D.D. Gajski, F. Vahid, S. Narayan, J. Gong. SpecSyn: An Environment Supporting the Specify-Explore-Refine Paradigm for Hardware/Software System Design. IEEE Transactions on VLSI Systems, Vol. 6, No. 1, pp. 84-100, 1998.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....best by reuse. Therefore, our goal is to control partitioning by the availability of IP. We believe that combining automated HW SW partitioning and code reuse bears a large potential for a further cut in product cost and design time. 2. Related Work Current codesign approaches like [1] 2] and [3] focus on interactive partitioning, performance estimation, cosimulation, communication synthesis and code generation. Automated partitioning has been under intense research (e.g. 2] 3] 5] 7] and [8] However, manual partitioning is still predominant in industry with its quality depending ....

....a further cut in product cost and design time. 2. Related Work Current codesign approaches like [1] 2] and [3] focus on interactive partitioning, performance estimation, cosimulation, communication synthesis and code generation. Automated partitioning has been under intense research (e.g. 2] [3], 5] 7] and [8] However, manual partitioning is still predominant in industry with its quality depending largely on the skills and experience of the designer. While code reuse is traditionally widespread in software engineering, the viability of this technique in the field of hardware design ....

Gajski, D., Vahid, F., Narayan, S., Gong, J.: SpecSyn: An Environment Supporting the Specify-Explore-Refine Paradigm for Hardware/ Software System Design, IEEE Trans. on VLSI Systems, Vol. 6, No.1, pp. 84-100, 1998

No context found.

D.D. Gajski, F. Vahid, S. Narayan, J. Gong. SpecSyn: An Environment Supporting the Specify-Explore-Refine Paradigm for Hardware/Software System Design. IEEE Transactions on VLSI Systems, Vol. 6, No. 1, pp. 84-100, 1998.

No context found.

D.D. Gajski, F. Vahid, S. Narayan, J. Gong. SpecSyn: An Environment Supporting the Specify-Explore-Refine Paradigm for Hardware/Software System Design. IEEE Transactions on VLSI Systems, Vol. 6, No. 1, pp. 84-100, 1998.

No context found.

Gajski, D., F. Vahid, S. Narayan and J. Gong. SpecSyn: An Environment Supporting the Specify-Explore-Refine Paradigm for Hardware/Software System Design. IEEE Trans. on VLSI Systems, Vol. 6, No. 1, pp. 84-100, 1998.

No context found.

D.D. Gajski, F. Vahid, S. Narayan, J. Gong. SpecSyn: An Environment Supporting the Specify-Explore-Refine Paradigm for Hardware/Software System Design. IEEE Transactions on VLSI Systems, Vol. 6, No. 1, pp. 84-100, 1998.

....cospecification, configurable logic, system on a chip, platformbased design. 1. INTRODUCTION Hardware software partitioning has been shown to provide excellent performance as well as power and or energy improvements compared to software only implementations in embedded computing systems [4] 7][9][11] 15] 16] 18] 25] 28] 29] Making such partitioning even more attractive is the appearance of single chip platforms, some of which are intended for consumer products, that include both a microprocessor and configurable logic [1] 3] 13] 23] 27] 30] Most approaches to hardware software ....

D.D. Gajski and F. Vahid and S. Narayan and J. Gong. SpecSyn: An Environment Supporting the Specify-ExploreRefine Paradigm for Hardware/Software System Design. IEEE Transactions on VLSI Systems, Vol. 6, No. 1, pp. 84100,

....partitioning [3] In functional partitioning, we take advantage of the recent trend of designers first specifying a system s functionality using a program like language, such as VHDL, Verilog, or C. We partition this program among packages, using SpecSyn estimators developed by UC Irvine [4] to guide the partitioning process. After partitioning, structure may be designed separately for each package. The advantages included greatly reduced I O (often 50 ) improved performance, and reduced synthesis runtimes (often by an order of magnitude) Such advantages were predicted by some ....

....or exlined to adjust the granularity) It is achieved by using any of several heuristics (such as simulated annealing, a modified Kernighan Lin heuristic, clustering or greedy improvement) or by manually moving objects. For further details on the partitioning approach, the reader is referred to [19, 4, 12, 20]. In addition to using the SLIF representation, our approach uses the FunctionBus for inter part communication [18] In contrast, many earlier multipackage functional partitioning approaches used a cut edges approach to I O implementation. In such an approach, a graph s nodes, representing ....

[Article contains additional citation context not shown here]

D. Gajski, F. Vahid, S. Narayan, and J. Gong, "Specsyn: An environment supporting the specify-explorerefine paradigm for hardware/software system design," IEEE Transactions on Very Large Scale Integration Systems, p. to appear, 1997.

....by combining pre estimated values, and is usually done hundreds or thousands of times during application of a partitioning heuristic. It may involve sophisticated data structures that can be incrementally updated as objects are moved. We have described estimation techniques elsewhere [5, 6]. Preclustering Granularity selection N way assignment Partitioned access graph Access graph estimation PreOnline estimation Fig. 2: Sequence of partitioning steps. 3 Partitioning methodology We apply a three step partitioning methodology to obtain an implementation. They occur in sequence as ....

....feasible solutions for each package that would satisfy overall constraints. Multipar [16] and Gebotys techniques [17] partitioned operations among packages simultaneous with scheduling and allocation, using integerlinear programming. Other e#orts partitioned procedurallevel operations. SpecPart [6] partitioned procedures among packages using clustering and iterative improvement. Peng and Kuchcinski [18] similarly partitioned among packages. None of these e#orts applied a threestep methodology as described in this work, although the idea behind multi stage clustering in Aparty is along ....

[Article contains additional citation context not shown here]

D. Gajski, F. Vahid, S. Narayan, and J. Gong, "Specsyn: An environment supporting the specify-explorerefine paradigm for hardware/software system design," IEEE Transactions on VLSI, p. to appear, 1998.

....with the aid of SpecSyn and highlevel synthesis, which is nearly a 10 times reduction from the 6 months required for the manual design. SpecSyn was used in the design of a recent commercial product: a 1 ms scan, 10k step programmable controller [20] Additional SpecSyn information can be found in [21]. 7 Conclusions We described the SpecSyn environment supporting the specify explore refine system design paradigm. Our exploration approach uses pre estimation and online estimation to achieve fast and accurate estimates, supports various partitioning heuristics, and is intended to be ....

D. Gajski, F. Vahid, S. Narayan, and J. Gong, "Specsyn: An environment supporting the specify-explorerefine paradigm for hardware/software system design," IEEE Transactions on VLSI Systems, vol. 6, no. 1, pp. 84--100, 1998.

....maintain a design model, and can be incrementally updated as objects are moved among components. In both estimation types, accuracy must be traded o# with speed. Estimation is an extremely important task and is closely related to partitioning, but we have described our techniques in other papers [6, 7]. 3 Partitioning methodology We apply a three step partitioning methodology to obtain an implementation. They occur in sequence as shown in Figure 2, though iterating over the sequence is certainly desirable. We now describe each step. 3.1 Step 1: granularity selection The goal of this ....

....feasible solutions for each package that would satisfy overall constraints. Multipar [17] and Gebotys techniques [18] partitioned operations among packages simultaneous with scheduling and allocation, using integer linear programming. Other e#orts partitioned procedural level operations. SpecPart [7] partitioned procedures among packages using a suite of heuristics, including clustering and iterative improvement. Peng and Kuchcinski [19] similarly partitioned among packages. None of these e#orts applied a threestep methodology as described in this work, although the idea behind multi stage ....

[Article contains additional citation context not shown here]

D. Gajski, F. Vahid, S. Narayan, and J. Gong, "Specsyn: An environment supporting the specify-explorerefine paradigm for hardware/software system design," IEEE Transactions on Very Large Scale Integration Systems, vol. 6, no. 1, pp. 84--100, 1998.

No context found.

Gajski, D.D; Vahid F.; Narayan S.; Gong J.; "SpecSyn: An environment Supporting the Specify-ExploreRefine Paradigm for Hardware/Software System Design", IEEE Transactions on Very Large Scale Integration Systems, Volume 6, Issue 1, 1998, Page(s): 84100.

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