Demos of the Adviser tech simplifying complex simulations and jobs. For complete documentation, see https://github.com/adviserlabs/docs.
Small molecule protein binding simulation. Creates proteins and ligands and then creates some mock data for binding affinity then generates results in the terminal itself.
Simulates heat transfer and fluid flow in a heated cavity with a moving lid on parallel CPU/GPU clusters; outputs temperature contours and velocity profiles to optimize cooling systems in electronics, engines, or solar panels.
Creates a bunch of synthetic tickers, generates trade data for the over a long period, then does analysis on the biggest winners/losers.
Runs historical market data through trading strategies on a cluster; outputs performance metrics (e.g., Sharpe ratio, drawdowns) to reveal profitable vs. risky models.
Models atmospheric/ocean dynamics over decades on HPC clusters; outputs temperature/rainfall maps to predict regional climate change impacts.
Evolves orbits of millions of particles (e.g., galaxies) on parallel clusters; outputs collision/formation visuals to study cosmic structure evolution.
Computes electron structures of materials on CPU/GPU clusters; outputs band gaps/energy levels to design better semiconductors or catalysts.
Simulates airflow around vehicles/wings on mesh-based clusters; outputs drag/lift coefficients to optimize aircraft fuel efficiency.
Analyzes millions of genetic variants across samples on distributed clusters; outputs SNP significance maps to pinpoint disease-linked genes.
Trains deep models (e.g., transformers) on massive datasets via parallel GPUs; outputs accuracy/loss curves to deploy production-ready AI models.
Docks millions of compounds against targets on cloud clusters; outputs hit lists (affinity scores) to prioritize lab-tested drug leads.
Models seismic waves through Earth layers on finite-element clusters; outputs ground motion maps to forecast building damage in quake zones.
Simulates stress/strain on bridges/buildings under loads on parallel solvers; outputs deformation/failure points to ensure structural safety.
Simulates particle collisions (e.g., LHC data) on massive clusters; outputs decay/event distributions to validate theories like Higgs boson.
Generates millions of samples from complex probability distributions on parallel clusters; outputs posterior parameter estimates and uncertainty bands to quantify beliefs about models (e.g., disease spread rates or climate sensitivity).
Modeling Infection rate (R0) for "Optimal quarantine duration"
Modweling CO2 sensitivity for "Price carbon tax accurately"
Modeling Volatility correlations for "Hedge fund risk budget"
Modeling Mutation rates for "virus evolution"
Solves supply chain/inventory problems via parallel MIP solvers; outputs optimal routes/costs to minimize logistics expenses for enterprises.
Folds amino acid sequences using distributed AlphaFold-like models; outputs 3D PDB files to reveal functional protein shapes for drug design.
Tracks chemical plumes in air/water on advection-diffusion clusters; outputs concentration maps to plan urban emission controls and health risks.