GridPP Public Service Summit

1
Big Questions and Big Answers
Professor David BrittonUniversity of
Glasgow IET, October 2009
2
(No Transcript)
3
The very big
On the largest distance scales the Universe
appears smooth with small fluctuations
Galactic super-cluster
Solar system
Galaxy
Universe
1012 m
1017 m
1023 m

• 1026 m

4
The very small
electron
nucleus
proton neutron
up quark down quark
10-10 m (thickness of human hair 10-5 m)
10-14 m
10-15 m
lt 10-18 m
5
Standard Model of Particle Physics

• There are 6 Quarks and 6 Leptons arranged in 3
  Generations of increasing mass.
• Each particle has a similar anti-particle
  with opposite properties.
• They interact via 4 force carrying particles
  the photon (?), gluon (g), Z and W.
• With known particles, theory is incomplete all
  particles would be massless.

Electro-magnetism
Why 3?
Strong Interaction
Where are they?
Weak Interaction
Why?
Matter as we know it all made from first
generation.
Origin of mass?
6
The Higgs Particle
Prime (but not only) objective of the LHC is to
understand the origin of mass is it a universal
Higgs field? (Like an electric field everywhere
but with the Higgs particle, rather than a
photon, as the exchange particle).
7
Higgs Mass
8
Big Questions

• What are the most basic building blocks of
  matter?
• Can the forces between particles be understood in
  a unified framework? How does gravity fit it?
• What unknown properties of these particles and
  forces drove the evolution of the universe from
  the Big Bang to its present state?
• Why is there more matter than antimatter in the
  Universe? What is the origin of this asymmetry
  (CP-violation)?
• What is the origin of mass (does the Higgs
  particle exist?)
• Why are there three generations?
• What is the cold dark matter that makes up 25 of
  the universe?
• Where/what is the missing (dark) energy that
  forms 70 of universe?
• Does supersymmetry exist?
• Are there extra dimensions?

9
Big Answers
10
The Large Hadron Collider
Mont Blanc
The LHC is a 27km accelerator that collides
counter-rotating beams of protons, of up to 7 TeV
each giving 14 TeV total energy. Energy is
determined by the circumference and the maximum
feasible magnetic field (8.4T). The energy
densities of the first fractions of a second
after the big-bang will be recreated, allowing
the Higgs to be momentarily produced.
Geneva
Airport
Switzerland
France
Switzerland
France
11
The LHC Complex
Protons accelerated in stages to 0.999999 speed
of light by a sequence of accelerators.
12
Building the LHC
In the main ring 1746 superconducting magnets
including 1232 15m SC dipoles weighing 27
tonnes each producing 8.36 Tesla and
running at 270c cooled by 700,000L liquid He
and 12 million of liquid N2 producing the
coldest place in the universe.
13
The LHC Complete
14
The LHC Experiments
100m
15
The LHC Detectors

• CMS
• General purpose detector
• 1,800 scientists from over 150 institutes

• ATLAS
• General purpose detector.
• 2,000 scientists from 34 countries

16
The CMS Detector
Brass/scintillator
tiles ? lt 5.0
HCAL
Total weight 12,500 t Overall diameter 15
m Overall length 21.6 m Magnetic field 4 T
17
(No Transcript)
18
(No Transcript)
19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
(No Transcript)
27
(No Transcript)
28
(No Transcript)
29
(No Transcript)
30
(No Transcript)
31
(No Transcript)
32
(No Transcript)
33
(No Transcript)
34
(No Transcript)
35
(No Transcript)
36
(No Transcript)
37
(No Transcript)
38
(No Transcript)
39
(No Transcript)
40
(No Transcript)
41
(No Transcript)
42
(No Transcript)
43
(No Transcript)
44
(No Transcript)
45
(No Transcript)
46
(No Transcript)
47
(No Transcript)
48
(No Transcript)
49
(No Transcript)
50
(No Transcript)
51
(No Transcript)
52
(No Transcript)
53
(No Transcript)
54
(No Transcript)
55
(No Transcript)
56
(No Transcript)
57
(No Transcript)
58
(No Transcript)
59
(No Transcript)
60
(No Transcript)
61
(No Transcript)
62
(No Transcript)
63
(No Transcript)
64
(No Transcript)
65
(No Transcript)
66
The CMS Barrel arrives
67
The ATLAS Detector
68
Mr Big and Mr Heavy
6 Floors
5 4 3 2 1 0
69
Final Piece
The Final Piece Closure of the LHC beam pipe
ring on 16thJune 2008 ATLAS was ready for data
taking in August 2008
70
Commissioning with Comics
Millions of cosmic muons recorded, allowing the
detector components to be commissioned and
aligned.
Access shaft
71
Start-Up 10th Sep 2008
LHC Control Room
ATLAS Control Room
72
Shutdown 19th Sep 2008
9744 high-current splices required with
resistance of less than 1 nW at 1.9k and less
than 10 mW after a quench.
In the last commissioning step in the last
sector, one splice had too large a resistance.
73
Catastrophic Consequences
Multi kA electrical arc
Helium enclosure punctured by arc and liquid
released into insulating vacuum chamber,
expanding by a factor of 1000x as it became a
gas. Relief valves opened but couldnt handle
volume. Pressure wave travels along accelerator.
Magnets Displaced
Vacuum Chamber Contamination
74
Recovery

• Repair of the damage
• 53 magnets returned to the surface for
  repair/replacement
• Vacuum chamber cleaned.
• Re-align interconnections cool-down power
  test.
• Prevention of similar incidents
• Search for indicators of faulty connections in
  commissioning data
• Retests of all suspicious locations
• Improvements on Quench Protection Systems
• Mitigation of consequences in case of similar
  incidents
• Increase number and discharge capacity of relief
  valves
• Reinforce external anchoring of cryostats at the
  locations of the vacuum barriers

75
Current LHC Status
76
Particle Collisions - Events

• TABLE OF MILLIONS
• Collisions 40-million times a second.
• 600-million protons collide per second.
• 150-million electronic channels.
• 15-million gigabytes of data recorded per year
  (one in a million events).
• Expect 1 in 10-million to be Higgs events.

• The raw data recorded is later reconstructed,
  filtered, and analysed (multiple times)!
• Comparable sized simulated data sets are
  generated to understand detector characteristics
  and backgrounds.

77
Computing Requirements
Starting from events like this
..we are looking for the Higgs signature that
may be there in 1 in 10 million events.
Requirements (2008 originally) CPU Power -
100,000 cores DISK storage of 40 PB Tape storage
of 40 PB requirements double by next year.
1998 Grid concept Foster, Kesselman
78
Web vs Grid
Web Focused on sharing information (high level
data - text, picture, music, video, etc.) Allows
a limited set of predetermined actions (data
processing) such as search, filter, sort, stream,
etc.
Grid The idea is to share storage and computing
power more directly, enabling much larger data
sets to be shared with user determined data
processing.
79
Why Grid?
80
Middleware
Your Program
Single PC
Grid
Application Layer
Your Program
User Interface Machine
Word/Excel
Games
Email/Web
Middleware Services
OPERATING SYSTEM
Middleware Services
Middleware Services
CPU
Disks, CPU etc
CPU Cluster
CPU Cluster
CPU Cluster
Disk Server
Middleware is the Operating System of a
distributed computing system.
81
Middleware
GRID INTERFACES (e.g. Ganga)
EXPERIMENT FRAMEWORKS
PRODUCTION/ANALYSIS SYSTEMS
AUTHORISATION/ROLES VIRTUAL ORGANISATION
MEMBERSHIP (VOMS)
BATCH SUBMISSION WORKLOAD MANAGEMENT SYSTEM
(WMS)
DATA MOVEMENT FILE TRANSFER SERVICE (FTS)
GRID MIDDLEWARE
STORAGE INTERFACE STORAGE RESOURCE MANAGER (SRM)
METADATA/REPLICATION LCG FILE CATALOGUE (LFC)
DISTRIBUTED CONDITIONS DATABASES ORACLE STREAMS
(3D)
Tier 0, Tier 1, Tier 2
WLCG FABRIC
82
Example workflow
VOMS
WMS
83
Tier Structure
CERN computer centre
Tier 0
Offline farm
RAL,UK
France
Italy
Germany
USA
Tier 1 National centres
Online system
Tier 2 Regional groups
ScotGrid
NorthGrid
SouthGrid
London
Glasgow
Edinburgh
Durham
Institutes
Useful model for Particle Physics but not
necessary for others
Workstations
84
Worldwide Resources
Worldwide

• 56 Countries
• 295 Sites
• 180,000 CPUs

UK

• 22 Sites
• 20,000 CPUs

(October 09)
85
GridPP

• UKs contribution to LHC computing
• 19 UK Universities, STFC and CERN
• GridPP1 (2001- 2004) 17M
• From Web to Grid
• GridPP2 (2004 - 2008) 16M
• From Prototype to Production
• GridPP3 (2008 2011) 30M
• From Production to Exploitation

May 2001 Oxfordshire
85
86
UK Resources
ScotGrid Durham, Edinburgh, Glasgow NorthGrid Da
resbury, Lancaster, Liverpool, Manchester,
Sheffield SouthGrid Birmingham, Bristol,
Cambridge, Oxford, RAL PPD London Brunel,
Imperial, QMUL, RHUL, UCL
New kit from Viglen at QMUL
Tier-1 4,500 PC Equivalents, 2.3PB Disk, 5PB Tape
Tier-2 10,000 PC Equivalents, 2PB Disk
New machine room in Glasgow
87
What is a Grid good for?
Problems that are highly parallelizable
Input data is independent e.g. Images
Simulation using different parameters
These pieces may be independent
Not so good for closely coupled problems
These pieces will have to interact
88
Other Applications
89
Other Grid Users
90
A final big question
Why pursue these big, and expensive, questions?
Answer-1 To advance human progress through basic
knowledge. Without such innate curiosity, the
modern world would not exist.
91
A final big question
Why pursue these big, and expensive, questions?
Answer-2 Research in curiosity-driven science is
an important driver for technological and
economic success.
92
A final big question
Why pursue these big, and expensive, questions?
Answer-3 Skills transfer Our future economic
competitiveness depends on maintaining a strong
technological base and a highly-skilled work
force. About one-half of all particle physics PhD
students eventually take up high-level careers in
business and industry. Particle physics is the
key subject in attracting young people to take up
science.
93
A final big question
Why pursue these big, and expensive, questions?
Answer-4 Direct economic impact Every 1 paid
by CERN to an industrial firm has been shown to
generate 3 of economic utility in the form of
increased turnover and cost-savings
94
Summary
95

• The END

96
Grid vs Cloud

• Grid computing is a collection of servers
  that are clustered together to attack a single
  problem. For a period of time, the entire
  resources of the grid are available to an user to
  tackle a particularly difficult compute problem.
  The engineering of such a grid requires complex
  inter-cluster networking, and usually the tuning
  of a grid is not for the faint of heart. Grid
  computing has been used in environments where
  users make few but large allocation requests. For
  example, a lab may have a 1000 node cluster and
  users make allocations for all 1000, or 500, or
  200, etc. So only a few of these allocations can
  be serviced at a time and others need to be
  scheduled for when resources are released. This
  results in sophisticated batch job scheduling
  algorithms of parallel computations. Cloud
  Computing is where users can get more or less
  resource on-demand. The critical difference
  vis-a-vis a grid is that a single user at a given
  point only gets a small portion of the utility or
  the cloud. Cloud computing really is about lots
  of small allocation requests. The allocations are
  real-time and in fact there is no provision for
  queueing allocations until someone else releases
  resources. This is a completely different
  resource allocation paradigm, a completely
  different usage pattern, and all this results in
  completely different method of using compute
  resources.

97
Supersymmetry

• Conventional method to fix Higgs mass
• Invoke SUSY
• Double the number of states in model
• Fermion/boson loops cancel
• Can also help to unify forces
• 105 new parameters (MSSM)

gt SUSY is a good pension plan for
experimentalists!
98
Dark Matter at the LHC?

• Characteristic signature for Dark Matter
  production at ATLAS
• Missing Transverse Energy (MET)
• Valid for any DM candidate
• (not just SUSY)
• Observation of MET signal necessary but not
  sufficient to prove DM signal (DM particle could
  decay outside detector)

Combine LHC and Astroparticle physics data in
order to prove that the neutralino hopefully
observed at LHC would be the DM particle
99
Extra Dimensions

• Hypothesize that there are extra space dimensions
• Volume of bulk space gtgt volume of 3-D space
• Hypothesize that gravity operates throughout the
  bulk
• SM fields confined to 3-D
• Then unified field will have diluted gravity,
  as seen in 3-D
• If we choose n-D gravity scaleweak scale then
• Only one scale -gt no hierarchy problem!
• Can experimentally access quantum gravity!
• But extra dimension is different scale from
  normal ones
• -gt new scale to explain

Extra dimensions are more of a lottery bet than a
pension plan!
100
BLACK HOLES AT THE LHC

• If two particles pass close enough with enough
  energy, they may form a black hole

• For 3 spatial dimensions, this will never happen
  gravity is too weak. But with extra dimensions,
  gravity may become stronger, micro black holes
  can be created in particle collisions

101
BLACK HOLES SEARCH FOR EXTRA DIMENSIONS
Theories which try to explain why gravity is so
much weaker than the other forces Gravity may
propagate in 4n dimensions, but we could
see strong effects only at very small distances,
reachable in pp LHC collisions
102
SEARCH FOR EXTRA DIMENSIONS
If theories with Extra Dimensions are true,
microscopic black holes could be abundantly
produced and observed at the LHC
They decay immediately through Hawking radiation
Simulation of a black hole event with MBH 8
TeV in ATLAS