DoubleZero

2Z is a permissionless protocol aggregating underutilized private fiber into a high-performance communication layer for distributed systems, solving network-level bottlenecks like spam, jitter, and bandwidth.

Modern distributed systems are bottlenecked not by compute power, but inefficient communication between nodes. Despite significant validator performance upgrades (e.g., Firedancer, Reth), block propagation and consensus remain throttled by spam, latency, and jitter on the public internet. Validators are burdened with filtering spam, verifying signatures, and deduplicating transactions, before producing a block. Public internet traffic doesn’t prioritize time-sensitive data, meaning critical transactions and messages are delayed or lost in noisy channels. Private networking (like co-location) solves this but sacrifices decentralization and permissionlessness. 2Z addresses this issue with a decentralized bandwidth marketplace and an optimized routing architecture, which filters transactions early and guarantees fast, secure message propagation.

2Z is a decentralized network protocol that targets the communication bottleneck at the network layer. It replaces noisy, public internet routes with a permissionless mesh of private fiber. 2 layer architecture: (i) outer ring filters spam and verifies messages then (ii) inner ring handles high-speed consensus over dedicated, low-jitter lines.

Austin Federa, prev Head of Strategy at the Solana Foundation. Senior PM and Bison Trail. Marketing @ Republic 2018. Andrew McConnell: Jump trading 7 years, US + Global Infra, Nasdaq Mateo Ward: Neutrona Networks: CEO & Co-Founder LATAM carrier, Nokia: Director + Nortel Manager.

Potential use cases outside crypto are compelling but underexplored in this version of the whitepaper. The paper doesn’t describe token models or economic guarantees, focuses heavily on architecture. I believe that the incentive design will make or break this protocol. There are many good idea in the space but alignment can make a protocol sustainable for the long run or not.

(i) How are bandwidth contributors paid? What are the tokenomics, if any? Even if open-source, do users trust FPGAs to filter fairly? Could censorship creep in? (ii) How practical is global coordination of private fiber contributions across jurisdictions and infrastructure? (iii) What motivates validators to switch to 2Z infrastructure versus sticking with cloud-based or colocated setups? (iv) Could this eventually trend toward centralized control of critical infrastructure like AWS did, just under a decentralized banner? (v) If 2Z drastically reduces fees and latency, how does that affect Solana’s fee market, MEV patterns, or consensus design? (vi) Could something like 2Z challenge the economic role of L1s more broadly (e.g., will L1s become scheduling layers while 2Z handles execution)?

(i) Traditional internet peering involves bilateral agreements between networks where they exchange bandwidth. This is often managed through legal contracts and physical presence in data centers. (ii) The "trust relationship is different" in peering compared to a simple payment model. (iii) Billing for overages and other specifics are negotiated in "peering agreements." (iv) Blockchain governance is seen as "particularly well suited for" global coordination of private fiber contribution. (v) A key advantage of blockchain-based coordination is reliance on "crypto cryptoeconomic guarantees on the chain" instead of the law as a backstop, offering a "net improvement over the current state of things." (vi) However, traditional peering in data centers involves "literal physical assets in that jurisdiction," making participants presumably "subject to the law of that jurisdiction." This contrasts with potential cross-jurisdictional issues in smart contract interactions.

(i) The discussion raised skepticism about whether "users trust FPGAs to filter fairly." (ii) It is argued that "that should be a again cryptoeconomic element to it to incentivize them to um to to be to be fair." (iii) While auditable FPGAs are possible if "someone puts up the money," there is doubt about the "economic justification for that." (iv) Practically, it's suggested that "people will just trust the FPGAs even though because it's not worth no one's going to want to spend the... time like $2 million to build and roll out verified FPGAs."

(i) 2Z aims to provide performant network infrastructure, initially focusing on Solana validators and infrastructure (med searchers, RPCs, etc.). (ii)Mainnet is expected to be "sufficiently more performant" than the current testnet. (iii) Joining the testnet is currently a "permissioned testnet," requiring filling out a form on the Doublezero.xyz/connect page. (iv) Participation can involve being a validator or providing "bandwidth and network infrastructure." (v) Bandwidth contribution can be at Layer 1 (direct fiber), Layer 2 (using existing network connectivity), or Layer 3 (a tunnel). (vi) Becoming a network contributor requires specific hardware, including Arista 7130 or 72A switches (some with FPGAs), costing potentially "$80,000 worth of equipment." (vii) The required hardware and expertise are described as "not particularly standard" for most network providers, making participation "much more specialized." (viii) The network aims to support "more models and more vendors" in the future, but they will need to be "essentially certified." (ix) The capital expenditure for contributing is acknowledged to be "a higher capital expenditure" and not a "retail platform" where "Anybody should be able to... contribute." This is due to the need for "robust infrastructure." (x) Despite the high barrier to entry, there is a belief in a "robust like economic model around folks to be able to profit from... setting this infrastructure up and running it reliably." (xi) Users of the 2Z network pay a fee, which is intended to have "5% of a variety of inflation and potential rewards." These fees are paid to contributors. (xii) Rewards can be earned in the supported token (e.g., the token for 2Z, or SOL for Solana). (xiii) Mainnet is anticipated for "late Q3... sometime in fall," with a "much broader metro set" (27-30 metros globally) and higher bandwidth (1000 gig depending on contributors). (xiv) Mainnet will have "various models... financial models or economic models" for connecting, with some "validation that... you've done certain activities."

(i) The discussion explores the "upside of participating" for network contributors. (ii) Validators are motivated to switch infrastructure because "Cloud-based costs literally a thousand times as much. Collocated, you're going to get better performance." (iii) For validators already collocated, 2Z is presented as a "net improvement to your collocation set up" that is "dramatically cheaper than EWS." (iv) For network contributors, the incentive is the economic model allowing them to profit from providing the infrastructure. (v) The network is expected to become "multi-talented" in the near future, allowing contributors to earn fees from "many networks," not just Solana. (vi) There are "quite a number of different use cases," including non-Web3 related distributed systems.

(i) A market maker seeks "fast prices" from centralized exchanges (like Binance in Tokyo) to trade on a decentralized chain (like Solana, with infrastructure primarily in Europe). (ii) The 2Z network could potentially help by allowing the market maker to "collocate either at both points at the service provider in Tokyo and in Europe" or have "one machine in Tokyo on the 2Z network" to subscribe to the Solana feed. (iii) However, the network itself doesn't subscribe to exchange price feeds; someone (like an RPC or validator) would need to "provide that data across the network." (iv) Latency can be "asymmetric" on a private link, meaning latency from one point to another might be lower than the reverse.

(i) One participant, with experience in Ethereum and low-latency networks, expresses concern that while 2Z is described as "permissionless," the high capital expenditure, specific hardware requirements, and need for specialized expertise make it "not really" permissionless in the same way as some other blockchain networks. (ii) The cost of participating is described as "several hundred thousands of dollars" in addition to the required expertise. (iii) Despite the high cost, this participant acknowledges that the capital outlay is "not actually that high" when viewed from the perspective of "private networking" rather than typical Ethereum participation.

(i) Clarify the specific economic model and fee structure for network contributors and users on mainnet. (ii) Provide more detailed information on the hardware certification process for new vendors and models. (iii) Discuss the roadmap for supporting additional protocols beyond Solana. (iv) Explore strategies for addressing the perception of permissioning due to the high capital and expertise requirements for network contribution. (v) Investigate how market makers and other low-latency users can effectively leverage the 2Z network to propagate external data sources like exchange price feeds. (vi) Detail the governance model for the global coordination piece facilitated by the blockchain.

2Z and Its economic impacts on Solana