Running a GPU burst for Multi-Messenger Astrophysics with IceCube across all available GPUs in the Cloud
1 like6,074 views
- IceCube is a neutrino observatory that detects high-energy neutrinos from astrophysical sources to study violent cosmic events. It uses over 5000 optical sensors buried in Antarctic ice to detect neutrinos. - A cloud burst was performed using over 50,000 GPUs across multiple cloud providers worldwide to simulate photon propagation through ice for IceCube data analysis. This was the largest cloud simulation ever and demonstrated the ability to burst at exascale scales. - The simulation helped improve IceCube's neutrino detection and pointing resolution to identify the first known source of high-energy neutrinos, a blazar, demonstrating IceCube's potential for multi-messenger astrophysics.
Running a GPU burst for Multi-Messenger Astrophysics with IceCube across all available GPUs in the Cloud
- 1. Running a GPU burst for Multi- Messenger Astrophysics with IceCube across all available GPUs in the Cloud Frank Würthwein OSG Executive Director UCSD/SDSC
- 2. Jensen Huang keynote yesterday 2 The Largest Cloud Simulation in History 50k NVIDIA GPUs in the Cloud 350 Petaflops for 2 hours Distributed across US, Europe & Asia On Saturday morning we bought all GPU capacity that was for sale in Amazon Web Services, Microsoft Azure, and Google Cloud Platform worldwide
- 4. IceCube 4 A cubic kilometer of ice at the south pole is instrumented with 5160 optical sensors. Astrophysics: • Discovery of astrophysical neutrinos • First evidence of neutrino point source (TXS) • Cosmic rays with surface detector Particle Physics: • Atmospheric neutrino oscillation • Neutrino cross sections at TeV scale • New physics searches at highest energies Earth Science: • Glaciology • Earth tomography A facility with very diverse science goals Restrict this talk to high energy Astrophysics
- 5. High Energy Astrophysics Science case for IceCube 5 Universe is opaque to light at highest energies and distances. Only gravitational waves and neutrinos can pinpoint most violent events in universe. Fortunately, highest energy neutrinos are of cosmic origin. Effectively “background free” as long as energy is measured correctly.
- 6. High energy neutrinos from outside the solar system 6 First 28 very high energy neutrinos from outside the solar system Red curve is the photon flux spectrum measured with the Fermi satellite. Black points show the corresponding high energy neutrino flux spectrum measured by IceCube. This demonstrates both the opaqueness of the universe to high energy photons, and the ability of IceCube to detect neutrinos above the maximum energy we can see light due to this opaqueness. Science 342 (2013). DOI: 10.1126/science.1242856
- 7. Understanding the Origin 7 We now know high energy events happen in the universe. What are they? p + g D + p + 0 p + g g p + g D + n + + n + + Co Aya Ishihara The hypothesis: The same cosmic events produce neutrinos and photons We detect the electrons or muons from neutrino that interact in the ice. Neutrino interact very weakly => need a very large array of ice instrumented to maximize chances that a cosmic neutrino interacts inside the detector. Need pointing accuracy to point back to origin of neutrino. Telescopes the world over then try to identify the source in the direction IceCube is pointing to for the neutrino. Multi-messenger Astrophysics
- 8. The ν detection challenge 8 Optical Pro Aya Ishiha Combining all the possible info These features are included in We re al a s be de eloping h Nature never tell us a perfec satisfactory agreem Ice properties change with depth and wavelength Observed pointing resolution at high energies is systematics limited. Central value moves for different ice models Improved e and τ reconstruction Þ increased neutrino flux detection Þ more observations Photon propagation through ice runs efficiently on single precision GPU. Detailed simulation campaigns to improve pointing resolution by improving ice model. Improvement in reconstruction with better ice model near the detectors
- 9. First evidence of an origin 9 First location of a source of very high energy neutrinos. Neutrino produced high energy muon near IceCube. Muon produced light as it traverses IceCube volume. Light is detected by array of phototubes of IceCube. IceCube alerted the astronomy community of the observation of a single high energy neutrino on September 22 2017. A blazar designated by astronomers as TXS 0506+056 was subsequently identified as most likely source in the direction IceCube was pointing. Multiple telescopes saw light from TXS at the same time IceCube saw the neutrino. Science 361, 147-151 (2018). DOI:10.1126/science.aat2890
- 10. IceCube’s Future Plans 10 | IceCube Upgrade and Gen2 | Summer Blot | TeVPA 2018 The IceCube-Gen2 Facility Preliminary timeline MeV- to EeV-scale physics Surface array High Energy Array Radio array PINGU IC86 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 … 2032 Today Surface air shower ConstructionR&D Design & Approval IceCube Upgrade IceCube Upgrade Deployment Near term: add more phototubes to deep core to increase granularity of measurements. Longer term: • Extend instrumented volume at smaller granularity. • Extend even smaller granularity deep core volume. • Add surface array. Improve detector for low & high energy neutrinos
- 11. Details on the Cloud Burst
- 12. The Idea • Integrate all GPUs available for sale worldwide into a single HTCondor pool. - use 28 regions across AWS, Azure, and Google Cloud for a burst of a couple hours, or so. • IceCube submits their photon propagation workflow to this HTCondor pool. - we handle the input, the jobs on the GPUs, and the output as a single globally distributed system. 12 Run a GPU burst relevant in scale for future Exascale HPC systems.
- 13. A global HTCondor pool • IceCube, like all OSG user communities, relies on HTCondor for resource orchestration - This demo used the standard tools • Dedicated HW setup - Avoid disruption of OSG production system - Optimize HTCondor setup for the spiky nature of the demo § multiple schedds for IceCube to submit to § collecting resources in each cloud region, then collecting from all regions into global pool 13
- 14. HTCondor Distributed CI 14 Collector Collector Collector Collector Collector Negotiator Scheduler SchedulerScheduler IceCube VM VM VM 10 schedd’s One global resource pool
- 15. Using native Cloud storage • Input data pre-staged into native Cloud storage - Each file in one-to-few Cloud regions § some replication to deal with limited predictability of resources per region - Local to Compute for large regions for maximum throughput - Reading from “close” region for smaller ones to minimize ops • Output staged back to region-local Cloud storage • Deployed simple wrappers around Cloud native file transfer tools - IceCube jobs do not need to customize for different Clouds - They just need to know where input data is available (pretty standard OSG operation mode) 15
- 16. The Testing Ahead of Time 16 ~250,000 single threaded jobs run across 28 cloud regions during 80 minutes. Peak at 90,000 jobs running. up to 60k jobs started in ~10min. Regions across US, EU, and Asia were used in this test. Demonstrated burst capability of our infrastructure on CPUs. Want scale of GPU burst to be limited only by # of GPUs available for sale.
- 17. Science with 51,000 GPUs achieved as peak performance 17 Time in Minutes Each color is a different cloud region in US, EU, or Asia. Total of 28 Regions in use. Peaked at 51,500 GPUs ~350 Petaflops of fp32 8 generations of NVIDIA GPUs used.
- 18. A Heterogenous Resource Pool 18 28 cloud Regions across 4 world regions providing us with 8 GPU generations. No one region or GPU type dominates!
- 19. Science Produced 19 Distributed High-Throughput Computing (dHTC) paradigm implemented via HTCondor provides global resource aggregation. Largest cloud region provided 10.8% of the total dHTC paradigm can aggregate on-prem anywhere HPC at any scale and multiple clouds
- 20. IceCube is ready for Exascale • Humanity has built extraordinary instruments by pooling human and financial resources globally. • The computing for these large collaborations fits perfectly to the cloud or scheduling holes in Exascale HPC systems due to its “ingeniously parallel” nature. => dHTC • The dHTC computing paradigm applies to a wide range of problems across all of open science. - We are happy to repeat this with anybody willing to spend $50-200k in the clouds. 20 Contact us at: help@opensciencegrid.org Or me personally at: fkw@ucsd.edu Demonstrated elastic burst at 51,500 GPUs IceCube is ready for Exascale
- 21. Acknowledgements • This work was partially sponsored by NSF grants OAC-1941481, MPS-1148698, OAC-1841530 and OAC-1826967. 21