My slides for the Innovate UK e-Infrastructure SIG meeting in August 2014, introducing the work we have been doing with HPC Midlands to create a standard heads of agreement for HPC services, to make it easier for academic supercomputer centres to share their facilities with other institutions and with industry.
2. 2
1. Why an HPC Framework?
2. Why Jisc?
3. How will it work?
4. Wider E-Infrastructure
Opportunity
Introducing the Jisc National HPC Agreement
(Technology Strategy Board E-Infrastructure SIG, August 2014)
3. 3
1. Why an HPC Framework?
2. Why Jisc?
3. How will it work?
4. Wider E-Infrastructure
Opportunity
Introducing the Jisc National HPC Agreement
(Technology Strategy Board E-Infrastructure SIG, August 2014)
4. 4
E-Infrastructure facilities for High Performance Computing and
Big Data are critically important to many branches of academic
and industrial R&D.
Given the high capital cost of the equipment involved, many
organizations have told us they are interested in:
- Sharing the cost of putting new facilities into place.
- Purchasing access to facilities for “cloudbursting”.
- Business continuity during maintenance/outages or for
routine ongoing use.
::Why an HPC Framework? ::
7. 7
Jisc agreement developed in collaboration with EPSRC’s
HPC Midlands regional Supercomputing Centre of Excellence
(http://hpc-midlands.ac.uk)
Market tested via HPC Midlands with a wide range of firms
ranging from SMEs to large corporates.
Key issues around protection of intellectual property and
information assurance factored in from day one.
::Why an HPC Framework? :: Origins
8. 8
1. Why an HPC Framework?
2. Why Jisc?
3. How will it work?
4. Wider E-Infrastructure
Opportunity
Introducing the Jisc National HPC Agreement
(Technology Strategy Board E-Infrastructure SIG, August 2014)
9. 9
We are a registered charity championing the use of digital
technologies in research and education.
We run a wide range of services for Universities & Colleges, e.g.
- JANET, world leading NREN.
- Groundbreaking content deals with publishers.
- Cloud brokerage.
R&D achievements such as:
- IETF standards track Moonshot project.
- Pioneering work in Open Educational Resources, Open
Access and Open Data, preparing the sector for "open by
default”.
::Why Jisc? :: About us
11. 11
::Why Jisc? :: Next gen products/services
Technology and
Infrastructure
59%
Futures
6%
Content and resources
21%
Customer experience
14%
£96m income in 2013/2014,
here’s where it goes.
Less R&D (Futures) than
you might imagine.
A solution provider,
not a funder.
Increasing role of
institutional subscriptions.
12. 12
::Why Jisc? :: Brokerage
National agreements for:
- Arkivum
- Box
- Dropbox
- Google Apps
- Office 365
Peerings with major cloud
service providers including
Amazon, Google and
Microsoft.
More at:
http://goo.gl/sLEbo3
13. 13
Financial X-ray overview:
http://goo.gl/I38gd4
- Easily understand and compare overall
costs for particular services
- Develop business cases for changes to IT
infrastructure
- Create an ongoing mechanism for
dialogue between finance and IT
departments using standard terms of
reference
- Provide a means of highlighting the
comparative cost of shared and
commercial third party services
::Why Jisc? :: Benchmarking
14. 14
::Why Jisc? :: ESISS
Information Security advice:
- Automated and manual
penetration testing
- Reporting for audit
compliance purposes
- Security consultancy
- Led byTigerTeam
accredited Senior
SecurityTester
More at:
http://goo.gl/6Jz5I7
15. 15
1. Why an HPC Framework?
2. Why Jisc?
3. How will it work?
4. Wider E-Infrastructure
Opportunity
Introducing the Jisc National HPC Agreement
(Technology Strategy Board E-Infrastructure SIG, August 2014)
16. 16
Jisc will not be a party to contracts made under this agreement.
Agreement sets no expectations around costs and charging
models for service provision – these are for service provider and
customer to agree between themselves.
Jisc’s role is principally brokerage, ensuring that education
sector and UK Plc gain maximum benefit from public
investments in HPC and Big Data.
Agreement is in no way exclusive, and nothing in it prevents
providers from offering their services in other ways.
:: How will it work? :: Jisc’s vision
17. 17
We envisage that Jisc’s agreement will:
Serve as a “kite mark” of sorts for HPC service providers, giving
customers confidence that the service they are considering
purchasing is being run professionally.
Help the UK to move towards more of a market in HPC services,
and we are working with the Technology Strategy Board to
ensure that it meets the needs of both academia and industry.
Accelerate adoption by reducing friction of engagement.
:: How will it work? ::
18. 18
:: How will it work? :: Differentiation
Provider Standard
service
Enhanced
assurance
Specialist
hardware
(GPU etc)
Specialist
software
Consultancy
services
A
B
C
D
Examples of differentiation include: access to advice and
consultancy services, specialist software and hardware, and
enhanced information assurance e.g. where required for export
control compliance.
19. 19
1. Why an HPC Framework?
2. Why Jisc?
3. How will it work?
4. Wider E-Infrastructure
Opportunity
Introducing the Jisc National HPC Agreement
(Technology Strategy Board E-Infrastructure SIG, August 2014)
20. 20
::Wider E-Infrastructure Opportunity ::
HPC equipment is just one type of E-Infrastructure - other
examples include mass spectrometers, gene sequencers, 3D
printers, wind tunnels etc.
Standardised terms and conditions, SLA, and information
assurance regime (where required) “Argos Catalogue” of
equipment that institutions are sharing with each other and
with industry.
However in many cases the real challenge is the human
expertise required to exploit the equipment.
21. 21
::Wider E-Infrastructure Opportunity ::
HPC equipment is just one type of E-Infrastructure - other
examples include mass spectrometers, gene sequencers, 3D
printers, wind tunnels etc.
Standardised terms and conditions, SLA, and information
assurance regime (where required) “Argos Catalogue” of
equipment that institutions are sharing with each other and
with industry.
However in many cases the real challenge is the human
expertise required to exploit the equipment.