Quick answer: Project Hamilton was a joint research effort by the Federal Reserve Bank of Boston and MIT’s Digital Currency Initiative to study the technology behind a central bank digital currency (CBDC). It built and open-sourced a high-speed transaction processor, with one design reaching 1.7 million transactions per second. It was deliberately technical research, not a recommendation or a decision to issue a US digital dollar. The project concluded at the end of 2022.
Updated 07/17/2026. By Jake Claver. Educational content, not investment advice.
A central bank digital currency is a frequently misunderstood topic, partly because the research gets read as policy. Project Hamilton is a good case study in the difference. It was a hands-on engineering project to find out whether a CBDC could be built to handle the speed and reliability of a national retail payment system, and to publish what was learned. It was not a plan to roll one out.
The work was a collaboration between the Federal Reserve system’s Boston Bank and MIT’s Digital Currency Initiative. Here is what it actually produced.
What Project Hamilton was
Project Hamilton was a multiyear research project to explore the CBDC design space and understand the technical challenges of a digital form of central bank money. The specific goal of Phase 1 was narrow and concrete: design a core transaction processor that could meet the speed, throughput, and fault-tolerance requirements of a large retail payment system. In February 2022 the team released a Phase 1 paper and, per the Boston Fed’s Phase 1 executive summary, published the code as open-source software.
The two architectures and the speed numbers
The team built and tested two different transaction-processing designs, and the performance gap between them is the technically interesting part:
- Atomizer architecture: reached roughly 170,000 transactions per second, with 99% of transactions settling in under two seconds. It keeps transactions ordered.
- Two-phase commit architecture: reached about 1.7 million transactions per second, with 99% settling in under one second, in a geographically distributed setting. The trade-off is that transactions are not necessarily registered in strict order.
For comparison, those throughput figures are far above typical card-network volumes, which was the point: to see how far a purpose-built processor could go. A notable finding was that the team did not need a conventional blockchain to hit these numbers. The high-throughput design decoupled the validation steps and co-designed the transaction format with parallel state updates, an approach documented in the peer-reviewed Hamilton paper published in 2023.
What it deliberately did not decide
This is the part most worth being precise about. The Boston Fed’s completion announcement framed the work as exploratory research into feasibility, not a signal that the United States would issue a CBDC. That decision is a matter for elected officials and the wider Federal Reserve, and Project Hamilton did not make it. It also did not settle questions of privacy, intermediary roles, or legal design; it focused on the transaction-processing engine. Reading it as a launch plan is the most common mistake made about it.
OpenCBDC and why the code is public
The software, released as OpenCBDC-tx under an open-source license, is described by MIT DCI as a centralized transaction processor for a hypothetical CBDC that implements the two architectures above. Publishing it lets other researchers, central banks, and engineers examine and build on the same code rather than starting from marketing claims. MIT notes the collaboration concluded in 2023 and the project is no longer actively maintained, but the research outputs remain a reference point in the digital-currency conversation.
How it fits the bigger tokenization picture
Project Hamilton studied a retail CBDC processor. A related but separate strand of official research looks at wholesale settlement and tokenized deposits. The Bank for International Settlements has laid out a widely-cited version of that in its work on a unified ledger, a shared programmable platform combining central bank money, tokenized commercial-bank deposits, and tokenized assets. The BIS argues that tokenization works only when it rests on the trust of central bank money, and it draws a sharp line between tokenized deposits (which keep the two-tier banking system) and stablecoins (tradable bearer instruments that can drift from par). Hamilton and the BIS blueprint are answering different questions, but they sit in the same official conversation about how digital money should settle.
Why this matters
Two practical takeaways. First, the engineering question, can central bank money move at internet scale with strong fault tolerance, has a documented, testable answer, and the code is public. Second, the technology being feasible is not the same as it being adopted; the policy choice is separate and was left open. For anyone trying to reason about digital money, keeping those two apart is most of the work. And as with any technology topic, none of this is a statement about the price or investment merits of any private token.
Common questions
What was Project Hamilton?
Project Hamilton was a joint research project by the Federal Reserve Bank of Boston and MIT’s Digital Currency Initiative to study the technology behind a central bank digital currency. Its Phase 1 goal was to design a core transaction processor fast and reliable enough for a large retail payment system. It concluded at the end of 2022.
Does Project Hamilton mean the US is launching a digital dollar?
No. The Boston Fed described it as exploratory technical research into feasibility, not a decision or recommendation to issue a US CBDC. That policy choice rests with elected officials and the wider Federal Reserve and was left open by the project.
How fast was the Project Hamilton system?
It tested two designs. An atomizer architecture reached about 170,000 transactions per second with 99% settling under two seconds, and a two-phase commit architecture reached about 1.7 million transactions per second with 99% settling under one second, though transactions were not necessarily kept in order.
Did Project Hamilton use a blockchain?
Not in the conventional sense. The team found it did not need a traditional blockchain to reach its highest throughput. The fastest design decoupled validation steps and co-designed the transaction format for parallel processing. The code was released as open source under the name OpenCBDC-tx.
How does it relate to tokenized deposits and the BIS unified ledger?
They are separate but related strands of official research. Project Hamilton studied a retail CBDC transaction processor, while the Bank for International Settlements’ unified ledger concept looks at combining central bank money, tokenized deposits, and tokenized assets on one platform for wholesale settlement. Both examine how digital money should settle, from different angles.
This content is educational only. It is not tax, legal, or investment advice. Check primary sources and speak with a qualified professional before making financial decisions.
