Transcript USC CSSE ARR 2011 - Center for Software Engineering

Why 21st Century Software Engineering is More
Like French Fries than Ever
Eric M. Dashofy
Computer Systems Research Department
The Aerospace Corporation
CSRD/CSTS/ETG
March 8, 2011
1
© The Aerospace Corporation 2011
I’ll give you the answer up front…
•
It comes in many different sizes:
–
–
–
–
•
•
2
Small
Medium
Large
Super-Size Me!
It’s cheap and it doesn’t last long
Nearly any alternative is healthier
Outline
•
I’ll talk about three areas of interest at different scales
– Small: Wireless sensor networks
– Medium: Tiny embedded computers, clustered
– Large: High-performance and cloud computing
•
•
3
And identify some unique challenges in each area
And describe some Aerospace research or investigations in each
one
Large Fries
High-Performance Computing
© The Aerospace Corporation 2011
High-Performance Computing
•
•
•
•
5
Critical domain to work of The Aerospace Corporation
Solving computationally-intensive problems, generally with large data
sets
• Sometimes data sets are from “real” sensor sources:
– Signal processing, image processing
• Sometimes they are synthesized
– e.g., find optimal satellite configuration through exhaustive
search or Monte-Carlo method
…using a variety of different, interconnected computing
architectures, techniques, and resources beyond a single core/serial
implementation.
HPC-style resources are everywhere: laptops, desktops, gaming
consoles, embedded processors…
Major trends in HPC
•
•
•
6
Increasing the speed of individual processing elements
– i.e., more megahertz!
Increasing the “smarts” within processors
– i.e., longer pipelines, branch prediction, speculative execution
Parallelism is the name of the game: dwarfs megahertz and smarts
At the level of…
Through…
Instructions
Multi-issue processors
Data elements
Vector processing, SIMD instructions
Threads
“Hyperthreading”
Processor elements
Multicore processors, Heterogeneous Multicore
Processors
Symmetric Multiprocessing (SMP), NUMA
Machines
Cluster-based computing
Clusters
Grid computing
Increasing Complexity in HPC
•
Put another way…
Chip
Multiissue
SIMD
Inst.
SIMD
Units
SMT
Multicore
VLIW/E MultiPIC
Proc.
486
X
Pentium
X
PII/III
X
X
P4
X
X
Core…
X
X
X
GPUs
Power6
X
X
X
X
X
X
X
X
Itanium
X
Cell
X/X
X
X
X
X
X/_
Not shown: FPGAs, tile-based computing…
7
H.M.P.
X/X
X
X
X
X
X
A Challenge Problem
•
What is the fastest way to transform a large amount of data through
three transformations (“Red”, “Green”, and “Blue”)?
– I’ll give you two options (assume both will result in the same output)
Data
Data
8
A Challenge Problem: Option 1
•
Do all the red, then all the green, then all the blue.
Data
9
A Challenge Problem: Option 2
•
Chunk up the data and do red, then green, then blue on each chunk.
Data
10
Ordinary parallelism is only one side of the story…
•
Moore’s Law: The number of transistors on a chip will double every
~18 months
Credit: User Wgsimon; used under Creative
Commons ShareAlike 3.0 License
11
…but what are we doing with those transistors?
•
Moore’s law: the number of transistors on a chip will double every
~18 months
Multicore VLIW/EPIC
Multicore + SMT
GPUs
VLIW/EPIC
Multicore
Heterogeneous Multiprocessor
SMT, multi-issue
SMT, in-order
Superscalar, multi-issue
12
Key Challenges
•
Parallelism is still hard for people to understand and master
– Some techniques – OpenMP, actor model – ease things
•
•
13
Trying out different parallelization strategies is still expensive and
labor-intensive
Good tooling is hard to get, because the platforms are evolving
faster than the tools can be built
A Little Research
14
Medium Fries
Kinda-Embedded Computing
© The Aerospace Corporation 2011
“Kinda-Embedded” Computing
•
New computing platforms and processors emerging in a niche we
haven’t seen before
– The power, programmability, and connectivity of an (old) desktop
computer
– In extremely small form factors
•
Largely driven by developments in cell phone technology
•
16
Meet Gumstix Overo
• 600Mhz ARM-based processor
w/extensions
• 256MB RAM, 256MB Flash, plus
SD card (up to 8GB)
• 10/100 Ethernet, 54mbps WiFi,
Bluetooth
• USB, HDMI
Applications
•
•
•
•
•
•
•
Cell phones, tablets, PDAs, other handhelds
“Plug Computers”
Car-puters
Walltops
Location-aware applications
Situational Awareness
Micro web-servers
•
Spaceborne(?)
17
Research: Gumstix for HPC-style Processing
Are they viable for future spaceborne applications?
•
•
Worked with a UCSB student
team and a summer intern
Developed two clusters
– One homemade
– One on Gumstix “Stagecoach”
backplane
•
•
18
Ported Range-Doppler SAR
algorithm to Gumstix and
parallelized it
Compared performance and
power usage to a small-formfactor desktop (Mac Mini)
Performance of various implementations
Comparing Gumstix implementation strategies to Mac Mini
SAR Image Processing Time
Comparing Different Compilation Tools with Several
Computers and Platforms
40
37.04
Image Processing Time (seconds)
35
30
24.51
25
20.39
18.32
20
16.79
14.37
15
13.64
6.288
6.47
2
3
17.09
16.29
15.88
13.07
12.93
12.85
6.558
6.656
6.718
6.75
6
7
5
0.422
0
1
19
BitBake FFTW Compilation
FFMPEG Compilation
13.25
10
5.732
Native FFTW Compilation
4
5
Number of Computers
Mac Mini
Energy Use
It’s slower, but how much energy does it use?
Energy Cost of Processing SAR Data Using Several Computers
160.0
142
140.0
120
120.0
Energy Cost (Joules)
105
100.0
88.5
101
Gumstix Energy Cost
(Computation Only)
75.7
80.0
Gumstix Energy Cost
(Including Cluster Hardware)
79.3
62.3
60.0
Mac Mini
65.2
45.9
49.2
40.0
36.9
20.0
24.5
12.6
0.0
11.5
1
2
3
4
Number of Computers
5
6
Gumstix extremely close to Mac Mini in terms of power
consumption
20
7
Key Challenges
•
Optimization can make performance vary widely
– Software engineering gives us few tools to do that optimization
• Biggest performance gains at very low levels of abstraction
•
Likely to see increasing diversity in processors at this scale
– Will what we learn on one apply to another?
•
True mobility requires batteries
– Power-aware computing probably increasingly important
– But the lure of these platforms is how similar they are to platforms where
we can blithely ignore power use.
21
Small Fries
Wireless Sensor Networks
© The Aerospace Corporation 2011
Wireless Sensor Networks
•
A wireless network of physically distributed small computing devices
(colloquially: “motes”) equipped with tiny sensors
•
Additional characteristics:
– Network topology may be fixed, variable, or ad-hoc
• Motes may be mobile, though infrequently
– Devices expected to run for months or years unattended
– Data collected generally forwarded to a central collection point
– Network should survive the loss of numerous motes
•
Applications:
– Environmental monitoring
– Factory/industrial monitoring
– Target detection and tracking
23
Meet Mica
•
MicaZ Mote
– 8Mhz Atmel ATMega128 Microcontroller (MIPS-like assembly
language)
– 4KB of RAM
– 512KB of Flash (usually partitioned into 4x128KB program
blocks)
– CC2420 802.15.4 “Zigbee”-compliant radio
• 7 power usage modes
– Several analog inputs, several digital inputs, I2C
•
Sensor boards stack on top
– Light, temperature, humidity,
orientation (magnetometer),
acceleration, location (GPS),
microphone (levels only)
Crossbow MicaZ Mote
24
How Motes Work
•
Base Station PC
Drop a bunch of motes in an area
GW
•
Distance between them: 50-100 meters
•
One distinguished mote is the “gateway;”
this is connected to an ordinary PC by
a serial connection or Ethernet
transceiver for data collection
25
How Motes Work
•
Drop a bunch of motes in an area
GW
•
Distance between them: 50-100 meters
•
One distinguished mote is the “gateway;”
this is connected to an ordinary PC by
a serial connection or Ethernet
transceiver for data collection
•
Software on the motes allows them
to self-organize, collect data, and
report that data to their ‘parent’
in the network
– Base station ultimately gets all data
26
A sample application we developed
•
27
Signal transmitter tracking application
A sample application we developed
•
28
Cell-phone tracking application
Socket
Device Control
Device Drivers
Timers/Utilities
User App
XServe
receives data
from base
station mote
and makes it
available to
userspace apps
XMesh forms and
maintains ad-hoc
network; also
provides interface
for other TinyOS
components to
send and receive
messages.
XMesh
Web UI
Tables
Map UI
The Software Stacks
TinyOS
XServe
Serial
29
USB
Ethernet
Motes communicate
over 802.15.4 protocol
(used in ZigBee devices)
Long-range,
low data rate.
Mote software based
on open-source
TinyOS.
A thin component
model is built on top of
that.
Software on the Motes
•
•
•
•
30
TinyOS developed concurrently with the Mica family motes
– Mica motes still probably the #1 platform for TinyOS
– Other platforms supported
Amazingly, there is a thin but useful
component model
Components written in nesC (a dialect
of C) with well-defined interfaces
Interfaces connected through a
configuration diagram that resembles
early module interconnection
languages (e.g., Polylith)
configuration Blink {
}
implementation {
components Main, BlinkM, SingleTimer, LedsC;
Main.StdControl -> BlinkM.StdControl;
Main.StdControl -> SingleTimer.StdControl;
BlinkM.Timer -> SingleTimer.Timer;
BlinkM.Leds -> LedsC;
}
…
interface Timer {
command result_t start(char type, uint32_t interval);
command result_t stop();
event result_t fired();
}
Key Challenges
•
•
•
•
•
31
Extreme power management is key
– How can we better integrate higher-level power management
principles into the sensor architecture?
Control flow is low-level and painful
– Can we alleviate this with higher-level models that are “compiled
down” to implementations without sacrificing power?
Are there architectural styles appropriate for these applications?
– How do you build applications that use computation and
aggregation in-the-mesh to reduce data transmission?
How do you build debuggable systems?
How do you transition to next-generation systems?
Conclusion
•
We are faced with yet-another “Cambrian Explosion” of
– Scales
– Platforms
•
Software engineering knowledge and insight lacking in these
domains
– For the smaller domains, abstraction (the key to software engineering) is
the enemy of performance and battery life
•
We are building platforms faster than we can build tools and far
faster than we can build skills
– Domain experts often have little formal training in software engineering
•
Need lightweight “force multipliers” with a very low cost:benefit ratio
All trademarks, service marks, and trade names are the property of their respective owners.
32