Anton, a Special-Purpose Machine for Molecular Dynamics Simulation

1. Anton, a Special-Purpose Machine for Molecular Dynamics Simulation By David E. Shaw et al Presented by Bob Koutsoyannis

2. The Anton Legacy • Anton van Leeuwenhoek “Father of Microscopy” • First to see bacteria and other micro organisms • Objective: Improve the tools available to scientists to further our understanding of organisms & diseases

3. Anton the Machine • Specialized Massively Parallel Machine being built to improve Molecular Dynamic Simulations. • In the works to be completed by 2009 • Biological processes spatially distributed among many nodes in a 3D torus. • MD specific hardware • Novel parallel algorithms

4. Molecular Dynamics Simulation • Models the motions and interactions of molecular systems • Proteins • Cell Membranes • DNA • (atomic level simulations)

5. Motivation • Life Saving… • Used to visualize biochemical phenomena that cannot be seen in lab experiments. • Protein Folding • Protein, Protein interactions • Protein, Drug interaction • Key for Developing Drugs

6. What makes one MD simulatorbetter than the Next? • Time Scale • Being able to simulate the interaction between molecules for more than a nanosecond. • Problem Size • Why is a millisecond of simulation out of the scope of our current technology? • Consider 200,000 molecules • 1012 time steps to simulate a millisecond • Each time step requires intense arithmetic computation on all 200,000 molecules

7. What makes one MD simulatorbetter than the Next? • Other Projects Addressing MD Sims • Folding@Home • Network of 200,000 PC’s • Large sample for independent molecular sims • But no millisecond simulations • FASTRUN, MDGRAPE, MD Engine • Good with larger molecular system sims • Have strong arithmetic units • Still limited by communication bottlenecks

8. MD Simulator Requirements • Force Calculation • (getting an idea of the level of computation needed) • Molecular mechanics force fields used to model the total PE of a system. • Input: X,Y,ZOutputs: Force Quantities M1 M2

9. MD Simulator Requirements • Force Calculation • (getting an idea of the level of computation needed) • For every time step, the force fields must be updated. • FFT, Convolution, Inverse FFT (Computationally expensive operations) • For 200,000 molecules/step… • 1) Need a huge number of arithmetic processing elements

10. MD Simulator Requirements • Integration • (getting an idea of the level of computation needed) • For every time step, updates of atomic positions and velocities must be made. • Global actions and Constraints must be enforced on the entire system (temperature, pressure, optimizations.)

11. MD Simulator Requirements • Parallelization • (getting an idea of the level of computation needed) • For every time step, every atom must communicate within its cutt-off radius with every other atom. • 2) A lot of inter-processor communication that can be scaled well is needed.

12. MD Simulator Requirements • Parallelization • (getting an idea of the level of computation needed) • Whole System is broken down into boxes (processing nodes) • Each node handles the bonded interactions within • NT method for non-bonded interactions (much more common). • NT method for Atom Migration

13. Why Specialized Hardware? • 1) Need a huge number of arithmetic processing elements • 2) A lot of inter-processor communication that can be scaled well is needed. • 3) Memory is not an issue • With 25,000 atoms (64bytes each) total=1.6MB over 512 nodes=3.2KB/node which is < most L1 Memory Communication Computation Needs

14. Memory Communication Computation Needs Why Specialized Hardware? • Consider Moore’s Law on 10X improvement in 5 years vs. Anton’s 1000X in 1 year. • Can great discoveries wait? • Can use custom pipelines with more precision, increased datapath logic speed, over less silicon area. • Have Tailored ISA’s for geometric calculations+ • Programmability for accommodating various force fields and integration algorithms • Dedicated memory for each particle to accumulate forces

15. Updating force field This node may update for them Communication Latency • Low-latency, high-bandwidthnetwork within and betweenASICs. • Push based communicationwith counters (reduce wait). • Set of Autonomous DirectMemory Access (DMA) Enginesallowing for greater overlap of communication and computation. • Admission Control Features

16. Subsystems of Anton • High-Throughput Interaction Subsystem (HTIS) • Flexible Subsystem • Communication Subsystem • Memory Subsystem

17. High-Throughput Interaction Subsystem • Executes Non-bonded MD interaction calculations (Charge Spreading & Force Interpolation) • Accumulates forces on each particle as data streams through. • ICB Controls flow of data through the HTIS, programmable ISA extensions, acts as a buffering, pre-fetching, synchronization, and write back controller

18. Flexible Subsystem • Initiates Force Computation Phase • Calculates bonded force terms • Force correction terms • All integration tasks Constraint Calculations (temp & pressure) Pos. Vel. Updates Atom Migration All Maintenance Activities (boot, diagnostic, self-test, loading sims, switching contexts, logging, check pointing, error reporting).

19. Flexible Subsystem • General Purpose Core w/ Caches • Remote Access Unit • Autonomous data transfers • Geometry Cores • MD calculations bonded • Correction Pipeline • Computes force correction terms • Racetrack • Local, internal connect for flex subsys components • Ring Interface Unit • Flex subsys to transfer packets to/from communication subsystem.

20. Communications Subsystem • Routing 48-bit address space • 16-bit node identifier 32-bit of address per node • Flow Control • Provided access to ASIC DRAM • Supports accumulation and synchronization Memory Subsystem

21. Simulation Evaluations • 500X NAMD 80-100X Desmond 100X Blue Matter

22. Accuracy Efficiency • Increase system simulation size leads to increase in efficiency. • Force Error measured in relative rms force error • Energy Drift