TOKENOMICS 2030:
THE RISE OF THE 
INTELLIGENCE ECONOMY

Adoption of artificial intelligence globally is growing, causing a paradigm shift in how we interact with the world. Interactions like simple content generation, search queries, and complex coding requirements are part of this new paradigm where interactions become both inputs and outputs reliant on tokens, the fundamental unit of AI processing.

These interactions are growing at an unprecedented rate as AI becomes more deeply integrated in the tools, products and systems we leverage every day. This serves as a springboard for global economic growth directly tied to the use of tokens. Our Tokenomics outlook explores what happens when AI shifts from a useful tool to a foundational layer of intelligence operating at a global scale.

A glowing red circular object with an integrated circuit pattern is displayed against a red dotted background. The design features a central chip surrounded by radiating circuit traces, suggesting advanced digital or electronic technology. Soft highlights and shadows give the token a three-dimensional appearance.

AI interactions begin with tokens

317 Billion
— AI tokens every second1 —
The infographic uses a black background with bright red text and graphics. The label “2030” appears in the upper-left corner, while “TODAY” is positioned near the lower-right corner. The visualization consists of 312 hollow red circles arranged in a uniform grid, along with 5 solid red circles. Each circle represents one billion tokens generated per second. The five solid circles indicate the current estimated rate of global token generation, while the remaining hollow circles represent the projected increase in token generation capacity by 2030. Together, the circles illustrate the scale of growth expected between today and 2030, highlighting both the current state and the anticipated future demand.
5 Billion

317 Billion
tokens/second1

By 2030, the world could generate approximately 317 billion AI tokens every second, equivalent to roughly 10 quintillion tokens annually.1 This is more than a growth story. It is the emergence of a new digital utility: intelligence delivered at unprecedented scale.
A black-background graphic emphasizes the massive scale of annual token generation using oversized red numbers outlined in red. The number is displayed across multiple lines, beginning with a tilde symbol (~) to indicate an approximation and reading as 10,000,000,000,000,000,000, representing ten quintillion. In the lower-right portion of the graphic, bold white text states “10 Quintillion Tokens Annually¹.” The superscript “1” indicates a referenced source or footnote not shown within the cropped image. The visual uses a minimalist design with very large numerals dominating the composition to communicate the enormous projected volume of tokens processed or generated each year. The strong contrast between the red outlined numbers and the black background draws attention to the scale of the figure.

Where We are Today

Five Billion Tokens Per Second and Growing1

The foundation for this future already exists. In 2026, global AI systems are estimated to generate approximately 150 quadrillion tokens per year, or roughly five billion tokens every second.1 This scale has moved beyond experimentation and into real world productivity in search, software development, content creation, customer engagement, and enterprise workflows. As we can do more with AI, the question becomes how much intelligence the world's infrastructure will be able to produce and deliver.

Streams of AI tokens flow through major application categories, including productivity, search, content creation, customer service, coding, and enterprise solutions. The graphic celebrates the extraordinary scale of AI adoption, reaching 150 trillion total tokens processed across diverse workloads and experiences.

HOW TOKEN GROWTH HAPPENS

Helping Optimize Infrastructure Performance

The Tokenomics model begins with a simple premise: token generation is ultimately constrained by the infrastructure available to produce it.

Systems Architects design an interconnected system of compute, memory, networking, storage and software. Each of these parts of the system have distinct optimizations to help enable maximum efficiency.

Efficient use of deployed components is what drives growth in token production by:

  • More compute capability per device
  • High memory bandwidth
  • Fast networking
  • Expanded storage hierarchies
  • Ideal infrastructure efficiency

Together, these advances increase the amount of intelligence that can be generated from watts of power and dollars invested in infrastructure.

A vertical visualization of the foundational infrastructure layers that power large-scale AI systems. The graphic highlights the progression from Compute to Memory, Storage, Networking, and Efficiency, illustrating how advancements across the technology stack enable exceptional performance. At the base, the image showcases a throughput of 317 billion tokens per second, emphasizing the scale and speed required to support modern AI workloads.

THE SHIFT TO CONTINUOUS INTELLIGENCE

From Questions to Autonomous Work

The first generation of AI was largely reactive. Users asked questions, generated content, or requested assistance one interaction at a time.

As infrastructure has shifted to better service the needs of scaling inference, or generating intelligence, AI is becoming increasingly persistent, connected, and autonomous. The next generations of AI will increasingly participate in ongoing workflows and decision-making processes.

Future generations of AI systems will continue to accelerate the need to maintain context, access information, coordinate actions, retrieve knowledge, and perform tasks across multiple steps.

The result is a world where intelligence becomes increasingly continuous rather than episodic.

A black-background infographic with red and white graphics compares a traditional reactive AI workflow with a continuous intelligence system. On the left side, a vertical sequence labeled "Reactive AI" shows a linear process. A user icon appears at the top, followed by a downward arrow to "Prompt", represented by several horizontal lines. Another downward arrow leads to "Response", represented by a speech bubble icon. This section illustrates a one-time interaction where a user provides a prompt and receives a response. A large red arrow points from the left section to the right section, indicating a transition to a more advanced model. On the right side, the phrase "Continuous Intelligence" appears at the center of a circular ecosystem. Surrounding the center are five connected components: Agent (top, represented by a robot icon), Data (upper right, represented by binary digits), Actions (right, represented by a play-button icon), Applications (bottom, represented by a four-square grid), and Tools (left, represented by crossed tools). Gray curved arrows connect the components in a loop, indicating ongoing interaction and feedback among agents, data, actions, applications, and tools. The visual contrasts a single prompt-response workflow with a continuous, interconnected intelligence system that operates through multiple inputs, outputs, and feedback cycles.

THE EVOLUTION OF AI USERS

From Freemium to Enterprise-Scale Intelligence

Not all AI usage is created equal.

FREEMIUM USER

A young man in a suit sits indoors, resting his chin on his hand in a thoughtful pose. He appears to be in a modern office setting with a window in the background. Soft natural light highlights his face and formal attire. The circular crop focuses attention on his expression and contemplative mood.

Complexity

Value

Token Intensity

Freemium users represent the broadest layer of the AI economy, spanning virtually every age group and use case. For many consumers, AI is beginning to replace traditional search as the primary way information is discovered, questions are answered, and content is created. The rapid adoption of frontier AI platforms has been among the fastest technology shifts in recent history, bringing AI into everyday life for hundreds of millions of users. While individual interactions may be relatively lightweight, the scale of this audience makes it a foundational driver of the token economy.

PREMIUM USER

A woman sits at a desk working on a laptop while pointing at a large monitor displaying data dashboards. The office environment appears modern and professional with multiple screens in use. Charts, graphs, and colored bars are visible on the monitor interface, suggesting analytics or business intelligence software. The overall scene conveys focused concentration and technology-driven data analysis.

Complexity

Value

Token Intensity

Premium users are individuals and organizations willing to pay for greater capability, productivity, and convenience. This group spans consumers, creators, developers, and business professionals who seek access to advanced features such as larger context windows, enhanced reasoning, multimodal capabilities, automation, and emerging agentic experiences. As competition among AI providers intensifies, subscriptions have become a key battleground, with new features and benefits continually introduced to attract and retain these higher-value users. Premium users generate significantly more activity than the average consumer, making them an important growth engine for token demand.

ENTERPRISE USER

Four professionals gather around a round table in a modern office setting. One man stands and speaks while the others sit and listen, suggesting a team meeting or presentation. A laptop and coffee cup are visible on the table, emphasizing a contemporary work environment. The background features a decorative white wall, adding a clean and professional atmosphere.

Complexity

Value

Token Intensity

Enterprise users integrate AI directly into business operations, connecting models with proprietary data, applications, workflows, and governance frameworks. Here, AI evolves from a productivity tool into an operational platform capable of supporting decision-making, automation, and business processes at scale. Enterprise deployments often involve persistent workflows, connected systems, and increasingly autonomous operations, making them among the most infrastructure-intensive workloads in the token economy. As organizations move from experimentation to deployment, enterprise AI is expected to become a major driver of long-term token growth.

Conclusion

Scaling the Infrastructure of Intelligence

The journey from five billion tokens per second to 317 billion tokens per second represents more than a technology forecast.1 It reflects the emergence of a world where intelligence becomes a readily-available resource driving economic growth.

Reaching that scale will require advances across the entire infrastructure stack—from compute to memory, networking, storage, and software optimization. No single breakthrough creates the token economy. It is the combined evolution of the entire ecosystem.

By 2030, AI will not simply be generating more content or answering more questions. It will be helping people, businesses, and autonomous systems make decisions, create value, and operate in new ways.

Image Title: Intelligence Economy Technology Stack

Image Description: A layered conceptual diagram llustrates the components that support the intelligence economy. At the top, a white banner contains the red title “The Intelligence Economy” and the subtitle “Driven by a continuously growing flow of tokens.”
Directly below, People and Business are shown side by side, represented by a user icon and a building icon. Beneath them is a horizontal divider line leading to Agents, represented by a robot icon. This placement suggests agents act as an intermediary layer between people, businesses, and underlying AI systems.
Below the agents layer, three bidirectional vertical arrows indicate continuous two-way interaction with Global AI Platforms, which is represented by globe and brain icons positioned on either side of the label.
Under the AI platform layer is a horizontal band labeled “Data Center Infrastructure.” This infrastructure serves as the foundation supporting all higher layers.
At the bottom of the diagram, five infrastructure components are displayed from left to right, each with a corresponding icon:

  • Compute (processor chip icon)
  • Memory (memory module icon)
  • Storage (database icon)
  • Networking (network device icon

The overall visual conveys a technology stack in which people and businesses interact through AI agents, which are powered by global AI platforms that rely on data center infrastructure, including power, compute, memory, storage, and networking resources.

The intelligence economy

Driven by a continuously growing flow of tokens.

— Data Center Infrastructure —

Why Sandisk

Enabling the Infrastructure of Intelligence

As AI scales, Sandisk is a strategic provider of storage that helps turn data into intelligence. NAND flash and SSDs play critical roles throughout the AI data pipeline, from storing massive datasets and staging data for compute, to supporting the increasingly memory-intensive demands of inference.

As AI moves toward longer context and agentic workloads, KV cache is emerging as another important opportunity for flash. Extending KV cache beyond scarce high-bandwidth memory can help preserve context, reduce repeated computation, and improve the power efficiency and utilization of AI infrastructure.

From data storage to inference, NAND technology helps keep AI infrastructure fed, efficient, and productive. Sandisk directly supports the continued growth of the token economy.

Learn more about Sandisk Enterprise SSDs

1. Information above is based on Sandisk’s internal analysis and estimates that reflect current assumptions and expectations and are intended for illustrative purposes only.

This presentation includes forward-looking statements within the meaning of federal securities laws, including forecasts and expectations regarding AI, data infrastructure, related market opportunities, and the potential role and benefits of Sandisk’s NAND technology in such markets. These forward-looking statements are based on management's current expectations and are subject to risks and uncertainties that could cause actual results to differ materially from those predicted. Key risks and uncertainties include: adverse changes in global or regional economic conditions; volatility in demand for our products; difficulties or delays in product ramps, manufacturing or other supply chain disruptions; the impact of competitive products and pricing; our development and introduction of products based on new technologies and management of technology transitions; and other risks and uncertainties listed in the Sandisk’s filings with the Securities and Exchange Commission, including our Annual Report on Form 10-K filed with the SEC on August 17, 2026. You should not place undue reliance on these forward-looking statements, which speak only as of the date hereof, and we undertake no obligation to update or revise these forward-looking statements, except as required by law.

Product specifications subject to change without notice. Pictures shown may vary from actual products. 

Compare