Effective Software
Code | Completion | Credits | Range | Language |
---|---|---|---|---|
BE4M36ESW | Z,ZK | 6 | 2P+2C | English |
- Relations:
- It is not possible to register for the course BE4M36ESW if the student is concurrently registered for or has already completed the course B4M36ESW (mutually exclusive courses).
- During a review of study plans, the course B4M36ESW can be substituted for the course BE4M36ESW.
- It is not possible to register for the course BE4M36ESW if the student is concurrently registered for or has previously completed the course B4M36ESW (mutually exclusive courses).
- Course guarantor:
- David Šišlák
- Lecturer:
- Michal Sojka, David Šišlák
- Tutor:
- Marek Cuchý, Joel Matějka, Michal Sojka, David Šišlák, Jiří Vlasák
- Supervisor:
- Department of Computer Science
- Synopsis:
-
Within the course of Efficient software you will get familiar with the area of software and algorithm optimization under limited resources. The course is focused on the efficient usage of modern hardware architectures - multi-core and multi-processor systems with shared memory. Students will practically implmenet and use presented techniques in C and Java. Main topics are: code optimization, effective data structures and processor cache usage, data structures in multi-threaded applications and implementation of efficient network servers.
- Requirements:
-
background knowledge in area of Computer architecture, C and Java programming languages
- Syllabus of lectures:
-
1. Introduction, modern computer architecture, C compilers.
2. Bentley's rules, C compiler, profiling.
3. Benchmarking, measurements, metrics, statistics, WCET, timestamping.
4. Scalable synchronization – from mutexes to RCU (read-copy-update), transactional memory, scalable API.
5. Memory access – efficient programming with caches, dynamic memory allocation (malloc, NUMA, ...)
6. Serialization of data structures – JSON, XML, protobufs, AVRO, cap'n'proto, mmap/shared memory.
7. Program run – virtual machine, byte-code, Java compiler, JIT compiler, relation to machine code, byte-code analysis, dissasembly of Java byte-code, optimization in compilers, program performance analysis, profiling.
8. Data concurrency in JVM – multi-threaded access to data, locks monitoring, atomic operations, lock-less/block-free data structures, non-blocking algorithms (queue, stack, set, dictionary), data races, synchronization.
9. Efficient servers, C10K, non-blocking I/O, efficient networking, threads.
10. JVM – Memory analysis (dynamic/static), data structures, collections for performance.
11. JVM – Object allocation, bloom filters, references, effective caching.
12. Virtualization (IOMMU, SR-IOV, PCI pass-through, virtio, …).
13. Memory Management in JVM – Memory Layout, Garbage Collectors.
- Syllabus of tutorials:
- Study Objective:
- Study materials:
-
[1] MIT: Performance-engineering-of-software-systems
[2] Oaks, S.: Java Performance: 2nd Edition. O'Reilly, USA 2020.
[3] Jones, R., Hosking, A., Moss, E.: The Garbage Collection Handbook - The Art of Automatic Memory Management. CRC Press, USA 2012.
[4] Herlihy, M., Shavit, N.: The Art of Multiprocessor Programming. Morgan Kaufman, 2008.
[5] Fog, A.: The microarchitecture of Intel, AMD and VIA CPU, 2016.
[6] Drepper U.: What every programmer should know about memory, 2007
[7] Jain, R.: The Art of Computer Systems Performance Evaluation. Wiley, New York 1991. (slides, book)
[8] Lilja, D. J.: Measuring Computer Performance: A Practitioner?s Guide. Cambridge University Press, 2000. (book web site, Supplemental Teaching Materials)
- Note:
- Further information:
- https://esw.pages.fel.cvut.cz/
- Time-table for winter semester 2024/2025:
- Time-table is not available yet
- Time-table for summer semester 2024/2025:
-
06:00–08:0008:00–10:0010:00–12:0012:00–14:0014:00–16:0016:00–18:0018:00–20:0020:00–22:0022:00–24:00
Mon Tue Wed Thu Fri - The course is a part of the following study plans:
-
- Open Informatics - Software Engineering (compulsory course of the specialization)
- Open Informatics - Computer Engineering (compulsory course of the specialization)