Computing Architecture

Slides:



Advertisements
Similar presentations
National Radio Astronomy Observatory June 13/14, 2005 EVLA Phase II Proposal Review EVLA Phase II Computing Development Bryan Butler (EVLA System Engineer.
Advertisements

Enhancing Paper Mill Performance through Advanced Diagnostics Services Dr. BS Babji Process Automation Division ABB India Ltd, Bangalore IPPTA Zonal Seminar.
ALMA Real Time Control System Jeff Kern Ralph Marson, Thomas Juerges.
Hunt for Molecules, Paris, 2005-Sep-20 Software Development for ALMA Robert LUCAS IRAM Grenoble France.
EVLA Computing Overview Gareth Hunt EVLA Advisory Committee 2002 June
MWA Data Capture and Archiving Dave Pallot MWA Conference Melbourne Australia 7 th December 2011.
ALMA Software B.E. Glendenning (NRAO). 2 ALMA “High Frequency VLA” in Chile Presently a European/North American Project –Japan is almost certainly joining.
Paul Alexander & Jaap BregmanProcessing challenge SKADS Wide-field workshop SKA Data Flow and Processing – a key SKA design driver Paul Alexander and Jaap.
Doug Tody E2E Perspective EVLA Advisory Committee Meeting December 14-15, 2004 EVLA Software E2E Perspective.
Overall Data Processing Architecture Review EVLA Monitor and Control Interfaces July , 2002EVLA Data Processing PDR Bill Sahr.
CRISP & SKA WP19 Status. Overview Staffing SKA Preconstruction phase Tiered Data Delivery Infrastructure Prototype deployment.
EVLA Transition to Science Operations: An Overview EVLA Advisory Committee Meeting, March 19-20, 2009 Bob Dickman AD, NM Operations.
An FX software correlator for VLBI Adam Deller Swinburne University Australia Telescope National Facility (ATNF)
2007Sep06 EAC Butler - Software Overview 1 Software Overview Bryan Butler.
Observing Modes from a Software viewpoint Robert Lucas and Philippe Salomé (SSR)
1 KFPA Critical Design Review – Fri., Jan. 30, 2009 KFPA Data Pipeline Bob Garwood- NRAO-CV.
SAGE meeting Socorro, May 22-23, 2007 EVLA Science Operations: the Array Science Center Claire Chandler NRAO/Socorro.
Atacama Large Millimeter/submillimeter Array Expanded Very Large Array Robert C. Byrd Green Bank Telescope Very Long Baseline Array Data Processing Progress.
Large Area Surveys - I Large area surveys can answer fundamental questions about the distribution of gas in galaxy clusters, how gas cycles in and out.
14 June, 2004 EVLA Overall Design Subsystems II Tom Morgan 1 EVLA Overall Software Design Final Internal Review Subsystems II by Tom Morgan.
Introduction to EVLA Software Bryan Butler. 2006Dec05/06EVLA M&C Transition Software CDR2 EVLA Computing (Terse) History The original EVLA Phase I proposal.
Recent progress in EVLA-specific algorithms EVLA Advisory Committee Meeting, March 19-20, 2009 S. Bhatnagar and U. Rau.
A real-time software backend for the GMRT : towards hybrid backends CASPER meeting Capetown 30th September 2009 Collaborators : Jayanta Roy (NCRA) Yashwant.
ICALEPCS 2005 Geneva, Oct. 12 The ALMA Telescope Control SystemA. Farris The ALMA Telescope Control System Allen Farris Ralph Marson Jeff Kern National.
N. RadziwillEVLA Advisory Committee Meeting May 8-9, 2006 NRAO End to End (e2e) Operations Division Nicole M. Radziwill.
M.P. RupenEVLA Advisory Committee Meeting September 6-7, Correlator Test Plan Michael P. Rupen.
Software Requirements for the Testing of Prototype Correlator Sonja Vrcic Socorro, December 11, 2007.
Metadata for the SKA - Niruj Mohan Ramanujam, NCRA.
Jeff Kern NRAO/ALMA.  Scaling and Complexity ◦ SKA is not just a bigger version of existing systems  Higher Expectations  End to End Systems  Archive.
The Science Data Processor and Regional Centre Overview Paul Alexander UK Science Director the SKA Organisation Leader the Science Data Processor Consortium.
August 3, 2016 US RMS Futures II, ngVLA Overview ngVLA Overview Mark McKinnon US RMS Futures II August 3, 2016.
Andreas Horneffer for the LOFAR-CR Team
Multi-beaming & Wide Field Surveys
Software Overview Sonja Vrcic
WP18, High-speed data recording Krzysztof Wrona, European XFEL
Bryan Butler EVLA Computing Division Head
NRAO VLA Archive Survey
LSST Commissioning Overview and Data Plan Charles (Chuck) Claver Beth Willman LSST System Scientist LSST Deputy Director SAC Meeting.
Applying Control Theory to Stream Processing Systems
EVLA Availability - or - When Can I Use It?
Simulation Requirements
EVLA Overall Software Design
Introduction to cosynthesis Rabi Mahapatra CSCE617
EVLA Computing Software Overview.
EVLA Prototype Correlator (PTC)
Oxford Algorithms Workshop
Antenna Monitor & Control Subsystem Engineering Requirements
Observational Astronomy
Observational Astronomy
Shared Risk Observing Claire Chandler EVLA SSS Review, June 5, 2009
VLA to EVLA Transition Plan
Degree-aware Hybrid Graph Traversal on FPGA-HMC Platform
Bryan Butler (for Bill Sahr)
EVLA Advisory Committee Meeting
Science Commissioning
Early indications of performance
Gustaaf van Moorsel September 9, 2003
Shared Risk Science with the EVLA
Mark McKinnon EVLA Project Manager
EVLA Advisory Committee Meeting, March 19-20, 2009
EVLA Algorithm Research & Development
EVLA Operations Jim Ulvestad
Prototype Correlator Tests on the Critical Path
Correlator Growth Path
Overview of Workflows: Why Use Them?
SDP Interface Identification
EVLA Construction Status
Rick Perley NRAO - Socorro
Molecular Imager: Focal Plane Array
L. Glimcher, R. Jin, G. Agrawal Presented by: Leo Glimcher
Presentation transcript:

Computing Architecture Jeff Kern

ngVLA Definition (from a software perspective) ngVLA (noun): A geographically distributed large N array for PI driven science with high sensitivity, wide instantaneous bandwidth, and low operational cost.

ngVLA Computing Architecture Constraints Geographically distributed PI Driven Science Antennas must be self protecting Need full suite of operations tools (Proposal, Observing Preparation, Data Processing) Local control for maintenance Minimize tight central communication loops Produce science ready products Large N Array (200+) High Sensitivity Avoid central bottlenecks (logging, monitoring) High calibration and imaging accuracy Snapshot observations likely Robust to elements entering and leaving the array Wide Instantaneous Bandwidth RFI concerns Large (but not unsupportable data volumes) Data Volume (again) Low Operational Cost Automation everywhere possible

Control and Monitoring In principle M&C of the ngVLA is very similar to existing arrays. Devices have gotten smarter, interface to software at a higher level. Use time tags for time critical operations and do synchronization in HW. Use standard communication protocols Large number of antennas and larger geographic area implies: Subarrays will be used nearly continuously (esp. during commissioning) Antennas need to be very independent, local monitoring and error detection Send sky position rather than encoder commands Operations Optimizations Dynamic Scheduling algorithm Need a more sophisticated weather model than typically used (variable across the array). Hardware drivers should self diagnose when updates to parameter data needs to be made, or operation conditions are out of normal. May need to include more self check modes in the hardware design. No RealTime OS

Correlator Software Correlator details TBD but assume: FX Architecture Assume delay tracking performed digitally in Correlator HW Subarrays will be needed early and often. Correlator Back End, cluster for processing and formatting More conditioning of visibilities in this stage than for VLA (Tsys, Sub-band stitching, Telcal Application) RFI Mitigation (or even further upstream) Data Flagging and blanking Data into final archiving / processing format Correlator functionality is easy to add in hardware. Cost is in commissioning and support (firmware and software complexity) ngVLA should limit correlator modes to those that have clear use cases. Commissioned

Data Archiving and Storage Estimated archived data rates are few TB per hour (visibility only) Not trivial, but possible (esp. with continued improvements in storage cost). Baseline design is to store the “raw” visibilities. But current paradigm doesn’t work Moving data, and multiple copies will likely be prohibitively expensive “Filling” data from an archival to working format, expensive and unnecessary Beginning to work on Measurement Set V3 with SKA Designed for massively parallel processing Robust to node failures, thread safe Support backwards compatibility to MS V2 Compute cluster should be “close” to storage Primary processing on PI hardware unlikely

Data Processing Algorithmic and performance requirements are still uncertain. Needs more study now that the system design is stabilizing. Preliminary analysis shows that most use cases are tractable Assumes that compute cost follows historical trend Wide-field low-frequency (<4 GHz) imaging is likely to be computationally prohibitive in early operations. Excluded from the baseline for this reason. Delivered products for most projects will be science quality images Science Ready Data Products (SRDP) project is a precursor to the ngVLA pipeline ALMA Calibration and Imaging Pipeline VLA Calibration Pipeline VLA Sky Survey Pipeline Key Differences from SKA: Higher frequency and later start date

Beginning End to End Operations ngVLA must be a telescope for all astronomers, not just radio “blackbelts”. Proposal process should focus on desired products, not on hardware functionality. Reprocessing will be expensive. Initial generation of the proper science images will be important Archival researchers also need to be supported (next talk). Product quality is an observatory deliverable, thus the observatory must control calibration strategy. Challenge is to allow the flexibility that traditional radio astronomers expect.

Software Architecture Summary: No blockers The RMS community knows how to design, implement, and operate the software to run the ngVLA. Some changes in emphasis because of large numbers of antennas and baseline length. Radio community is gaining experience with full lifecycle software support (Proposal to Delivery). Lessons to be learned from current generation of NRAO telescopes. SRDP project is pioneering science quality image production across the frequency range of the ngVLA.