Wispr Flow: Dictation That Works EverywhereStop typing. Start speaking. 4x faster.
Promoted
Z6 PQC-Gateway
Denial-of-service protection for post-quantum endpoints
8 followers
Denial-of-service protection for post-quantum endpoints
8 followers
ML-KEM-768 decapsulation costs 791 µs of server CPU, and any client triggers it with 1,088 bytes. In one month Cloudflare handled 89 separate attacks above 100m. Google mitigated one peaking at 398m. Absorbing that means owning 80k cores that idle the rest of the year. Nobody provisions that at any price, so the endpoint goes down. Or a 24-byte token, 2.2% overhead on the ciphertext it gates, no server state, no session memory, no shared secret.
the 824b sub-mtu result is genuinely impressive for pqc traffic, that latency flattening under concurrent load is exactly the kind of real-world metric we need more of in this space. one thing i'd love to see is a side-by-side benchmark against a hardware tls accelerator baseline, since most enterprise teams will want to know how a software gateway like this compares before replacing their hsm-backed setup. would make the case for adoption much stronger.
Honestly the latency numbers here are kind of wild, 6.8ms versus 38ms is a pretty huge gap if that holds up in real workloads. Cutting ML-KEM payloads under MTU is exactly the kind of unsexy fix that actually matters at scale.
A consolidated dashboard view showing all ten apps on one screen with toggleable status indicators would save a lot of clicking. Right now I imagine users have to jump between separate tools to compare noise floor readings, which seems to defeat the point of having a unified tensor engine underneath.
honestly sounds wild, the tensor engine stuff is way over my head but post quantum security is something i keep waiting to see more accessible. one thing that would help a lot is some kind of simple benchmark or comparison page so people can actually see how it stacks up against classical encryption. would make it way easier to recommend to non technical folks