The Changing Business Case for Supercomputing: An Industrial Perspective

1
The Changing Business Case for Supercomputing An
Industrial Perspective

• Dr. Kenneth W. Neves
• Senior Technical Fellow
• Manager, Computer Science
• Seattle, WA

2
Topics

• Indicators of market health and viability of
  supercomputing
• 1970
• late 1980s early 1990s
• Today
• Boeing high performance computing challenges
• Production computing
• Research computing
• Enterprise-wide computing
• Product visualization
• Conclusions Common research issues
• technical
• system

3
Key Factors to Monitor

• Market for high performance computers
• Applications - the need
• Computer Power
• Computer Architecture

4
Concepts of Key Factors
Big Market - sustained by commercial sales and
not just research
Very Novel
Big Market
Pacing
100X
Pacing - the next generation applications are
fundamental to business success
100x - Supercomputers offer 2 orders of magnitude
over next best alternative
Very novel - to achieve per- formance the
architecture requires large modifications of
existing software
Market
Need
SC Power
Architecture
5
1970s

• The market was new
• The requirements were scientific and led directly
  to improved products, research, and understanding
• The performance over the next most powerful
  market-based machines was enormous
• Required vector computing understanding, yet most
  applications had long loops to exploit
• oil, aero, structures, weather

Very Novel
Big M
Pacing
100X
6
Late 1980s- Early 1990s
Very Novel
Big
Pacing
100X

• Market split into vector parallel and highly
  parallel
• Vector parallel was well understood with a base
  of applications
• The performance of vector machines relative to
  other alternatives began to wane
• micros turned new generations every 18 months
• custom hardware lost edge
• The new breed parallel computers lacked software
  base and were very novel and hard to use

Vector
Very Novel
Big
Pacing
100X
Parallel
7
Now Facts of Life

• Today, SC companies have all but died or been
  absorbed into a more commodity market
• Micros dominate
• Cutting edge computational research MUST resort
  to highly parallel machines (separates the men
  from the boys)
• The cost of novel architectures both in
  hardware and software has thinned the market
• Many supercomputer users of old are workstation
  users today

Like 1970
Parallel
8
Boeing Applications

• CAD/CAM (billion dollar investment)
• Product Data Management and Manufacturing
  Resource Control (multi-billion dollar
  investment)
• Scientific Computing (important, but
  multi-million dollar investment) that tends to be
  cyclic
• Super Computing Problems, e.g.,
• CFD highly separated flows
• multi-disciplinary optimization
• constrained design
• electromagnetics

9
High-end Computing Activity

• Production computing
• Scientific research computing
• Enterprise-wide computing
• Product visualization

10
Production Computing

• Requires repeatable, controllable process
• Can be big problems (CFD for cruise wing design,
  structural analysis)
• Done on more ordinary architectures (Cray T-90)
• Migration from central computing
• as workstation and server capability improved
  many of the central users migrate to more
  affordable environments
• department level supercomputers
• application dedicated platforms (can be novel
  architectures, but not shared with many users)
• secret computing

11
Scientific Research Computing

• Grand challenge problems are often
  multidisciplinary, can involve optimization
• Often offer opportunity for macro-level
  parallelism
• Airfoil Constrained Optimization

0.1
12
Unconstrained
13
With Manufacturing Constraints
14
Factory Modeling
Models physics of metal cutting
15
Enterprise-wide Computing

• Distributed data
• 700 terabytes
• 20 business units
• secure, reliable, coherent
• Parallel SMP servers
• Oracle as middleware for 4 major applications
• Re-engineering of 315 legacy applications
• 50,000 users world wide (not including
  subcontractors)

16
Enterprise System Complexity
UFS File Server
DCE Security Server
Sequent Clusters
Master NIS
Scheduling Server
BNN Token Ring
Data Center FDDI Ring
Vital Production Systems
NFS Cluster (ServiceGuard)
Utility/Method Servers Clusters (ServiceGuard)
Routers
NT Resource Server (S3, Print)
Routers
NT WINS and MAD
NIS
Campus Server Room FDDI Ring
Router
DHCP Server
Application Servers (BaaN, Cimlinc,
ShopView Capp, Linkage, Web)
STAC Servers
DNS/NFS Cluster (ServiceGuard)
Switches
Printers Workstations
17
Product Visualization
Machining from CAD Generative Design Neural
Network design retrieval System complexity rivals
enterprise wide computing
ALSO
18
Research Issues

• Goal The network is the computer
• Power Grid (NASA term)
• computing resources are managed like a power
  system
• data movement is minimized, access time is
  minimized
• fail safe
• networking queuing, agent assisted
• Threads maintained
• Synchronization of process managed by middleware
  rather than individuals
• data authentication and time stamping for
  coherency
• Parallel data based performance (unsolved
  problem)
• Scientific computing approach, but applied to new
  application areas of the enterprise

19
Old Style Performance Enhancement
CPU Time
Analysis Application
20
New Style Performance Enhancement
CPU Time
21
What Questions to Ask
22
What Questions to Ask
23
NASA Power Grid Concept
24
System Performance Pyramid
Storage Systems
25
Conclusions

• Older performance improvement techniques are
  fundamental and necessary, but not sufficient
• New system level attack on performance and
  scalability is needed
• need to address response time
• system throughput (of the entire process)
• Looking at performance for (system level)
  analysis is similar to enterprise-wide computing
• Scientists, hardware vendors from the SC
  community, computer scientists, and
  enterprise-wide system developers need to
  collaborate
• The traditional supercomputing community needs to
  diversify its interests!