GRUN, P., BALASA, F., AND DUTT, N. 1998. Memory size estimation for multimedia applications. In Proceedings of the Sixth International Workshop on Hardware/Software Codesign (CODES/CASHE '98, Seattle, WA, Mar. 15--18), G. Borriello, A. A. Jerraya, and L. Lavagno, Chairs. IEEE Computer Society Press, Los Alamitos, CA, 145--149.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....dataflow (SDF) programs. A SDF program models the data (memory) access explicitly using arcs between the computational actors and hence the memory minimization problem is well formulated. Second, a number of research groups, for example, the IMEC Acropolis group [2] and Irvine ACES group [6], have studied the memory minimization problem for regular data structure. In their study, efficient analysis techniques are used to optimize programs, typically those in multimedia domain, which operate on multi dimensional arrays. The major difference between our work and the abovementioned ....

P. Grun, F. Balasa, and N. Dutt. Memory size estimation for multimedia applications. In Workshop on Hardware /Software Codesign, Seattle, March 1998.

....for each array. To generate the production time axis, the execution ordering has to be fully fixed. Also, since each array is treated separately, only in place mapping internally to an array is considered, not the possibility of mapping arrays in place of each other. Another approach is taken in [5]. The data dependency relations between the array references in the code are used to find the number of array elements produced or consumed by each assignment. From this, a memory trace of upper and lower bounding rectangles as a function of time is found. The total storage requirement equals the ....

....alive basic sets gives the storage requirement. L.1 (j: 0 . 3) k: 0 . 2) begin I.1 B[0] j] k] f( A[j] k] end; L.2 (i: 1 . 5) j: 0 . 3) k: 0 . 2) begin I.2 B[i] j] k] g( B[i 1] j] k] I.3 C[i] j] k] if (j = 2) h( B[i] j 2] k] fi; I. 4 D[j] k] if (i = 5) l( B[5][j] k] fi; Figure 1: Code example (Silage language) Since the rest of this paper in part is based on this work, we now present a simple illustrative example in the single assignment language Silage. Note that very small loop bounds are used here to allow a graphical representation, but the ....

Grun, P., Balasa, F., and Dutt, N., "Memory Size Estimation for Multimedia Applications", Proc. Sixth Int. Workshop on Hardware/ Software Codesign (CODES/CACHE), March 1998, pp. 145-9

....consumption time for any two depending instances, giving the storage requirement for one array. The total storage requirement is the sum of the requirements for each array. Only in place mapping internally to an array is considered, not the possibility of mapping arrays in place of each other. In [8] the data dependency relations between the array references in the code are used to find the number of array elements produced or consumed by each assignment. From this, a memory trace of upper and lower bounding rectangles as a function of time is found with the peak bounding rectangle yielding ....

P. Grun, F. Balasa, and N. Dutt, "Memory Size Estimation for Multimedia Applications", Proc. Sixth Int. Workshop on HW/SW Codesign (CODES/CACHE), March 1998, pp. 145-9

No context found.

GRUN, P., BALASA, F., AND DUTT, N. 1998. Memory size estimation for multimedia applications. In Proceedings of the Sixth International Workshop on Hardware/Software Codesign (CODES/CASHE '98, Seattle, WA, Mar. 15--18), G. Borriello, A. A. Jerraya, and L. Lavagno, Chairs. IEEE Computer Society Press, Los Alamitos, CA, 145--149.

....these two traditionally distinct compiler phases. Furthermore, there is an urgent need for estimation tools that can predict the impact of parallelism on the amount of memory size at the system level to help guide these coarse and finegrain compiler transformations (e.g. the MEMOREX project [38]) Memory aware Instruction Level Parallelization (ILP) Techniques. Last, but not the least, more aggressive memory aware ILP techniques are further required to optimize both the memory requirements, as well as to compile efficiently (e.g. for low power) given a specific memory architecture and ....

P. Grun, F. Balasa, N. Dutt. Memory Size Estimation for Multimedia Applications. In Proc. CODES/CASHE, 1998.

....descriptions for a range of architectures including VLIW (TI C6X [20] DSP (Motorola 56K [15] and Superscalar (Intel Pentium [11] have been implemented. EXPRESSION is currently used to drive several projects at U.C. Irvine, including MEMOREX (a memory estimation and exploration environment) [5], 16] and EXPRESS (an optimizing, retargetable compiler and exploration environment for Embedded Systemson Chip) Our on going work targets strengthening the coupling between the compiler and the architecture, by inserting more architecture dependent optimizations and providing an architecture ....

P. Grun, F. Balasa, and N. Dutt. Memory size estimation for multimedia applications. In Proc. CODES/CACHE '98.

....descriptions for a range of architectures including VLIW (TI C6X [17] DSP (Motorola 56K [15] and Superscalar (Intel Pentium [13] have been implemented. EXPRESSION is currently used to drive several projects at U.C. Irvine, including MEMOREX (a memory estimation and exploration environment)[9], 16] and EXPRESS (an optimizing, retargetable compiler and exploration environment for Embedded Systems on Chip) Our on going work targets strengthening the coupling between the compiler and the architecture, by inserting more architecture dependent optimizations and providing an architecture ....

P. Grun, F. Balasa, and N. Dutt. Memory size estimation for multimedia applications. In Proc. CODES/CACHE '98.

....toolkit generation. EXPRESSION descriptions for a range of architectures including VLIW (TI C6X [17] and DSP (Motorola 56K [14] have been implemented. EXPRESSION is currently used to drive several projects at U.C. Irvine, including MEMOREX (a memory estimation and exploration environment)[9], 15] and EXPRESS (an optimizing, retargetable compiler and exploration environment for Embedded Systems on Chip) Our on going work targets strengthening the coupling between the compiler and the architecture, by inserting more architecture dependent optimizations and providing an architecture ....

P. Grun, F. Balasa, and N. Dutt. Memory size estimation for multimedia applications. In Proc. CODES/CACHE, 1998.

No context found.

Grun, P., Balasa, F., and Dutt, N., "Memory size estimation for multimedia applications", Proc. Sixth Int. Workshop on Hardware /Software Codesign (CODES/CACHE), March 1998, pp.145-9

No context found.

P. Grun, F. Balasa, and N. Dutt, "Memory Size Estimation for Multimedia Applications", Proc. Sixth Int. Workshop on HW/SW Codesign (CODES/CACHE), March 1998, pp. 145-9

No context found.

Grun, P., Balasa, F., and Dutt, N., "Memory size estimation for multimedia applications", Proc. Sixth Int. Workshop on Hardware /Software Codesign (CODES/CACHE), March 1998, pp.145-9

No context found.

Grun, P.; Balasa, F.; Dutt, N., "Memory size estimation for multimedia applications", in Proc. 6th International Workshop on Hardware/Software Codesign, 1998, Page(s): 145-149.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC