This powerpoints presention presents our new developed FOCAP tool (Framework for optimizing the Computer Architecture Performance) in order to gain a better understanding and familiarity of the students with new advanced learning methods and tools in the Microarchitecture Simulation and Optimization. At this stage, FOCAP allows a mono-objective automatic design space exploration (DSE) of a superscalar processor by varying several architectural parameters. Such DSE tools are very useful, since it is impossible to simulate all the configurations of a highly parameterized microarchitecture. Therefore, heuristic methods, local search algorithms and advanced machine learning methods are good candidates to find near-optimal configurations by evaluating a reduced number of configurations from the huge design space. Our application falls in the European trend of Framework projects that aim expanding research towards the advanced education field using modern learning technologies. The FOCAP tool facilitates the comprehending of theoretical questions, thus allowing students to feel more confident when studying microarchitecture optimization and DSE-related issues.
1. InfoMatrix 2013
Programming Section
FOCAP Tool
Framework for Optimizing Computer Architecture
Performance
Associate Professor:
Students:
Adrian FLOREA, PhD
Andrei Klein Florin
Badea Victor
Advanced Computer Architecture & Processing Systems Research Lab http://acaps.ulbsibiu.ro/index.php/en/
2. Teaching Tool
- in Computer Architecture -
Understand the importance of each component of the processor
Understand the need for Automatic Design Space Exploration
See the impact on productivity vs. manual exploration
Good start in Research
Combining concepts from different Computer Science domains:
microarchitecture simulation and optimization, networking,
evolutionary computing, programming, reliability, database
Open Source code
http://webspace.ulbsibiu.ro/adrian.florea/html/simulatoare/FOCAP_Tool.rar
3. Optimization Algorithms in
Design Space Exploration
Search and optimization
algorithms
Enumerative
(exhaustive search)
Heuristic
Deterministic
Hill Climbing
Stochastic
Simulating Annealing
Genetic Algorithms
6. Client – Server Architecture
Client
• Send client performance & nr of threads
• Wait for work from the server
• Process work
• Send results back to the server
Server
• Receive client performance & nr of threads
• Send request
• Store connection info for future use
• If request times out, send request elsewhere
11. Some Simulation Results
600000
500000
C 400000
P
U
c
y 300000
c
l
e
s 200000
(b)
600000
100000
500000
0
0
100
200
300
400
500
600
Simulated configurations
(a)
(a) Microarchitecture optimization
using Hill-Climbing
(b) Optimal configuration at each step
C 400000
P
U
c
300000
y
c
l
e
s 200000
100000
0
0
5
10
15
Number of Hill-Climbing iterations
20
25
12. Conclusions
FOCAP is:
very good education tool in Computer Architecture
free availability for use, flexibility, extensibility and portability
Encourages the student to take a research path
Encourages Open Source Projects
Provides a good code implementation for a stable and
reliable asynchronous communication system
Emphasizes the importance of a adequate architecture
13. And that’s not all… Further Work
Extending the performance analysis in order to find
optimum configuration from a multi-criteria point of view
(processing performance, power consumption)
Processor power consumption measurement (integration with CACTI).
Notification System
Bring the UI to the mobile world
Add support for Linux machines