Weitere ähnliche Inhalte Ähnlich wie Engineering Simulation Meets the Cloud (20) Kürzlich hochgeladen (20) Engineering Simulation Meets the Cloud1. Dr. Dennis Nagy, BeyondCAE
April 8, 2014
8:00 AM PDT, 11:00 AM EDT, 17:00 CET
• * Engineering Simulation Veteran’s Series, Parts 1 and 2:
• “Where We Are and How We Got Here,” and
• “Where Are We Going?
were webcast on February 12th and 25th, 2014, respectively.
• Slides and recorded versions of both webcasts are available at
www.TheUberCloud.com
Engineering Simulation Meets The Cloud:
Current Status and Future Prospects
2. The Full Abstract (for reference only)
The Cloud is all the rage for the past couple of years and shows no signs of reduced hype any time
soon. What is “The Cloud?” There are many evolving viewpoint-centric definitions, which make it
even harder to separate the wheat from the chaff (remember the World-Wide Web run-up to the
dot.com bubble burst in the late 1990s?). This presentation will provide an overview of the Cloud-
relevant concepts and terminology, and their evolving practical implementations, from an
engineering simulation software viewpoint, covering HPC, Big Data, IaaS, PaaS, SaaS,…(had
enough already?). Beyond all these vendor/I.T.-driven concepts and viewpoints, what are
product/process development engineers and their employer enterprises actually trying to
accomplish more effectively? How does The Cloud fit in now and, more importantly, for the
near future.
Special emphasis will be given to how small and medium enterprises’ (SME’s—the under-served
“missing middle”) engineering needs are starting to be met by Cloud-related simulation access
opportunities. Computer hardware vendors seem to be leading the Cloud wave in general, but
major simulation software vendors are quite varied in their articulated Cloud strategies and
implementations so far, which has partly given rise to CAE-oriented start-ups emerging with a
Cloud-based value proposition. The UberCloud’s impressive success to date will be
summarized as one example of how simulation software vendors, hardware resource
providers, engineering end users, and HPC experts are combining their time and
interests, on a (so far) voluntary basis to discover, understand, and overcome the
practical hurdles in conducting industry-useful engineering simulations in The Cloud.
24 April 2014Copyright © 2014 by Dennis A. Nagy 2
3. Topics: Questions to Answer
At whom is this presentation targeted?
What are engineers actually doing?
How does Engineering Simulation (CAE) help
them do it “better”?
What are the obstacles to more engineers
making more “effective” use of Engineering
Simulation?
What is “The Cloud”?
Current Status of Engineering Simulation in
“The Cloud”
Future prospects for better engineering via
prudent use of “The Cloud”
The role of The UberCloud HPC Experiment
24 April 2014Copyright © 2014 by Dennis A. Nagy 3
4. Who Am I?
424 April 2014Copyright © 2014 by Dennis A. Nagy
A broad expert in engineering simulation (CAE), over 42
years of experience: from R&D, university teaching, through
commercial software development, support, sales and
marketing management, to executive management.
• Former Sr. VP of worldwide Sales at MSC.Software
• CEO of Engineous (now part of DS/SIMULIA)
• VP of Marketing and Business Development at CD-adapco
• VP of Marketing and Asia-Pacific at Fluent (now part of ANSYS)
• VP of International Business, Blue Ridge Numerics (now part of Autodesk)
Currently Principal at BeyondCAE, a global
strategy and business development
consulting activity located in Chapel Hill, NC, USA.
Mentor, TheUberCloud HPC Experiment
Member of the NAFEMS Americas Steering Committee
5. At Whom is this Presentation Targeted?
I have spent almost all my career on the vendor (ISV) side
However, there are plenty of sources of vendor (Cloud and ISV) views
on “how The Cloud does something beneficial for (fill in the blank)”
So I am trying to look at this topic from the engineering end-
user/management viewpoint and provide some introduction for that
audience to navigating the Cloud opportunities and hype.
24 April 2014Copyright © 2014 by Dennis A. Nagy 5
INNOVATION
6. Back to Basics:
What Are Engineers Actually Doing?
Engineers are developing (”designing” in a
physical/functional sense) products, systems, and processes
to meet (or even exceed!) a set of captured and understood
requirements
The functional requirements involve the expected physical
behavior of what is being developed
Anything that will assist engineers (and their corporate
management) in judging whether requirements will be met
is helpful
The “Design-Build (physical prototypes)-Test-Modify”
workflow cycle has been more and more
supplemented/replaced by Engineering Simulation
24 April 2014Copyright © 2014 by Dennis A. Nagy 6
7. How Does Engineering Simulation (CAE)
Help Engineers Do Their Work “Better”?
Provides the possibility of examining the functional
behavior of product/process designs much earlier and
more quickly than physical prototype testing
Can allow engineers to gain further functional insight (the
“aha” moments) and try more (and more radical “outside the
box”) ideas during development and/or allow for a significant
speed-up in the development cycle (freedom to trade off)
IF
…the simulations are “realistic enough” and “fast enough”
(joint judgment call by development engineers, simulation
experts, and management)
Engineering Simulation has reached this “if” milestone in
many major industries and larger manufacturing
companies today
24 April 2014Copyright © 2014 by Dennis A. Nagy 7
8. Advanced Information Technology Solutions:
An Engine of Innovation
24 April 2014Copyright © 2014 by Dennis A. Nagy 8
Greg Schroeder
Senior Research Engineer
Center for Automotive Research (CAR)
Ann Arbor, MI, USA
March 25, 2014
9. What are the Obstacles to More Engineers Making
More “Effective” Use of Engineering Simulation?
Despite simulation’s success so far, there are still key
obstacles
The majority of practicing engineers do not use simulation
(spreadsheets don’t count )
Users and potential users of simulation encounter three main
obstacles:
1. Cost of simulation software
2. Cost of adequate computer hardware to run it on
Budget justification of both is “chicken and egg”
3. Enough knowledge/expertise to use today’s generation of
engineering simulation software “effectively” (or even at all)
…but this is a topic for a whole different presentation
24 April 2014Copyright © 2014 by Dennis A. Nagy 9
Here is where
“The Cloud”
comes in
10. What is “The Cloud”?
Wikipedia:
In common usage, the term "the cloud" is essentially a metaphor for the
Internet.[2] Marketers have further popularized* (hyped?) the phrase
"in the cloud" to refer to software, platforms and infrastructure that
are sold "as a service", i.e. remotely through the Internet.
“Cloud computing is the next stage in the Internet's evolution…”
NetLingo:
A style of computing in which dynamic, scalable and virtual resources
are provided over the Internet. Cloud computing refers to services that
provide common business applications online, which are
accessed from a Web browser, while the software and data are
stored on the servers.
In other words, access to both computer power and data storage
capacity remotely via the Internet (browser or very thin client user
interface)
HPC (“high”-performance computing) HPCC
“Big Data”
24 April 2014Copyright © 2014 by Dennis A. Nagy 10
11. What is “The Cloud”?
NIST’s 7-Page “Definitive” Answer and Recommendations
(2011-12)
24 April 2014Copyright © 2014 by Dennis A. Nagy 11
“Cloud computing is a model for
enabling ubiquitous, convenient,
on-demand network access to a
shared pool of configurable
computing resources (e.g.,
networks, servers, storage,
applications, and services) that
can be rapidly provisioned and
released with minimal
management effort or service
provider interaction. This cloud
model is composed of five
essential characteristics, three
service models, and four
deployment models.” (Excerpt)
12. Public vs. Private Clouds
PUBLIC CLOUD:
All infrastructure components — including servers — are shared among clients
Potential contention for server resources
Lower cost than virtual private cloud
VIRTUAL PRIVATE CLOUD:
Servers dedicated to an enterprise
Offers greater control, security, and application performance than public cloud
Higher cost than public cloud
“Either one of these models [owned/operated by a 3rd party] is less expensive than
building your own or another type of infrastructure service.”
-- Lynda Stadtmueller, Program Director, Cloud Computing, Frost & Sullivan
ACTUAL PRIVATE CLOUD:
Your company owns the hardware, too, and your I.T. Department makes it look
like “A Cloud” to you, the end-user (browser-base access, no need to understand
what’s behind the curtain).
This may be cost-competitive with “external” Clouds, if there is enough
demand/use. The key here is “looks like a Cloud” = browser-based ease-of-use.
24 April 2014Copyright © 2014 by Dennis A. Nagy 12
13. Potential Benefits of The Cloud
for Engineering Simulation
When software and hardware access is priced
“appropriately,” both overflow needs and new-user
needs can be met incrementally (“by the drink”)
More realistic, affordable business model for small and
medium businesses/enterprises (SMB/SME) …“The
missing middle”
More streamlined user access/experience, to allow
engineers to focus on engineering instead of I.T. details
“Overflow” vs. “replacement of in-house I.T.
infrastructure by Cloud (=3rd-party remote) service
providers is a separate debate/discussion
24 April 2014Copyright © 2014 by Dennis A. Nagy 13
14. Potential Benefits of The Cloud
for Engineering Simulation
There are many articles and white papers
Last month: results of a LinkedIn user’s ad-hoc survey of
top benefits:
(go to LinkedIn’s “Cloud Computing, SaaS & Virtualization Group” for more detailed
discussions)
1. (18 votes) - no barrier (cost) to entry / capex vs opex
2. (13 votes) - elasticity / flexibility / agility
3. (3 way tie with 12 votes each) - on demand / pay per use / pay as
you go, time to market, and scalability
4. (5 votes) - self service / services / API / tools
5. (2 way tie with 4 votes each) - availability / fault tolerance, and
real-estate / location benefits / anywhere access
6. (4 way tie with 2 votes each) - risk management, security,
reduced hardware dependencies / mobility, and automation
24 April 2014Copyright © 2014 by Dennis A. Nagy 14
15. The Main Layers of The Cloud
24 April 2014Copyright © 2014 by Dennis A. Nagy 15
“Cloud computing
specifically refers to
incorporating
software as a service
(SaaS), platform as a
service (PaaS), and
infrastructure as a
service (IaaS). Users
do not need to have
knowledge of,
expertise in, or
control over the
technology
infrastructure in the
"cloud" that
supports them. “
-NetLingo
(From Frost & Sullivan)
17. SaaS Further Refinement
Common user interface/GUI as a service
Compute modules (batch “solvers”) as a service
This is relevant because it’s mainly (but not exclusively) the
number-crunching that can take most advantage of “Big Compute”
infrastructure (HPCaaS) to handle larget problems and run them
faster
Interactive vs. batch Cloud services
“where is the pre/post-processing done?
Cost and speed of data transmission to/from The Cloud
Visualization from The Cloud (pushing only pixels)
24 April 2014Copyright © 2014 by Dennis A. Nagy 17
18. Gartner’s Magic Quadrant for IaaS
24 April 2014Copyright © 2014 by Dennis A. Nagy 18
“I believe this chart is the most important
chart in enterprise IT. It covers the new
trend of IaaS, where instead of buying
storage, operating systems, deployed
applications, and other infrastructure,
enterprises rent these products from
vendors in a low risk pay-as-you-go
system. It is a new way of doing
enterprise computing …Computing
infrastructure is the $150B a year “Big
Kahuna”, where all the money changes
hands, and where companies from
Microsoft, to IBM to HP make all their
profits. … if IaaS becomes the dominant
computing model, Amazon Web
Services is the only possible winner “ *
– Rory O’Driscoll, Partner at SCALE
Venture Partners, October 31, 2013
* $250B + for Cloud architecture/infrastructure
by 2017?
So where is Google?
“…this market is highly competitive, and
that the situation may change very fast.”
21. 24 April 2014Copyright © 2014 by Dennis A. Nagy 21
Engineering Simulation (CAE):
Relevant Cloud Players
SimScale/AWS
Ciespace
Nimbis Services
AWS
Rescale
Cycle Computing
Nimbix
Sabalcore
CloudBroker GmbH
CPU 24/7
CPUsage
TotalCAE
AWS
HP
SGI Cyclone
Microsoft Azure
Gompute/Gridcore
Outscale
Autodesk
OCF enCORE (Hartree)
Fujitsu TC Cloud
Altair HWOD
Bull
Penguin OD
IBM
NCMS
Ohio SC - AweSim
CESGA (Spain)
SICOS/HLRS
HSRapperswil
FCSCL
San Diego SC
Nonprofit
(gov./universities)
CommercialCloudVendors
SaaSPaaSIaaS
23. 24 April 2014Copyright © 2014 by Dennis A. Nagy 23
Vendors:
Company Software SimScale/AWS
Ciespace
NimbisServices
AWS
Rescale*
CycleComputing
Nimbix
Sabalcore
CloudBrokerGmbH
CPU24/7
CPUsage
TotalCAE
AWS
HP
SGICyclone
MicrosoftAzure
Gompute/Gridcore
Outscale
Autodesk
OCFenCORE(Hartree)
FujitsuTCCloud
AltairHWOD
Bull
PenguinOD
IBM
NCMS
OhioSC-AweSim
CESGA(Spain)
SICOS/HLRS
HSRapperswil
FCSCL
SanDiegoSC
proprietary
AltairHWUA
MCAE:
Autodesk SIM 360 x x
ANSYS ANSYS-Mechanical x x x x x x x x x x x x x x
ANSYS Fluent x x x x x
COMSOL Multiphysics x x
Mathworks MATLAB/SIMULINK x x x x
Siemens PLM NX CAE x
DS/Simulia Abaqus x x x x x x x x x
CD-adapco Star-CCM+ x x x x x x x x
Livermore Softw. LS/Dyna x x x x x
MSC MSC.NASTRAN x x
MSC MSC.MARC x x
Exa PowerFlow
Convergent Converge x
ESI OpenFOAM x x x x x x x x x x x x
ESI CFD-ACE+ x
AVL AVL/Fire x
Cambridge Flow BOXERMesh x
Ciespace/CMU BubbleMesh x x
NASA CFL3D x
Friendship Systems Framework x
VR&D Genesys x
Fidesys Fidesys x
Altair Hyperworks x x x
Altair AcuSolve x x x x x
NUMECA Turbo-fine x
FluiDyna Culises x
FluiDyna LBultra x
Metacomp CFD++
Stanford U. SD++
Applied Math M. CoolSim x
MTU Aero CalculiX x
TotalCAE Portal x x
CSC-Finland Elmer x x
TESIS FlowVision x x
ECAE/EDA:
CST Microwave Studio x
ANSYS HFSS x
ANSYS SI-Wave x
Synopsys VCS x
Mentor HyperLynx SI x
CST Studio Suite x x
PaaS
(vendor-publicized or 3rd-party published information)
SaaS (government/universities)IaaS
Matrix of CAE Vendors and Cloud Providers (partial-work in progress)
Commercial Cloud Vendors Nonprofit In-house
I.T.
24. Cloud-related Efforts to Reach the
“Missing Middle” SMEs/SMBs
24 April 2014Copyright © 2014 by Dennis A. Nagy 24
Brief overview here; go to TheUberCloud
website (click on logo) for more details
Go to websites (click on logos) for
information
25. TheUberCloud HPC Experiment:
An open voluntary collaborative community
Objective:
Making HPC as a Service available, for everybody, on
demand
How?
For SMEs and their engineering applications
to explore the end-to-end process
of using remote computing resources,
as a service, on demand, at your finger tips,
and learning how to resolve the roadblocks.
2524 April 2014Copyright © 2014 by Dennis A. Nagy
26. Where Are We with the Experiment?
Started August 2012: today (April 2014) ~1000
participating organizations and individuals
Participants are from 48 countries
Round 5 started February 1, 2014: ~25-30 new teams
140+ teams have been formed in Rounds 1-5
Registration at:
www.hpcexperiment.com
www.cfdexperiment.com
www.compbioexperiment.com
www.bigdataexperiment.com
2624 April 2014Copyright © 2014 by Dennis A. Nagy
27. UberCloud Poll about Roadblocks
The UberCloud, June 13, 2013—but generally still true now
2724 April 2014Copyright © 2014 by Dennis A. Nagy
28. Cloud Security/Privacy: An Issue?
Security and privacy are in the eye of the beholder
Major Cloud datacenters and hosting are well aware of these issues
and have good technical and legal security/privacy infrastructure
Many “in-house” users of software and Internet access are not aware
of how (in)secure and exposed their data and computing really is
“The NSA has/sees everything anyway”
not really a joke for non-U.S. users/prospects for U.S.-based Cloud data
centers
Autodesk whitepaper:
Autodesk®360: Work Wherever You Are – Safely
“With over 10 years of experience providing cloud-based solutions,
Autodesk knows that data security is critical to your success.”
24 April 2014Copyright © 2014 by Dennis A. Nagy 28
29. Major Technical Hurdles Encountered
Downloading speed of large results files for local
visualization
Remote Visualization of results
General-purpose remote visualization tools
The “last mile” situation
Pixel-pushing and the speed of light
Complicated, non-intuitive Cloud access mechanisms
& details (ports, files,…)
24 April 2014Copyright © 2014 by Dennis A. Nagy 29
30. More Details on the
25 Best Teams So Far (Rounds 1 & 2)
UberCloud HPC Experiment Compendium
25 selected use-cases from 60 teams in Rounds 1 & 2
Sponsored by Intel’s HPC Organization
3024 April 2014Copyright © 2014 by Dennis A. Nagy
31. Further Cloud-Simulation Relevant
Information in LinkedIn Group Discussions
New Trends in CAE Simulation
The Cloud: What does it mean for software re-sellers?
Software Licensing in the Cloud
CFD/CAE software on Amazon Web Service (AWS) cloud
The UberCloud invites you to a free experiment with Autodesk
Sim 360 in the Cloud
2013, Marked by Cloud and Changing Licensing Methods
Why the CAE in cloud not picking up ?
Engineering Simulation (CAE) in "The Cloud"?
Cloud Computing, SaaS & Virtualization
What are the top 3 benefits of cloud computing?
The evolution of the hybrid cloud
Amazon vs Google? This "cloud war" is like Target vs Walmart
Welcome to the cloud price wars
IAAS , Cloud Computing , HPC , Parallel Computing
24 April 2014Copyright © 2014 by Dennis A. Nagy 31
32. The Cloud as a Utility: Future Vision
The toaster analogy (today):
The engineer using The Cloud
(in the future):
24 April 2014Copyright © 2014 by Dennis A. Nagy 32
33. 24 April 2014Copyright © 2014 by Dennis A. Nagy 33
“Sometimes, it’s hard to tell when a novelty becomes a trend or a trend becomes the
new normal. This is not one of those times. The era of the on-premise server is
clearly behind us, with the cusp of change literally on our calendars.”
--Peter Coffee, Salesforce.com March 29, 2014
The Cloud is Coming to Engineering Simulation (or vice-versa):
It’s Only a Question of How Rapidly
34. Some Other Terms
Apache CloudStack is a top-level project of the
Apache Software Foundation (ASF). The project
develops open source software for deploying public
and private Infrastructure-as-a-Service (IaaS) clouds.
Hadoop is a free, Java-based programming framework
that supports the processing of large data sets in a
distributed computing environment. It is part of
the Apache project sponsored by the Apache Software
Foundation.
24 April 2014Copyright © 2014 by Dennis A. Nagy 34
35. PDF copies of these slides (with active hyperlinks) available from
TheUberCloud: www.TheUberCloud.com
or from me at:
35
Dennis.Nagy@BeyondCAE.com
24 April 2014Copyright © 2014 by Dennis A. Nagy
Hinweis der Redaktion Let’s start with a brief look at where we are with the experiment. Since the value each of you individually get out of the experiment increases by the number of total participants, it is important for us to follow this indicator closely.We currently have 400 participants and about 100 of that is on the call today. So a big thank you for taking the time not to only participate in the experiment but to attend the half time webinar.