Do We Need More Blockspace? Why Next-Gen Chains Are Building for Demand That Doesn’t Exist Yet

TL;DR

• Critics argue that new blockchains add blockspace that nobody uses, pointing to low utilization across most Layer-1 networks.
• Solana was once the textbook example of this critique. Cheap, abundant blockspace later enabled DePIN, compressed NFTs, and high-frequency on-chain activity.
• Demand often follows supply in infrastructure. Roads, bandwidth, and cloud compute all scaled before their dominant use cases existed.
• Next-generation chains like Solana, Sui, Monad, and NEAR target emerging categories. These include AI (micropayments, commerce, verification), real-time gaming, on-chain order books, and machine-to-machine settlement.
• Everstake operates validator infrastructure across several of these networks. We hold a conviction in their architectural direction, while maintaining compliant, institutional-grade operations forward-looking for agentic volumes.

The Excess Blockspace Argument

The central take is simple: the industry already has more capacity than it uses. Critics look at average transactions per second across major chains and see:

• Ethereum settles a fraction of its theoretical ceiling. 
• Most L2s run well below capacity. 
• New high-throughput L1s often post double- or low triple-digit TPS under real load.

The “ghost chain” label emerged from the initial usage expectation and reality. A fast chain with no applications is treated as evidence that throughput itself is a rather misleading metric. The argument continues: if existing blockspace is underutilized, building more only dilutes activity across fragmented networks. This view treats blockspace as a finished commodity, similar to disk storage.

The argument demonstrates a specific, narrow reading of market structure. Activity tends to concentrate where liquidity and users already exist. On this view, dominant chains compound their advantage, and challengers rarely catch up. Critics then ask why new networks keep shipping when the demand picture looks flat.

Utilization data appears to support the skeptics at first glance. The table below shows how theoretical capacity compares with observed usage across several well-known networks.

Network	Theoretical TPS (peak)	Observed live TPS (typical)	Launch year
Ethereum	~30	~15	2015
Solana	~65,000	~4,000	2020
Sui	~297,000	~400 to 900	2023
Monad	~10,000	early mainnet phase	2025

The numbers seem to confirm the claim. They also miss what infrastructure history usually looks like in the early years.

Solana’s Case Study: Demand Followed Supply

Solana is the clearest counterexample to the excess blockspace thesis. 

• In 2020 and 2021, commentators routinely dismissed the chain as a marketing story. 
• The network promised 65,000 TPS while real usage stayed modest. 
• After the FTX collapse in late 2022, many declared the chain dead. Key applications broke, diminishing the sentiment further. 

However, the tables turned. Abundant, cheap blockspace turned out to be a precondition for categories that did not exist at scale before. When fees are fractions of a cent and confirmation is sub-second, application design space widens. 

Developers stopped asking “can we afford to put this on-chain” and started asking “what becomes possible if we do”.

Three categories illustrate the developer exploration on Solana:

1. DePIN. Helium migrated from its own L1 to Solana in April 2023. Hivemapper, Render, and others followed. By 2024, Solana DePIN represented a combined fully diluted value above $10 billion. None of this was feasible on an expensive blockspace.
2. Compressed NFTs. State compression made minting millions of NFTs economically workable. That unlocked loyalty programs, ticketing, and gaming assets at a scale that older chains could not support.
3. High-frequency on-chain activity. Order books, market making, Solana DeFi summer and arbitrage bots came on-chain once fees and latency crossed a threshold. Infrastructure like ShredStream reduced block-propagation latency further, which made Solana viable for latency-sensitive flow.

Supply conditions the imagination of builders. When you tell developers they have effectively unlimited, near-free blockspace, they build things that only make sense under those conditions. Nobody forecasts these applications in advance because the pre-existing chain could not host them. The underlying pattern matches broader infrastructure history:

• Interstate highways preceded the logistics economy. 
• Broadband preceded streaming video. 
• Cloud compute preceded the modern SaaS stack. Capacity was the impulse for demand categories to form around it.

To learn how this infrastructure translates into delegation review our comprehensive Solana report and the ShredStream product overview.

Demand Categories That Don’t Exist at Scale Yet

The most interesting demand categories for next-generation chains are not on-chain today in any meaningful volume. They are either in early pilots or waiting on infrastructure that can support them. Four stand out.

AI micropayments

Autonomous agents that pay for data, compute, and inference need settlement rails with near-zero cost, sub-second finality, and reliable throughput. Today’s chains can handle this in a demo or theoretically. They cannot handle millions of agents transacting continuously. 

The agentic wave predicted for 2026 through 2028 will require infrastructure that treats micropayments as the base unit, not the exception.

Agent-to-agent commerce

A new class of “headless” merchants is emerging. 

• Services with no checkout page, storefront or sales team. 
• Agents read schemas, send requests, pay, and receive outputs in a single exchange. 
• Stripe and Tempo’s MPP marketplace processed over 34,000 transactions in its first week, with fees as low as $0.003. 
• Coinbase’s x402 is processing roughly $1.6 million per month in agent-driven payments. 

Traditional payment processors struggle to underwrite merchants with no website and no legal entity. Programmable stablecoins on open networks remove that friction.

Verifiable agent identity

As agents transact, govern systems, and act on behalf of humans, every transaction needs a cryptographic receipt that shows who acted, under what authority, and who bears liability when something fails. This is the premise behind emerging KYA (know your agent) standards, signed credentials linking an agent to its:

• principal, 
• permissions, 
• constraints, 
• reputation. 

Non-human identities already outnumber human employees in financial services by roughly 100 to 1. Infrastructure designed for human-paced throughput cannot carry this at volume.

Real-time gaming

On-chain gaming has been promised for years. The blocker has always been latency and fee predictability. A player cannot wait several seconds for a sword swing to confirm. They cannot pay $0.10 per action in a session with hundreds of actions. Networks that finalize in 400 milliseconds and charge fractions of a cent per transaction change that equation.

On-chain order books

Decentralized exchanges have mostly used automated market makers because order books were too expensive to maintain on-chain. Faster chains could make central limit order book designs workable. Professional market makers can quote and cancel at rates comparable to centralized venues, but with on-chain transparency and settlement.

Machine-to-machine settlement

Supply chains, energy grids, and IoT networks increasingly require automated value transfer between devices. 

• A smart meter paying a utility per kilowatt-hour. 
• A vehicle paying a toll or a charging station. 
• This category overlaps with AI agents and requires infrastructure that is boring in the best sense: fast, cheap, and always available.

The table below matches each emerging category with the infrastructure properties it requires.

Demand category	Primary requirement	Secondary requirement
AI micropayments, commerce, verification	near-zero fees	sub-second finality
Real-time gaming	low latency under load	fee predictability
On-chain order books	high throughput	parallel execution
Machine-to-machine settlement	continuous availability	deterministic confirmation

None of these categories exist at the scale their advocates describe. That is precisely the point. Infrastructure built for today’s demand locks in today’s ceiling. This is the case for building for the future, which could often become another guessing game.

Sui, Monad, and the Next Generation

Several chains are building explicitly for these emerging categories. Their architectures diverge, but their premises converge: current blockspace is the constraint, and relaxing it unlocks new behaviors.

Sui uses an object-centric data model with parallel execution built around it. Independent objects can be modified simultaneously without consensus on total order. The network has demonstrated 866 TPS under live conditions and 297,000 TPS in controlled testing. Its 2026 roadmap includes protocol-level private transactions, the USDsui stablecoin, and DeepBook enhancements targeting margin trading and revenue-sharing features. The combination reads as a deliberate push into institutional payments and trading use cases. 

Monad launched mainnet on November 24, 2025. It targets 10,000 TPS with 400–millisecond block times and sub-second finality while preserving full EVM bytecode equivalence. The team comes from Jump Trading, and the architecture reflects that background. Optimistic parallel execution, deferred execution, superscalar pipelining, and a custom state engine called MonadDB make up the core stack. A hard fork in March 2026 introduced a reserve balance precompile and a linear memory model for the EVM. Monad’s thesis is that high-frequency trading logic belongs on an EVM chain that developers already know.

NEAR has shifted its narrative toward AI and agentic applications. Its Intents product treats user requests as structured objects that resolvers compete to fulfill. This fits the pattern of AI agents that express goals rather than individual transactions.

The table below compares Sui, Monad and Near primary optimization targets.

Chain	Execution model	Primary target demand	Mainnet year
Sui	object-centric parallelism	payments, trading, gaming	2023
Monad	optimistic parallel EVM	high-frequency trading, DeFi	2025
NEAR	intent-based routing	AI agents, cross-chain	2020

Superior technology does not automatically produce adoption. Plenty of elegant architectures have failed to attract users. 

The point is narrower: these architectures are the only ones that could plausibly support the demand categories described above. Judging them by current usage assumes the demand environment is static. It isn’t.

Everstake’s Infrastructure Thesis

Everstake validates on several of these networks because the infrastructure thesis is credible. 

Everstake is supporting Solana, Sui, Monad, NEAR, and other next-generation chains. Everstake also runs certified compliant infrastructure that meets institutional requirements, including transparent reporting, geographically distributed nodes and operational procedures aligned with common regulatory frameworks.

There’s no way to predict which next-generation chain will capture the most activity. We look for networks whose architectures plausibly support categories that do not yet exist at scale, run by teams with the execution track record to ship at production quality. Validator participation is how we commit to that view concretely.

Chain selection is an analysis-heavy task, and innovative architecture does not guarantee success.

Some teams miss their delivery timelines. Some miss the market. Some ship products that are ahead of their time and find no demand. Strong technical ideas and good intentions are not enough on their own.

A project needs a team with a delivery track record, a clear roadmap, a workable economic model, and favorable market conditions. Survival precedes adoption.

—Everstake’s R&D Team.

FAQ

Is the excess blockspace argument wrong? 

Not exactly. Most current blockspace is underutilized at this moment. The argument becomes wrong when it concludes that future demand will resemble current demand. Infrastructure history suggests capacity often precedes the use cases that justify it.

Why did Solana escape the ghost chain narrative? 

Accessible, abundant blockspace enabled categories that older chains could not host. DePIN, compressed NFTs, and high-frequency on-chain flow grew specifically because fees and latency fell below thresholds that made these applications viable.

What makes Sui architecturally distinct? 

Sui uses an object-centric data model. Transactions that touch independent objects execute in parallel without requiring global ordering. This fits workloads with many concurrent users modifying separate state, which describes gaming, payments, and agent activity.

How is Monad different from existing EVM chains? 

Monad delivers full EVM bytecode equivalence with optimistic parallel execution, approximately 1-second finality, and a target of 10,000 TPS. Developers deploy existing Solidity contracts without modification while the execution layer handles concurrency underneath.

What does “agentic demand” mean for blockchains? 

It refers to AI agents acting autonomously and settling value on-chain. An agent paying for data, compute, or inference generates transaction patterns very different from human users. Volumes can be higher, amounts smaller, and timing less predictable.

Why does Everstake validate on multiple next-generation chains? 

Because infrastructure conviction does not require picking a single winner. Each of these networks targets a plausible category of future demand. Validator operations let us support the networks directly while maintaining compliant, institutional-grade service for delegators who want participation across the group.

Disclaimer:

This article is for informational and educational purposes only. It does not constitute financial, investment, legal, or tax advice, and should not be relied upon as the basis for any investment decision. Nothing in this article is a forward-looking statement, prediction, or guarantee of future performance. Protocol roadmaps, technical specifications, and market conditions are subject to change. Readers should conduct their own research and consult qualified professionals before making any staking, investment, or infrastructure decisions.