Research on how reductions in latency create value in electronic markets—from adverse selection and inventory risk to queue priority and information constraints.
Latest Research →
MEV in Binance Builder: Two-Swap Dominance
Fresh arXiv preprint showing arbitrage profit concentrates in simple 2-3 swap paths, not complex cycles—latency kills complexity.
Arbitrum Timeboost: Express Lane Auctions
Arbitrum's new mechanism for protocol-level MEV capture via priority auctions. Changes the game from pure speed competition to speed + auction strategy.
Innovation Coding: Transmitting Less When You're Faster
Key insight: lower latency means less innovation accumulates before transmission, so you need to send less data. Foundational for understanding latency-reliability tradeoffs.
NVIDIA's Hybrid Expert Parallel for MoE Training
Deep dive on using NVLink (1-2μs) for intra-node + RDMA (5-10μs) for inter-node expert routing. The 5-10μs RDMA latency is the bottleneck.
Monitoring 500+ GPUs: The Pulse of Large Clusters
Production experience running 504 B200 GPUs shows monitoring requires correlating signals across GPU/network/storage/power subsystems.
Analysis →
In-depth analysis coming soon as research develops.
About
TernQED is a research collective investigating latency value across electronic markets, high-frequency trading, MEV extraction, and distributed AI training. We combine perspectives from market microstructure, information theory, trading practice, and systems architecture.