As widely known, Solana has long harbored a dream of building a PerpDEX.

When founder Toly conceived the Solana architecture back in 2017, his goal was to create a blockchain that was "fast enough and cheap enough to run a viable order book."

Nine years later, Solana’s most successful PerpDEX to date, Drift, was hacked for $285 million, never recovering since.

Drift’s abrupt demise did not shake Solana’s conviction. In the foundation’s narrative, Solana’s ultimate vision is ICM—“on-chain capital markets”—a permissionless Nasdaq where stocks, commodities, and derivatives are all brought on-chain.

Previously, Solana was essentially a blockchain incapable of sustaining a high-performance PerpDEX, making it ill-fitted for this grand narrative. But last month, Solana’s Alpenglow consensus upgrade entered community validator testing. This new consensus mechanism compresses transaction confirmation times from roughly 12.8 seconds down to around 150 milliseconds—approaching the matching speeds of traditional stock exchanges. Technical barriers once standing in the way are now being systematically removed.

The foundation is more determined than ever to secure a “flagship” PerpDEX.

This PerpDEX is Phoenix, developed by Ellipsis Labs.

On Solana, Ellipsis Labs is a name with weight. Led by Jarry Xiao and Eugene Chen, the team launched their proprietary AMM, SolFi, at the end of 2024—a trading venue without a frontend, rejecting retail liquidity provision, and relying solely on market makers’ own capital for quotes. After SolFi’s launch, trading volume surged rapidly, peaking at 60% of Solana’s total AMM volume.

Phoenix is the product of applying the same expertise to perpetual contracts. According to Ellipsis Labs, market makers can update quotes at lower cost, enabling tighter spreads and deeper order books, while traders avoid gas fees entirely.

A fully on-chain, order-book perpetual contract platform running directly on Solana’s mainnet without any off-chain components had never been achieved before Phoenix.

The foundation clearly agrees. Over the past two weeks, Toly and multiple foundation members have repeatedly shared and commented on Phoenix, with promotional intensity visibly evident.

However, the official endorsement sparked backlash within the community.

Other PerpDEXs built on Solana—Pacifica, Bulk, Jupiter—are doing the same thing but have never received official mention.

Pacifica’s founder Constance expressed restraint: “Our team chose Solana in 2025 without receiving any funding from the foundation or raising capital from investors. We simply wanted to build a solid product first and let the market decide.” She added that people often ask whether Pacifica’s growth stems from foundation support—her answer: no. In her view, such tribalism and forcing developers to cater to specific demands will burn down existing bridges before others even get built.

Bulk’s co-founder kdotcrypto expressed regret over the foundation’s bias: “They promote what they believe benefits them most. Pushing aside others just because one team fits a certain criterion turns friends into enemies.”

The foundation’s logic, however, is internally consistent.

Toly’s core argument is that he favors protocols fully operating on-chain and interoperable with other smart contracts, as only these generate revenue for Solana validators. PerpDEXs like Pacifica and Bulk that perform order matching off-chain, no matter how much volume they generate, bring zero direct revenue to Solana itself.

Solana developer Anza’s Max Resnick put it bluntly: “Pacifica is to Solana what Hyperliquid is to Arbitrum. Even if Pacifica truly surpasses Hyperliquid, it brings no tangible benefit to Solana.”

Solana’s shareholder Multicoin partner also endorsed this “brand alignment” behavior: “A chain can remain neutral, but the foundation need not be. Concentrating limited resources on the team you believe in most is perfectly legitimate.”

The logic behind “Solana Help” is roughly this: a public blockchain functions like a shopping mall sustained by rental income and transaction fee cuts. Fully on-chain protocols are tenants renting storefronts and paying cut based on turnover. Off-chain matching PerpDEXs are more like companies that only list the mall’s address on their business cards while conducting real operations elsewhere. Customers come in under the mall’s banner, but no money stays in the mall.

kdotcrypto and Pacifica reject this analogy.

Bulk counters that its storefront is indeed located inside the mall. Bulk is operated by Solana validators, sharing the same validator set as the mainnet, with 12.5% of exchange revenue distributed to Solana validators. kdotcrypto calculated that this revenue share delivers far greater value than ten thousand on-chain transactions.

Constance of Pacifica points out that users must cross-chain funds into Solana to use Pacifica—this is already a tangible network effect—and Pacifica has even made SOL the first and only non-stablecoin collateral asset on its platform.

Toly acknowledges overlap between Bulk and validators, but insists this is merely “the same group of people doing extra work,” not “the same security model.”

On the mainnet, SOL stakers can vote to enable a feature switch that forcibly alters fee rules. But on Bulk, they cannot—because Bulk’s revenue distribution is controlled by its operators, and staked SOL holds no sway. While both systems share the same operator pool, cost efficiency does not equal structural integration. Toly compares this relationship to the overlap between Wormhole’s guardians and Solana validators: superficially the same group, but functionally autonomous.

Regarding validator revenue, Toly argues that the mainnet itself can generate income through the destruction of base fees from high-frequency transactions—even tiny burns per transaction accumulate to a staggering $1–2 billion annually at 100,000 TPS.

Solana has every right to make its own choices.

It is the foundation’s prerogative to allocate its limited resources toward the architecture it deems correct. Blockchain neutrality does not require the foundation to treat every team equally; investing in its most promising direction is neither wrong nor illegitimate. Toly and Tushar are not mistaken on this point.

Yet the arrogance displayed by the Solana team is another matter entirely.

Toly’s tweet referred to other Solana-based PerpDEX teams as enemies destined to be “drowned in infinite forks.” Foundation members repeatedly shared an unverified tweet claiming Phoenix is “six times cheaper” than Hyperliquid.

This “six times” figure holds true only on a narrowly selected trading pair—the gold RWA contract on Hyperliquid that has disabled its growth mode and thus incurs full fees. On Hyperliquid’s standard crypto pairs, Phoenix offers only about 1.23x and 2.88x savings on posting and taker fees respectively. When accounting for the common practice of self-referral discounts among traders, the gap narrows further.

Compared to most HIP-3 contracts using growth modes, Hyperliquid is actually cheaper.

WhitWhale, a large-cap trader, delivered a scathing verdict on Phoenix after real-world testing: mandatory desktop-only access, visible liquidity depth at only about one-fifteenth of Hyperliquid’s, no advanced order types, and requiring an invite code to join.

Phoenix may not be bad—but it’s simply too early.

A product unprepared to handle traffic is thrust into the spotlight with maximum fanfare, and the damage is not to Hyperliquid—it’s to Phoenix’s own reputation.

PUMP, not a native Solana team, still dominates on Solana. Pacifica, having received no foundation funding, remains the highest-volume PerpDEX on Solana. Real user data, voted with feet, cannot be bought with the foundation’s marketing budget.

Traders do not blindly accept the foundation’s endorsements. The winner is always the product with the best experience.

Original: BlockBeats