HyperEVM Under Fire: Is Hyperliquid's Composable Finance Layer Dying or Evolving?

The debate over HyperEVM has been one of the most heated in crypto circles over the past few months. Critics have called it “dead” or “dying,” while proponents point to recent surges in activity as proof of resilience. The most accurate assessment lies somewhere in between: HyperEVM, as the application layer of Hyperliquid, has undergone a severe “stress test” that exposed deep structural tensions—yet it is not dead. Instead, amid controversy and market volatility, it is exploring a new evolutionary path, including the introduction of a Layer 2 solution.

The “Dying” Narrative: Why So Much Criticism?

Sceptics argue that HyperEVM is either dead or being actively killed, citing several troubling observations:

Dismal ecosystem projects: Prominent KOL katexbt noted that out of 18 HyperEVM projects, 13 were “complete wastes of time”—including 7 already dead, 3 that stalled after their token generation events (TGE), and others with little to show.

Misaligned resource allocation: Hyperliquid places its core trading business (HyperCore) within a high‑performance, closed engine, while the open “application layer” (HyperEVM) is treated as a secondary component. Community members criticise the team for concentrating resources on core products like HIP‑3 and HIP‑4, leaving ecosystem development underfunded.

Poor developer experience: HyperEVM was not designed as a general‑purpose Ethereum environment, but rather as a “specialised layer” for interacting with HyperCore. Its reliance on corewriter and precompiled contracts creates a steep learning curve and complicates cross‑asset operations.

Lackluster application‑layer metrics: In stark contrast to HyperCore’s strength, HyperEVM’s DeFi protocols collectively generated less than $6 million** in fees by early August 2026, while HyperCore’s trading‑related fees amounted to **$56 million over the same period—a nearly ten‑fold difference.

Declining TVL and user activity: Hyperliquid’s total value locked (TVL) dropped from $1.44 billion at the end of Q2 to about $1.2 billion by early August, with only a small fraction genuinely locked in HyperEVM applications—and that fraction was shrinking. Daily active sending addresses hovered around 8,000, far below those of Base (over 250,000) and Arbitrum (over 110,000).

Security incidents: Ecosystem projects suffered breaches, including the lending protocol Purrlend, which lost approximately $1.52 million due to a compromised admin multisig, and a phishing attack on HyperSwap that further eroded user confidence.

The “Evolving” Evidence: Turning Points Are Emerging

Despite the harsh criticism, recent developments suggest that HyperEVM is far from inert—and indeed, a dramatic turning point has arrived:

Pump.fun integration acts as a “stress test”: In late August 2026, the meme‑coin launchpad Pump.fun integrated with HyperEVM. This sent network activity soaring: gas fees spiked from ~0.15 Gwei to 60 Gwei in just two days—a 400× increase. While this caused congestion, it also proved HyperEVM’s ability to handle extreme demand and its latent appeal to high‑volume traders.

Strong data rebound: Driven by the meme frenzy, HyperEVM metrics exploded:

Daily revenue exceeded $500,000 on August 23.

Weekly fees surpassed $1 million that same week.

24‑hour DEX trading volume reached $503 million on August 27.

Daily active DEX addresses hit 25,500, the highest since September 2025.

Launch of “Elysium” Layer 2: In response to HyperEVM’s constraints, the leading liquid staking protocol on Hyperliquid, Kinetiq, announced a dedicated Layer 2 named Elysium.

Performance leap: Elysium promises orders of magnitude higher block speed and throughput than the current HyperEVM from launch.

Better data access: It plans to improve the L1 Read precompile, giving developers richer and fresher market‑depth data from HyperCore than what the current HyperEVM provides.

Strategic division: This move tacitly acknowledges HyperEVM’s limitations and creates a new layer to offload risk—allowing HyperCore to focus on core trading, HyperEVM to evolve cautiously, and Elysium to serve as a more aggressive experimentation ground.

Conclusion: Dying or Evolving?

Looking at the full picture, the controversy around HyperEVM is best understood as the inevitable “growing pain” of a nascent ecosystem.

The “dying” accusations are not without merit—they highlight fundamental problems in strategic positioning, resource allocation, developer experience, and ecosystem building. The core contradiction is that Hyperliquid, as a trading‑first platform, naturally channels its best assets (liquidity, users) toward HyperCore, leaving HyperEVM—the application layer—perpetually undernourished.

Yet the “evolving” evidence is equally compelling: the Pump.fun‑driven surge proved HyperEVM’s potential as a venue for high‑volatility asset trading. And the birth of Elysium L2 offers a pragmatic answer to the impossible triangle—protecting the core, maintaining general composabil