Journal article
MULTI-PHASE ROTARY CLOCK SYNCHRONIZATION OF LEVEL-SENSITIVE CIRCUITS
Journal of circuits, systems, and computers, v 18(5), pp 899-908
01 Aug 2009
Abstract
Resonant clocking technologies provide clock networks with improved frequency, jitter and power dissipation characteristics, however, often require novel automation routines. Resonant rotary clocking technology, for instance, entails multi-phase and nonzero clock skew operation and supports latch-based design. This paper studies the effects of multi-phase synchronization schemes on the minimum clock period for rotary-clock-synchronized circuits, which necessitate the application of clock skew scheduling and employ level-sensitive registers. In experimentation, single, dual, three- nd four-phase clocking schemes generated by rotary clock synchronization are applied to a suite of level-sensitive-transformed ISCAS'89 benchmarks. Average clock period improvements of 30.3%, 24.8%, 17.7% and 12.0%, respectively, are observed on average compared to the flip-flop based, zero clock skew circuits. As the number of clock phases increases, smaller improvements are observed due to lesser overall effectiveness of the complementary effects of clock skew scheduling and time borrowing. It is shown, however, that for some circuits (23% of the benchmarks), multi-phase synchronization leads to significant performance benefits in operating frequency.
Metrics
Details
- Title
- MULTI-PHASE ROTARY CLOCK SYNCHRONIZATION OF LEVEL-SENSITIVE CIRCUITS
- Creators
- Baris Taskin - Drexel UniversityIvan Kourtev - University of Pittsburgh
- Publication Details
- Journal of circuits, systems, and computers, v 18(5), pp 899-908
- Publisher
- World Scientific
- Number of pages
- 10
- Resource Type
- Journal article
- Language
- English
- Academic Unit
- Electrical and Computer Engineering
- Web of Science ID
- WOS:000269448400003
- Scopus ID
- 2-s2.0-68249138197
- Other Identifier
- 991019168962204721
InCites Highlights
Data related to this publication, from InCites Benchmarking & Analytics tool:
- Web of Science research areas
- Computer Science, Hardware & Architecture
- Engineering, Electrical & Electronic