# Private Reads Update Jan 2026 ## Key Highlights - [Team space](https://www.notion.so/efdn/Private-Reads-Privacy-Acceleration-2f5d9895554181dda471fd9d7d3f7464) moved to EF Notion - PIR: finished [spec and reference implementation](https://github.com/keewoolee/rms24) of Plinko & RMS PIR scheme and with [optimizations](https://github.com/keewoolee/rms24/commit/35926deb8eeea0143af1a23aecc5513c449c2376). A production-grade implementation anchored on this spec is [WIP](https://github.com/igor53627/rms24-rs/tree/main). - [Released PAPs](https://www.notion.so/efdn/PAPs-0cbd98955541825296e201936c5361f2): proposals to boost our work or to accelerate other privacy-related initiatives through iterative micro-grants - Tor: collab with Tor Project to unlock [Arti](https://arti.torproject.org) embedded in the browser kicked off [with a PR](https://gitlab.torproject.org/tpo/core/arti/-/merge_requests/3600) and is [progressing](https://github.com/voltrevo/arti/tree/wasm-arti-client) very [well](https://www.notion.so/efdn/EF-Tor-Weekly-04-02-2fdd98955541801da1d4f9e459ac0ff3) - [Aligned](https://www.notion.so/efdn/Team-weekly-13-01-44cd98955541829da383814278bc1812?source=copy_link#db2d989555418237b07281151a88d8d4) on Q1 [roadmap](https://www.notion.so/efdn/Roadmaps-ca8d989555418381ae430193d009cec0): follow spec-driven approach in [PIR](https://www.notion.so/efdn/PIR-Implementations-2c8d98955541820e9bb5816e57077ce0) decoupled from [UBT](https://www.notion.so/efdn/UBT-Node-ae1d9895554182aa90e30181e242ac35) work, get [Arti](https://www.notion.so/efdn/Arti-in-Browser-950d989555418284b12e01b8f04c4f52) fully functional in-browser, push towards more [acceleration](https://www.notion.so/efdn/PIR-Acceleration-4b3d989555418206a0de017a5cfe8592) and better [communication](https://www.notion.so/efdn/PrivaETHy-Pulse-895d989555418369ae1f01c41a6772b6) - The [Sync'ing](https://github.com/ethereum/devops/issues/2183) of UBT-enabled Geth fails but [this fix](https://github.com/ethereum/go-ethereum/pull/33739#issue-3887073466) from Stateless should alleviate that - [PR to integrate Tor-js into ethersjs](https://github.com/ethers-io/ethers.js/pull/5083) so wallets integrating Ethersjs can route RPC calls through Tor with minimal effort. Tor-js was intended as a temporary unblocker for Tor in browser [until Arti is ready](https://www.notion.so/efdn/Arti-in-Browser-950d989555418284b12e01b8f04c4f52). - Initial sync with EcoDev Digital Studio on making Spotlight series, short videos on privacy work worth highlighting ## Top 1–3 Priorities for Next Month (Feb) - Publish a system design for PIR decoupling clients from sharded server - Solve performance issues in Arti client in-browser - Solve edge issues of sync'ing [UBT](https://eips.ethereum.org/eips/eip-7864)-Geth and prove its execution with OpenVM ## Is there something another EF team can help you with? - Feedback on our sharded PIR system design - Feedback or tips on our approach to [proving UBT-Geth](https://www.notion.so/efdn/UBT-Proof-cf0d98955541820ba5570115224b65c9)'s equivelance to mainnet MPT ## Updates and Progress ### PIR 1. Finished spec and reference implementation for RMS and Plinko scheme 2. Started production-grade implementation of RMS based on the reference implementation 3. Engaged a grantee who will be specc’ing an additional PIR scheme 4. Created initial sketch of a sharded PIR design tailored to different Ethereum dataset which collectively serve the entire Ethereum state/logs/history. 5. GPU-acceleration of hint generation in Plinko tested, <$50 cost to generate hints for 100GB dataset, indicating the viability of the option of delegated hint generation to the server inside a TEE 6. Held initial syncs with 2 teams specialized in FHE/GPU/PIR who are interested in contributing to PIR efforts - **Next**: - Grantee to spec and reference implement an additional scheme next week - Implementations of RMS/Plinko based on Python reference implementation and tests - Share out PIR system design on ethresear[.]ch - **Areas of collaboration:** - GPU acceleration of server-side PIR - Spec and reference implementation of FHE-native PIR schemes ### Verifiable UBT - Spun a UBT-enabled Geth instance and started sync'ing - The MPT-to-UBT conversion flow is connected with the witness pipeline in Geth, and mainnet sync is currently in progress. - Still solving a sync halt issue at the ~3M block which is likely due to stumbling on a large block, leading to a large witness, leading to conversion process falling behind (witness cleared faster the the conversion tool consuming it) - **Next**: - Fix the edge cases of sync'ing - Run Keeper alongside MPT-UBT conversion - Prove Keeper with OpenVM - **Collaboration opportunities:** - Proving [stateless] Geth in other zkVMs besides OpenVM - Implementation of UBT (EIP7864) in other clients amenable to proving in zkVM ### Tor Arti - Aligned with the Tor Project team on the goal of our collab: push PRs from our side, review and merge from their side, culminating end of March in a fully embeddable Tor client (Arti, in wallets, frontends, ..) - The work to get Tor Arti client embeddable in browser context kicked off [with a PR](https://gitlab.torproject.org/tpo/core/arti/-/merge_requests/3600) and is [progressing](https://github.com/voltrevo/arti/tree/wasm-arti-client) very [well](https://www.notion.so/efdn/EF-Tor-Weekly-04-02-2fdd98955541801da1d4f9e459ac0ff3) - Internal [security review](https://github.com/voltrevo/tor-js/pull/5) of tor-js - [Tor-js into ethersjs](https://github.com/ethers-io/ethers.js/pull/5083), now any wallet relying on Ethersjs can route RPC calls through Tor with minimal effort - **Next**: - Continue the weekly checkups with Tor Project - Solve performance issues in wasm Arti ### Ecosystem support and collabs - Initial [sync with ZAN team](https://docs.google.com/document/d/1HTlzNfEqnVdO9P1adUeRy0Cd00IZ-s1JI-SqlgmTOkw/edit?tab=t.cz879ymu3l80) on GPU/FHE/PIR - SNU researchers (Korea) [preseted their work](https://docs.google.com/document/d/1Y_HiNRbOei6e-Ww0pLUZhPN4E6JGy83VRIwW5KAptvA/edit?usp=sharing) and will be coordinating with them on FHE ASIC/GPU acceleration after we publish our ethresear . ch post to see if there are schemes they can take on.