From NV1 missteps to RTX leadership
NVIDIA's first graphics accelerator was the NV1, released in 1995. It used quadrilateral-based rendering instead of the triangle primitives that Microsoft's DirectX API standardized. The mismatch made the NV1 largely incompatible with the emerging PC gaming ecosystem, and sales were poor. That failure forced NVIDIA to pivot.
The company launched the RIVA 128 graphics card in August 1997, a product designed around Direct3D triangle rendering from the ground up. It shipped over a million units in its first four months. NVIDIA Corporation invented the graphics processing unit in 1999 with the GeForce 256, the first consumer chip to handle hardware transform and lighting on the GPU itself. That branding, "GPU," stuck and redefined how the industry talked about graphics hardware.
Subsequent generations built on that foundation. NVIDIA's GeForce 10 series launched on May 6, 2016, using the Pascal architecture to deliver a generational leap in performance per watt. The RTX 2080 GPUs were released on September 27, 2018, introducing real-time ray tracing and the first version of DLSS (Deep Learning Super Sampling) to consumer graphics. NVIDIA's Ampere microarchitecture was announced on May 14, 2020, pushing both gaming and data center performance forward.
These advances in computer graphics processors turned out to have a second life: the same massively parallel architectures that render millions of pixels per frame are ideally suited for the matrix math at the core of deep learning.
Accelerated computing and CUDA: The shift beyond graphics
Accelerated computing means offloading parallel workloads from CPUs to specialized processors, primarily GPUs, that can handle thousands of operations simultaneously. Instead of running tasks one at a time on a general-purpose CPU, accelerated computing splits work across thousands of GPU cores, delivering orders-of-magnitude speedups for the right workloads.
NVIDIA's CUDA platform, introduced in 2006, was the catalyst. CUDA allows developers to use GPUs for general-purpose parallel computing, writing code in familiar languages like C++ and Python rather than graphics-specific shaders. NVIDIA's CUDA enhances software development for AI and scientific applications, and the ecosystem of libraries (cuDNN for deep learning, cuBLAS for linear algebra, RAPIDS for data analytics) became the de facto standard for GPU-accelerated research.
By the time deep learning exploded in the 2010s, CUDA already had years of adoption among researchers in physics, computational biology, and quantitative finance. That head start created a self-reinforcing cycle: more developers wrote CUDA code, more frameworks (TensorFlow, PyTorch) optimized for CUDA, and more buyers chose NVIDIA GPUs. The company has also increased open-source support in recent years, open-sourcing Linux kernel GPU modules and some AI model components, though the core CUDA runtime remains proprietary.
Artificial intelligence, deep learning, and agentic AI
In the early 2010s, deep learning researchers discovered that NVIDIA GPUs, which are highly parallel processors suitable for neural network training, could accelerate model training by 10x to 100x compared to CPUs. Modern AI requires substantial computational resources for training, and NVIDIA's data center GPUs became the default choice. NVIDIA's high-performance data center chips are essential for generative AI systems, powering training and inference for large language models, image generators, and multimodal AI.
NVIDIA launched the world's most powerful AI supercomputer in 2023, built around DGX systems. NVIDIA's H100 GPUs are in high demand for AI projects, with the H100 delivering a performance of 1,979 TFLOPs for AI workloads. NVIDIA's Blackwell architecture aims to enhance AI capabilities further, purpose-built for the next era of reasoning-heavy and agentic AI, systems that can plan, use tools, and act autonomously across data sources.
NVIDIA's AI data center revenue is projected to reach $1 trillion by 2028, according to industry estimates. The company facilitates the development of AI applications through its enterprise software, including NVIDIA AI Enterprise, NeMo for building custom AI models, and a generative AI model library that includes the Edify model for image and video generation. NVIDIA sells not just chips but full AI infrastructure stacks: hardware, SDKs, frameworks, and cloud services that enterprises use to deploy artificial intelligence solutions at data center scale.
Data center, compute networking, and cloud computing
NVIDIA's fastest-growing revenue stream centers on data center GPUs and systems sold into hyperscale and enterprise data centers. In fiscal Q1 2026 (the quarter ended April 27, 2025), data center revenue reached $39.1 billion, up 73% year over year, out of $44.1 billion in total revenue (source: NVIDIA Q1 FY2026 results). Cloud service providers account for a large share of data center revenue. That same quarter also carried a $4.5 billion charge tied to excess H20 inventory and purchase obligations under new China export-licensing rules, a reminder that policy can hit results directly.
The 2019 acquisition of Mellanox Technologies for $6.9 billion gave NVIDIA a networking segment it previously lacked. Mellanox's InfiniBand and Ethernet products let NVIDIA build end-to-end compute networking solutions, connecting thousands of GPUs into unified clusters. Networking revenue grew roughly 105% year over year in recent periods, driven by NVLink fabric, Spectrum-X Ethernet for AI, and InfiniBand switches.
On the cloud computing side, NVIDIA hardware is available through every major cloud platform. Cloud-based DGX and GPU instances let users rent AI compute rather than building out their own infrastructure. This model lowers the barrier for enterprises, independent software vendors, and researchers who need access to AI infrastructure without capital expenditure on physical servers.
Key markets and use cases: Gaming, pro visualization, and Omniverse
NVIDIA GPUs power high-end PC gaming and are crucial for advanced gaming applications. GeForce GPUs remain the benchmark for PC gaming performance, and NVIDIA's RTX technology is used for professional visualization and rendering across tools like Adobe Premiere, Autodesk Maya, and Unreal Engine. Gaming revenue reached $11.4 billion in fiscal 2025, a modest 9% increase; the gaming market is mature, but the install base is massive.
Beyond local hardware, NVIDIA's cloud-gaming service streams games from remote servers to user devices, letting players access high-fidelity titles on low-power devices like tablets and smart TVs.
NVIDIA's Omniverse is a platform for building industrial metaverse applications. It enables physically accurate digital twins and real-time 3D collaboration, used in industrial design, robotics simulation, and virtual factory planning. BMW, Siemens, and other manufacturers use Omniverse to simulate production lines before building physical infrastructure. These platforms blend computer graphics with AI, using accelerated computing to simulate and visualize complex physical environments.
Automotive, autonomous vehicles, and robotics
NVIDIA provides platforms for robotics and autonomous systems, including DRIVE and Jetson. The DRIVE platform powers driver assistance, in-car infotainment, and autonomous driving systems for automotive manufacturers and robotaxi programs. Automotive platforms generated $1.7 billion in fiscal 2025 revenue, with growth accelerating as electric vehicle solutions increasingly depend on onboard AI compute.
Jetson and Isaac platforms power autonomous machines beyond cars: warehouse robots, delivery drones, surgical systems, and smart-city infrastructure. NVIDIA extends its technology to healthcare and industrial simulation through these edge AI platforms, connecting factory-floor robotics back to data center AI training pipelines.
The automotive markets and robotics segments remain small relative to data center, but they represent a long-term bet on autonomous vehicles, physical AI, and the convergence of simulation, deep learning, and real-world deployment.
Financial profile, valuation scale, and market status
NVIDIA sits in the Electronic Technology / Semiconductors sector and trades on NASDAQ under the ticker NVDA. The company's market capitalization growth trajectory tells the story of AI demand:
May 2023: market cap reached $1 trillion
2024: crossed $2 trillion, then $3 trillion
October 29, 2025: exceeded $5 trillion, the first company to hit that mark
August 2026: $5.32 million
As of August 19, 2026, NVIDIA's stock price is $219.74. A 10-for-1 forward stock split completed on June 7, 2024, brought the per-share price into a more accessible range for retail traders. NVDA shares are among the most actively traded equities globally.
Key metrics traders watch include the price-to-earnings ratio (roughly 30 to 40x in recent quarters depending on forward vs trailing), quarterly net income growth, and data center revenue acceleration. Fiscal 2025 operating income was $81.5 billion, with diluted earnings of $2.94 per share. The dividend yield is minimal; NVIDIA reinvests heavily in R&D. For a deeper look at historical price cycles and NVDA price patterns, see the dedicated article on NVIDIA stock price history and cycles.
Regulation, controversies, and the competitive landscape
NVIDIA operates in a regulatory environment that has grown more complex as its chips became central to AI and national security debates.
US export controls have restricted sales of certain NVIDIA data center GPUs (A100, H100) to China since 2022. NVIDIA designed modified chips (A800, H800, H20) for compliance. In January 2026, the Commerce Department revised licensing policy to allow H200 exports to approved Chinese buyers under controlled conditions. For a full analysis, see the article on China and export-control risk to NVDA.
In November 2025, the US Department of Justice charged four individuals with conspiring to illegally export NVIDIA GPUs to China, involving 3.89 million US dollars in illicit proceeds (source: US Department of Justice), and lawmakers subsequently pressed NVIDIA on its compliance oversight.
Past controversies include the GTX 970 memory specification dispute (2015), criticism of the GeForce Partner Program (2018), and a 2022 SEC settlement over inadequate disclosure of cryptomining revenue's contribution to gaming segment sales.
NVIDIA's failed attempt to acquire Arm Holdings for $40 billion collapsed in 2022 under regulatory pressure from the UK, EU, and US antitrust authorities.
Competitive threats come from AMD's MI-series data center GPUs and from custom silicon efforts by hyperscalers (Google TPU, Amazon Trainium, Microsoft Maia). A detailed comparison lives in the article on NVIDIA's competitive moat vs custom silicon and AMD.
NVIDIA's role in the global AI and data center economy
NVIDIA's hardware and software stack underpins much of the global AI boom. GPUs are deployed in hyperscale data centers run by companies like Microsoft, Google, Amazon, and Meta, as well as in sovereign AI projects where governments build national compute capacity. Enterprises run on-premise AI clusters using DGX systems for tasks ranging from drug discovery to financial modeling.
The broader economic ripple effects are substantial. Demand for new data centers has triggered a construction boom, with billions invested in power infrastructure, cooling systems, and high-bandwidth networking. Data centers currently consume an estimated 1 to 2% of global electricity, and that share is growing. NVIDIA argues that accelerated computing reduces total energy consumption per unit of compute compared to CPU-only approaches, because GPUs finish workloads faster and allow servers to idle sooner.
NVIDIA's narrative centers on "AI factories," digital twins via Omniverse, and robotics as productivity multipliers across industries from healthcare to logistics. The company positions itself not just as a chip vendor but as the platform layer for the next era of computing, analogous to what Intel was for the PC era.
NVIDIA and tokenized exposure: Why it matters to BloFin traders
NVDA's liquidity, volatility, and role as an AI infrastructure company bellwether make it appealing to traders beyond traditional equities markets. On BloFin, you can access NVIDIA xStock (NVDAX) price exposure through tokenized assets, without needing a brokerage account or dealing with traditional settlement times.
Tokenized NVDA (NVDAX) tracks the price of the underlying NVDA share. You can trade NVDAX spot (NVDAX/USDT) or the NVDA perpetual contract (NVDAXUSDT) for leveraged exposure. For step-by-step instructions, see the guide on how to buy tokenized NVIDIA on BloFin, and for leveraged trading specifics, read how to trade NVIDIA with leverage.
BloFin's infrastructure supports this type of trading activity with a high-performance trading engine, deep liquidity pools, unified accounts that let you manage spot and perpetual positions from one balance, and institutional-grade security partnerships with Fireblocks and Chainalysis.
How NVIDIA news and fundamentals affect NVDA and NVDAX
NVIDIA's stock price responds to a specific set of catalysts:
Quarterly earnings reports, particularly data center revenue growth and forward guidance
Product launches like new GPU architectures (Blackwell announcements at GTC, for example)
Export-control policy changes that expand or restrict sales to China
Large cloud computing deals, such as multi-billion-dollar GPU orders from hyperscalers
Shifts in macro sentiment around AI spending and semiconductor demand
When NVIDIA has reported sharp data center revenue growth, the shares have moved hard; when export policy tightened in late 2022, the stock sold off on concerns about lost China revenue. These same drivers affect NVDAX, since tokenized NVDA is designed to track the underlying share price. For a detailed breakdown, see what moves NVIDIA's stock price, and for reading quarterly results, NVIDIA earnings explained.
How NVIDIA fits into a broader trading strategy on BloFin
NVIDIA represents one component in a diversified trading approach. If you already trade crypto majors like BTC and ETH, adding NVDAX exposure lets you express a view on AI, semiconductor, and data center trends without leaving the crypto ecosystem.
Correlations between NVDA and assets like Bitcoin shift over time depending on macro conditions and risk appetite. When both assets trade as "risk-on" bets, they can move together; in other environments, they diverge. The article on NVIDIA stock vs Bitcoin correlation covers the data in detail.
Practical uses for NVDAX on BloFin include:
Going long NVDAX/USDT if you believe AI infrastructure spending will continue to accelerate
Shorting the NVDAUSDT perpetual to hedge existing tech-heavy portfolio exposure
Using basis or cash-and-carry structures, as described in the hedging NVDA with the perpetual guide
BloFin features that support these strategies include unified accounts, cross-margin, copy trading (follow experienced NVDAX traders), and trading bots that can be configured around NVDA / NVDAX price levels. See BloFin features explained for a full platform walkthrough.
The bottom line
NVIDIA has evolved from a graphics chip startup that nearly went bankrupt in 1996 into a central player in artificial intelligence, data center infrastructure, and accelerated computing. Its GPUs train the AI models that power search engines, chatbots, autonomous vehicles, drug discovery, and industrial simulation. Its networking products connect those GPUs into clusters that process data at scales that would have been unimaginable a decade ago. That is why NVDA is one of the most watched tickers in global markets and why its market capitalization has reached multi-trillion-dollar levels.
Understanding NVIDIA's business, from its revenue concentration in data center sales to the regulatory risks around export controls, helps you interpret NVDA and NVDAX price moves with more precision. It does not eliminate risk; NVIDIA faces real threats from custom silicon competitors, supply chain concentration, and potential slowdowns in AI capital expenditure. But informed trading beats guessing.
If you are ready to act on this knowledge, start with the guides on what tokenized NVIDIA (NVDAX) is, how to buy tokenized NVIDIA on BloFin, and how to trade NVIDIA with leverage. Each one walks you through the specific steps to turn understanding into a trading plan.
Frequently asked questions
What does NVIDIA actually do?
NVIDIA designs GPUs, data center systems, networking hardware, and AI software platforms. It started in computer graphics for gaming and now generates most of its revenue from data center accelerated computing and AI infrastructure sold to cloud providers and enterprises.
Why is NVIDIA important for AI and data center workloads?
NVIDIA GPUs are massively parallel processors that accelerate neural-network training and inference. Combined with CUDA software and networking products, NVIDIA provides a complete stack for building and running AI models at scale, and no other vendor currently matches that breadth.
How is NVDA stock different from tokenized NVDAX on BloFin?
NVDA is a share of NVIDIA Corporation traded on Nasdaq. NVDAX is a tokenized asset designed to track the NVDA share price, and it does not grant voting rights or direct ownership of NVIDIA shares. See the comparison in tokenized NVIDIA versus real NVIDIA stock.
Does NVIDIA pay dividends, and do NVDAX holders receive them?
NVIDIA pays a small dividend, but its yield is minimal because the company prioritizes R&D investment. Whether NVDAX holders receive dividend equivalents depends on the tokenized asset's structure; see the article on whether NVIDIA pays dividends and how that applies to NVDAX.
What risks should I consider before trading NVDAX?
Key risks include NVIDIA's revenue concentration in data center (roughly 88 percent of total sales), export-control restrictions that could limit China sales, dependency on TSMC for manufacturing, and the possibility that hyperscalers develop custom AI chips that reduce reliance on NVIDIA hardware. NVDAX also carries the risks inherent in tokenized assets, including platform and counterparty risk.
Is NVDAX available around the clock?
Yes. Unlike NVDA on Nasdaq, which trades during US market hours, NVDAX on BloFin is available around the clock, so the token price can move on weekends and overnight in response to global news. See the guide on NVDAX trading hours.
Who founded NVIDIA and who runs it now?
Jensen Huang (also written Jen-Hsun Huang), Chris Malachowsky, and Curtis Priem founded NVIDIA in 1993. Jensen Huang remains CEO, making him one of the longest-serving chief executives in Silicon Valley.
Researched and written by the BloFin Academy editorial team with AI-assisted drafting. Updated August 2026. NVIDIA's fiscal-2025 results, its June 2024 ten-for-one stock split, and its market-capitalization milestones are drawn from the company's SEC filings and investor communications. Revenue, market-capitalization, and valuation figures are described in rounded, dated terms because they move every earnings cycle; check NVIDIA's latest quarterly report and a live market source for current numbers.
This article is educational and general in nature, not financial or investment advice, and it gives no price target. Both cryptocurrencies and tokenized stocks like NVDAX are volatile, and prices can fall as fast as they rise. Tokenized products like NVDAX add platform and custodial counterparty risk and convey no shareholder rights, voting power, or direct ownership, and trading with leverage can produce losses that exceed your initial deposit. Nothing here is a recommendation to buy, sell, or hold NVIDIA stock or tokenized NVIDIA. Do your own research, and consider speaking with a licensed professional before making financial decisions. BloFin does not provide investment advice.