This week: the waitlist hit 80K and the private beta launches on mainnet next week, Noir reached 1.0, Microsoft Research unveiled a faster post-quantum proving system, and Ethereum researchers explored outcome checks for transactions.
In House
A few things from our side first.
What happened at Raycash this week:
We’re at 80,000 people on the waitlist.
The private beta launches on mainnet next week.
Episode 3 of They’re Watching You is Live
Join the waitlist at raycash.xyz →
On Our Radar
Here is what got us excited this week:
Noir 1.0 gives ZK developers a stable base
Noir 1.0 is out, the first stable release of the open-source language for zero-knowledge applications. Syntax and semantics stay stable across the 1.x series, and breaking changes now require a major version with a deprecation period. The toolchain follows semantic versioning, the compiler went through internal and external audits plus fuzzing, and more than 100 contributors work on the language. Aztec Alpha V6 lands days later and targets Noir 1.0.
A ZK compiler decides what a proof actually proves. It generates the constraints, so a compiler bug can break an application whose code is correct. That makes audits and a stable interface the unglamorous work this stack needed. With the V5 proving bug still fresh, V6 arriving on an audited 1.0 toolchain is the right order of events.
Microsoft Research unveils Magnetar, a fast post-quantum zkVM
Magnetar is a zero-knowledge virtual machine that proves ordinary programs written in Rust, Go or C by proving RISC-V execution. The authors report 7.2 to 8.7 million processor instructions per second on a laptop, 10 to 39 times faster on average than the fastest comparable zkVMs across 15 benchmarks, with proofs that verify in about 12 milliseconds. Security rests only on standard hash functions like SHA-256 and SHA-3, with no elliptic curves, lattices or trusted setup.
That last part makes it quantum-resistant. Proofs from normal code, built on the most heavily studied primitives, is a strong combination. The speed figures are Microsoft’s own benchmarks and Magnetar is a research result for now, so the test is whether others reproduce them.
Ethereum researchers explore transaction assertions
Ethereum Foundation researchers are looking at native transaction assertions, a way to judge a transaction by what it did and not only by what was signed. EIP-7906 would add a read-only check at the end of a transaction that reverts it if the net result breaks a rule, such as a minimum amount received. It builds on EIP-8141 frame transactions, remains a draft considered for the Hegotá upgrade, and isn’t expected before 2027. The Foundation says a rule locking Bybit’s Safe implementation contract would have reverted the transaction behind the $1.5 billion theft.
A signature proves you approved certain bytes. It says nothing about the result, and users sign complex transactions blind every day. Assertions only help if the rule is strong, and the EIP’s own text warns that one that checks too little gives a false sense of security. A private payment does nothing for you if you sign something that drains the account. Outcome checks guard the signing step, and that step has already cost people real money.
Thanks,
From the Raycash team
