Article

Article title FUNCTIONAL ORIENTED PROCESSORS – KEY ELEMENTS OF EMBEDDED SUPERCOMPUTERS FOR THE REAL-TIME SYSTEMS
Authors N.A. Lookin
Section SECTION I. PRINCIPLES OF CONSTRUCTION, ARCHITECTURE AND HARDWARE BASE SUPERCOMPUTERS
Month, Year 12, 2014 @en
Index UDC 004.2; 004.38; 621.38
DOI
Abstract There are many various kinds of supercomputers exists in the modern research and development activity. The supercomputers embedded into real-time systems are key components for high performance of these systems. Optimization of the architectures for embedded supercomputers is important for its top level application efficiency. Increase in performance by means of functional-oriented processors (FOP) including in the architectures of embedded supercomputers is a subject of our paper. The results of research and development of the certain kinds of FOP describes in the paper. It may be used for ASIC design. In particular technical parameters and brief description are given for the following FOP"s: table-algorithmic coprocessors for fast computations of mathematical functions with guaranteed accuracy; parallel FOP with SIMD architecture for implementation of the algorithms of air-vehicle correlation-extremal navigation; 2D-arrays containing a large number of bit-level processor elements; there are two types of the architecture of these arrays – 2D-SIMD and 2D-MIMD architectures for real-time image and signal processing.

Download PDF

Keywords Functional-oriented processors; special-purpose parallelism; VLSI; complexity of computations; board control systems; homogenous computer environment.
References 1. Lukin N.A. Osnovy teorii proektirovaniya arkhitektur funktsional'no-orientirovannykh protsessorov dlya sistem real'nogo vremeni [Fundamentals of the theory of design architecture
functionally-oriented processors for real-time systems], Materialy Pyatoy Mezhdunarodnoy nauchnoy molodezhnoy shkoly i Pyatoy Mezhdunarodnoy molodezhnoy nauchno-tekhnicheskoy konferentsii «Vysokoproizvoditel'nye vychislitel'nye sistemy», 31 avgusta –
6 sentyabrya 2008, Taganrog [The proceedings of the Fifth International youth scientific school and the Fifth International youth scientific and technical conference "High performance computing", 31 August - 6 September 2008, Taganrog]. Taganrog: Izd-vo TTI YuFU, 2008, pp. 115-166.
2. Cui J., Bordim J.L., Nakano K., Hayashi T., Ishii N. Multithreaded Parallel Computer Model with Performance Evaluation, Workshop on Advances in Parallel and Distributed Computational Models. IPDPS 2000 Workshop.
3. Cristina Nicolescu and Pieter Jonker. Parallel low-level image processing on a distributed-memory system, Parallel and Distributed Computing in Image Processing, Video Processing, and Multimedia. IPDPS 2000 Workshop.
4. Available at: http://en.wikipedia.org/iki/Embedded_Supercomputing.
5. Reddi V. Supercomputer Performance on a Chip Powers Next-Generation Embedded Image Processing. Available at: http://rtcmagazine.com/articles/view/102184.
6. Vodyakho A.I., Smolov V.B., Plyusnin V.U., Puzankov D.V. Funktsional'no orientirovannye protsessory [Functionally oriented processors]. Leningrad: Mashinostroenie. Leningr. otd-nie, 1988, 244 p.
7. Lupanov O.B. Kurs lektsiy po diskretnoy matematike [Lectures on discrete mathematics]. Moscow: Izd-vo MGU, 2006, 38 p.
8. Kolmogorov A.N. Teoriya informatsii i teoriya algoritmov [Information theory and theory of algorithms]. Moscow: Nauka, 1987, 305 p.
9. Karatsuba E.A. Bystrye algoritmy i metod BE [Fast algorithms and the method of BWE]. Available at: http://www.ccas.ru/personal/ karatsuba/alg.htm.
10. Popov B.A. Vychislenie funktsiy na EVM [Calculation functions on the computer]. Kiev: Naukova Dumka, 1984, 600 p.
11. Lukin N.A. Arkhitekturnyy sintez funktsional'no-orientirovannykh protsessorov ma-tematicheskikh funktsiy [Architectural synthesis of functionally-oriented processors mathematical functions],
Giroskopiya i navigatsiya [Gyroscopy and navigation], 2003, No. 3, pp. 109-120.
12. Lukin N.A. Parallel'nye algoritmy vychisleniya matematicheskikh funktsiy i ikh kolichestvennye kharakteristiki [Parallel algorithms for computing mathematical functions and their quantitative characteristics], Novye informatsionnye tekhnologii v issledovanii diskretnykh struktur. Doklady II Vserossiyskoy konferentsii [New information technologies in the study of discrete structures. The reports of the II all-Russian conference]. Ekaterinburg: UrO
RAN, 1998, pp. 157-172.
13. Lukin N.A. Sistemnoe proektirovanie funktsional'no-orientirovannykh protsessorov dlya bortovykh korrelyatsionno-ekstremal'nykh navigatsionnykh sistem [System design functionally-oriented processors for on-Board correlation-extremal navigation systems], Vestnik Samarskogo gosudarstvennogo aerokosmicheskogo universiteta im. akademika S.P. Koroleva
[Vestnik of Samara state aerospace University. academician S.P. Korolev], 2009, No. 4 (20), pp. 218-236.
14. Voevodin V.V. Matematicheskie modeli i metody v parallel'nykh protsessakh [Mathematical models and methods in parallel processes]. Moscow: Nauka, 1986, 296 p.
15. Lukin N.A. Funktsional'no-orientirovannye protsessory s odnorodnoy arkhitekturoy dlya realizatsii algoritmov bortovykh sistem upravleniya [Functionally-oriented processors with
homogeneous architecture for implementing algorithms on-Board control systems], Vysokoproizvoditel'nye vychislitel'nye sistemy. Parallel'nye vychisleniya i zadachi upravleniya PACO’2010: Trudy V Mezhdunarodnoy konferentsii, g. Moskva, 26–28 oktyabrya 2010 g
[High-performance computing systems. Parallel computing and control problems PACO’2010: Proceedings of the V International conference, Moscow, 26-28 October 2010]. Institut problem upravleniya im. V.A. Trapeznikova RAN, pp. 1177-1185.
16. Bogachev M.P. Arkhitektura vychislitel'noy sistemy s odnorodnoy strukturoy. V kn. Odnorodnye vychislitel'nye sredy [Architecture of computing systems with a homogeneous structure. In the book of Homogeneous computing environment]. L'vov: FMI AN USSR. 1981.
17. Shmoylov V.I. Pul'siruyushchie informatsionnye reshetki i superkomp'yutery klassa A [Pulsating information grids and supercomputers class A]. L'vov: Merkator, 2005.
18. Available at: http://www.minitera.ru/rus/win/pbm_proc_arh.html.

Comments are closed.