lubluniky

lubluniky / bs-p

Public

Ultra-low latency AVX-512 Polymarket market-making kernel (Logit Jump-Diffusion + Avellaneda-Stoikov in logit space)

avx512 c core hft market-making
54
8
89% credibility
Found Feb 25, 2026 at 18 stars 3x -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
C
AI Summary

A performance-optimized library for computing inventory-aware buy and sell quotes across batches of prediction markets using specialized financial mathematics.

How It Works

1
🔍 Discover the quote calculator

While searching for ways to trade smarter on prediction markets like Polymarket, you find this ultra-fast tool for figuring out the best buy and sell prices.

2
📦 Add it to your setup

You simply include this speedy calculator in your trading program, ready to handle tons of markets at once.

3
📊 Prepare your market details

You gather simple lists about current odds, your holdings, risk levels, time left, and adjustment factors for each market.

4
⚡ Get instant quotes

In a flash, it computes perfect bid and ask prices for thousands of markets, adjusting for your inventory and risks.

5
📈 Feed your trading bot

Your bot now uses these smart prices to place orders quickly and keep your trades balanced.

6
🚀 Handle live updates smoothly

A fast message queue lets your bot receive market changes without missing a beat, keeping everything in sync.

🎉 Master prediction markets

You now run a high-speed market-making operation that feels effortless and puts you ahead of the competition.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 18 to 54 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is bs-p?

bs-p delivers an ultra-low latency Rust crate for Polymarket market making, computing batch bid/ask quotes via Logit Jump-Diffusion and Avellaneda-Stoikov models in logit space. Feed it arrays of current logits, inventory, volatility, risk aversion, time horizons, and spread params; it spits out probability quotes with zero hot-path allocations. Built on AVX-512 SIMD in C with Rust FFI, it's tuned for HFT backtesting or live bots handling thousands of markets.

Why is it gaining traction?

It crushes latency at 6.71 ns per quote on 8K-market batches, plus a lock-free SPSC ring buffer for 29M msgs/sec – pure ultra low latency github firepower akin to ultra low latency mode nvidia or ultra low latency NIC setups. Users love the pre-allocated SoA API for seamless batch processing, fast sigmoid/logit transforms, and numerically stable clamping, outpacing scalar finance libs. The Avellaneda-Stoikov adaptation shines for inventory-aware quoting without the usual overhead.

Who should use this?

HFT traders building Polymarket quoting bots, quants backtesting prediction market strategies, or infra devs piping ultra low latency video streaming feeds into Avellaneda-Stoikov engines. Ideal for those chasing bs-performance autohaus e.k.-level speed in live trading loops.

Verdict

Solid niche pick for ultra-low latency prediction markets despite 14 stars and 0.9% credibility score – benchmarks and Rust API deliver real value now. Test it for bots; maturity will grow with adoption.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.