The Structure of the Universe: Cosmology, Exoplanets and Lattice QCD

Lead Research Organisation: University of Cambridge
Department Name: Applied Maths and Theoretical Physics

Abstract

This proposal aims to address key questions about the fundamental structure of the Universe and the origin and nature of the galaxies, stars and planets within it. This proposal is to enhance the STFC DiRAC Facility which provides the primary computational platform for UK particle physicists, cosmologists and astrophysicists. The proposal will fund two high performance computing systems: First, there is a flexible shared-memory (SMP) node with 16TB, which will be the largest such system in Europe. Secondly, there is a powerful tightly-coupled cluster capable of 200Tflops (HPCS). This proposal will advance our understanding in 4 key scientific areas:

A. Science exploitation of the CMB and Large-scale Structure surveys:
The cosmic microwave background (CMB) remains the premier source for cosmological information. Planck satellite data dramatically supersedes the previous WMAP data and we will use it to constrain fundamental cosmology. With Planck data releases scheduled for 2013 and 2014, this upgrade will leverage the proprietary data, maximizing science exploitation by COSMOS members. Work on large-scale structure encompasses using surveys to constrain the properties of the Universe and understanding the hierarchical formation of galaxies.

B. Observational consequences of the Early Universe.
The consortium has pioneered the use of lattice simulations to understand the physics of non-linear phenomena in the early universe. This is expected in most cosmological models, including those with phase transitions, cosmic defects and extra dimensions. The challenge using the new SMP node will be to calculate the observable consequences of these theories, which can range from signatures in the CMB or large-scale structure through to the production of dark matter or primordial gravitational waves.

C. Extra-solar planets and their atmospheres.
The SMP node will also support key UK research in extra-solar planets. This is concerned with the observation and characterisation of exoplanets and their atmospheres, developing key numerical codes which are vital to this international endeavour. The new SMP system will allow much more information to be extracted from spectroscopic data from exoplanet environments. This will help us answer some of the oldest questions in science such as: Are there worlds beyond our solar system? Are they numerous or rare? How many of them have the right conditions for life?

D. Quantum chromodyanamics (QCD)
We shall use the HPCS to carry out the research programme of the UKQCD Consortium: numerical solution of QCD and related theories using techniques known collectively as lattice field theory.
QCD describes how quarks bind into observable particles known as hadrons. By solving QCD on supercomputers, we gain information about the rich physics which results from the strong interactions between quarks. Furthermore, precise QCD calculations are necessary to connect experimental measurements to fundamental parameters governing how quarks change their type, or "flavour."
So far, the Standard Model (SM) of particle physics (which unites the electromagnetic, weak, and strong forces) appears to be sufficient to describe quark flavour-changing interactions. However many physicists expect there to be physics "beyond" the Standard Nodel (BSM); precise measurements combined with accurate QCD computations may reveal hints of this new physics.
In several cases, lattice QCD is the only known method for achieving the few-percent precision necessary to distinguish between Standard and non-Standard physics.

Lattice field theory also offers accurate methods for exploring field theories with strongly interacting particles. UKQCD groups are active in categorizing possible BSM theories. In addition, UKQCD are studying these types of theories under extreme conditions such as high temperature and/or high particle density. These calculations complement experimental programs at RHIC and LHC.

Planned Impact

The primary goal of the COSMOS consortium is to enable UK researchers to maintain international leadership as they advance the confrontation between fundamental and observational cosmology. COSMOS work has a significant economic impact through (I) training in high performance computing and (II) collaboration and interaction with HPC technology developers.
The COSMOS consortium has the largest user base of the STFC HPC consortia (20..40 new users annually, mainly PDRAs and graduate students) and maintains a special emphasis on user accessibility and education, ensuring maximum benefit to UK science and subsequent employers. Significant numbers of graduate students and postdoctoral fellows have been and continue to be trained and given transferable skills, including analysis and problem-solving, software engineering and numerical simulation and techniques for massively parallel programming. The computational techniques employed involve a constant interplay between HPC codes and intuition on systems of huge inherent complexity. This sort of activity is highly regarded in the business world and in high-tech industries, providing an important platform for future employment.
The COSMOS consortium also contributes to the public understanding of science through its high media profile, thereby improving the social acceptability of new technology and attracting young people to consider careers in the sciences. In addition, COSMOS consortium members have had a high public profile and are engaged in outreach. Cosmology remains of major interest to the public. The group contains members who are particularly active in maximizing the impact of our science in the outside world to audiences of all sorts. Hawking is probably the most well known scientist in the world and his lectures, television coverage, and books have very widespread impact that acts as an important standard-bearer for the whole area of the Research Council's scientific work in cosmology and fundamental physics.

The Miracle consortium will continue and broaden its researchers' current involvement in Knowledge Exchange and Outreach programmes:
(i) Quantemol, a UCL-based 'spin-out' firm related to the Miracle Exomol project
(ii) Providing accurate and complete molecular data to Industry and climate research programmes
(iii) Working with commerce, research design labs and other academic areas to produce new IT (hardware and software) solutions transferrable to Industry and academic areas such as Biomedicine;
(iv) Working with leading IT vendors to test and design new hardware and software solutions which will eventually end up in the marketplace.

UKQCD:
The demands for HPC facilities in the research in lattice field theory helps stimulate technological and algorithmic development in that area. Several of our PhD students and postdocs have moved into education, science industry, financial services, and biological sciences research. Their training in physics and mathematics, in problem solving, in numerical techniques, and in the arts of report writing and making presentations makes them valuable to the UK economy. We engage in outreach activities directly associated with this research. In Cambridge, we give talks in schools and present the physics underlying our research at University open day. A group of major users of this facility with UKQCD is the High Precision QCD collaboration of which the UK members are in the universities of Cambridge, Glasgow and Edinburgh. The HPQCD website, www.physics.gla.ac.uk/HPQCD, gives links to "HPQCD in the news" which directly shows the impact of our research on the wider community. We undertake outreach at many levels; for example, at the University of Glasgow members of UKQCD give Open Day talks and contribute to
masterclasses for school students and to public lectures.

Publications

10 25 50
publication icon
Achúcarro A (2014) Evolution of semilocal string networks: Large-scale properties in Physical Review D

publication icon
Al-Refaie A (2015) ExoMol line lists - VIII. A variationally computed line list for hot formaldehyde in Monthly Notices of the Royal Astronomical Society

publication icon
Ali-Haïmoud Y (2013) An efficient implementation of massive neutrinos in non-linear structure formation simulations in Monthly Notices of the Royal Astronomical Society

publication icon
Amin M (2013) A scattering theory of ultrarelativistic solitons in Physical Review D

publication icon
Amin MA (2013) Clash of kinks: phase shifts in colliding nonintegrable solitons. in Physical review letters

 
Description High energy grazing collisions of black holes
Amount $48,207 (USD)
Funding ID NSF-XSEDE Grant No. PHY-090003 
Organisation National Science Foundation (NSF) 
Sector Public
Country United States
Start 01/2017 
End 12/2017
 
Description Probing fundamental physics with gravity
Amount € 1 (EUR)
Funding ID RACE grant Tier-0 PPFPWG 
Organisation European Commission 
Sector Public
Country European Union (EU)
Start 12/2016 
End 11/2017
 
Description Intel/SGI/COSMOS Collaboration 
Organisation Intel Corporation
Country United States 
Sector Private 
PI Contribution The present relationship is a longstanding collaboration and partnership in high performance computing with SGI since 1997 and with Intel since 2003. On the basis of taking delivery of the largest deployment of Xeon Phi processors in a single-image shared memory system, we have been testing the architecture and system stack while developing codes to exploit this new many-integrated core (MIC) technology. We influenced design of the initial "MG" blade accommodating the Xeon Phi within the SGI UV2, and we have worked with Intel to improve their Xeon Phi offload implementation.
Collaborator Contribution Our partner SGI has contributed half the salary GPB£18000 so that COSMOS could employ a parallel programmer, James Briggs. Intel have contributed 50% of an Intel employee, John Pennycook, to assist with development work and to act as a direct link to Intel software engineering teams. SGI has also contributed substantial software consultancy support through Cheng Liao which has led to new scalable SMP applications.
Impact There have been many positive outcomes from this collaboration, notably the Xeon Phi acceleration of important COSMOS consortium codes, including Camb (main CMB cosmological parameter estimation code), WALLS (early universe field theory code), MODAL (non-Gaussianity code), etc. A major development has been the implementation of a scalable shared-memory eigensolver capable of decomposing dense matrices up to 1M x 1M; this has important applications for the TROVE pipeline, determining exoplanet spectral lines (and elsewhere). Work from this collaboration was reported at Supercomputing 2013, held in Denver, November 2013.
Start Year 2013
 
Description Intel/SGI/COSMOS Collaboration 
Organisation Silicon Graphics Inc
Country United States 
Sector Private 
PI Contribution The present relationship is a longstanding collaboration and partnership in high performance computing with SGI since 1997 and with Intel since 2003. On the basis of taking delivery of the largest deployment of Xeon Phi processors in a single-image shared memory system, we have been testing the architecture and system stack while developing codes to exploit this new many-integrated core (MIC) technology. We influenced design of the initial "MG" blade accommodating the Xeon Phi within the SGI UV2, and we have worked with Intel to improve their Xeon Phi offload implementation.
Collaborator Contribution Our partner SGI has contributed half the salary GPB£18000 so that COSMOS could employ a parallel programmer, James Briggs. Intel have contributed 50% of an Intel employee, John Pennycook, to assist with development work and to act as a direct link to Intel software engineering teams. SGI has also contributed substantial software consultancy support through Cheng Liao which has led to new scalable SMP applications.
Impact There have been many positive outcomes from this collaboration, notably the Xeon Phi acceleration of important COSMOS consortium codes, including Camb (main CMB cosmological parameter estimation code), WALLS (early universe field theory code), MODAL (non-Gaussianity code), etc. A major development has been the implementation of a scalable shared-memory eigensolver capable of decomposing dense matrices up to 1M x 1M; this has important applications for the TROVE pipeline, determining exoplanet spectral lines (and elsewhere). Work from this collaboration was reported at Supercomputing 2013, held in Denver, November 2013.
Start Year 2013
 
Description 6th form talk (Chelmsford) 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Schools
Results and Impact October 2017: Ulrich Sperhake gave a talk to a 6th form college in Chelmsford. Subject: The direct detection of gravitational waves by LIGO and the insights already obtained from this new window on the Universe as well as the enormous potential for our search for answers to the most profound questions in astrophysics and fundamental physics.
Year(s) Of Engagement Activity 2017
 
Description IOP Engineering Society (Chelmsford) 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Professional Practitioners
Results and Impact October 2017: Ulrich Sperhake gave a talk to the IOP Engineering Society in Chelmsford. Subject: The direct detection of gravitational waves by LIGO and the insights already obtained from this new window on the Universe as well as the enormous potential for our search for answers to the most profound questions in astrophysics and fundamental physics.
Year(s) Of Engagement Activity 2017
 
Description Science Festival 2013 
Form Of Engagement Activity Participation in an open day or visit at my research institution
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Public/other audiences
Results and Impact Substantial effort was made during the Cambridge Science Festival to publicise the role of the COSMOS supercomputer at the Centre for Mathematical Sciences, where hundreds of members of the general public visited on Saturday 23 March 2013. Group tours were offered
every half hour which proved a particular highlight for the Open Day. The role of COSMOS for key science exploitation of
the Planck satellite (announced on 21st March with considerable media coverage) was also emphasised.

Considerable interest shown by members of the public.
Year(s) Of Engagement Activity 2013
 
Description Science Festival 2015 
Form Of Engagement Activity Participation in an open day or visit at my research institution
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Public/other audiences
Results and Impact Science Week 2015 within the University of Cambridge includes a popular open day at the Centre for Mathematical Sciences in which we participated actively. Apart from visualisation and other poster displays, a particularly popular activity was the tour of the COSMOS supercomputer.
Year(s) Of Engagement Activity 2015
 
Description Web sites 
Form Of Engagement Activity Engagement focused website, blog or social media channel
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Public/other audiences
Results and Impact Our Centre for Theoretical Cosmology and COSMOS supercomputer websites include news stories of wide interest to the general public as well as more specialist audiences.

These can be found at:

www.ctc.cam.ac.uk
www.cosmos.damtp.cam.ac.uk
Year(s) Of Engagement Activity 2012,2013,2014,2015,2016
URL http://www.ctc.cam.ac.uk