Nagu R. Dhanwada

Architecting voltage islands in core-based system-on-a-chip designs

Voltage islands enable core-level power optimization for System-on-Chip (SoC) designs by utilizing a unique supply voltage for each core. Architecting voltage islands involves island partition creation, voltage level assignment and floorplanning. The task of island partition creation and level assignment have to be done simultaneously in a floorplanning context due to the physical constraints involved in the design process. This leads to a floorplanning problem formulation that is very different from the traditional floorplanning for ASIC-style design. In this paper, we define the problem of architecting voltage islands in core-based designs and present a new algorithm for simultaneous voltage island partitioning, voltage level assignment and physical-level floorplanning. Application of the proposed algorithm to a few benchmark and industrial examples is demonstrated using a prototype tool. Results show power savings of 14%-28%, depending on the constraints imposed on the number of voltage islands and other physical-level parameters.

A power estimation methodology for systemC transaction level models

Majority of existing works on system level power estimation have focused on the processor, while there are very few that address power consumption of peripherals in a SoC. With the presence of complex cores in current day embedded system-on-chip devices, the problem of complete system level power estimation is gaining significance. Transaction level models for SoCs are gaining increasing attention with emerging architectural modeling standards like SystemC. In this paper we present a methodology for performing system power estimation for different scenarios or applications being executed on these transaction level models. We describe techniques and a setup for transaction level power characterization, and an approach to augment SystemC transaction level models to perform transaction level power estimation. We also present experimental results to validate the accuracy and speed of our approach.

Behavioral synthesis of analog systems using two-layered design space exploration

This paper presents a novel approach for synthesis of analog systems from behavioral VHDL-AMS specifications. We implemented this approach in the VASE behavioral-synthesis tool. The synthesis process produces a netlist of electronic components that are selected from a component library and sized such that the overall area is minimized and the rest of the performance constraints such as power, slew-rate, bandwidth, etc. are met. The gap between system level specifications and implementations is bridged using a hierarchically-organized, design-space exploration methodology. Our methodology performs a two-layered synthesis, the first being architecture generation, and the other component synthesis and constraint transformation. For architecture generation we suggest a branch-and-bound algorithm, while component synthesis and constraint transformation use a genetic algorithm based heuristic method. Crucial to the success of our exploration methodology is a fast and accurate performance estimation engine that embeds technology process parameters, SPICE models for basic circuits and performance composition equations. We present a telecommunication application as an example to illustrate our synthesis methodology, and show that constraint-satisfying designs can be synthesized in a short time and with a reduced designer effort.

Exploring power management in multi-core systems

Power dissipation has become a critical design metric in microprocessor-based system design. In a multi-core system, running multiple applications, power and performance can be dynamically traded off using an integrated power management (PM) unit. This PM unit monitors the performance and power of each core and dynamically adjusts the individual voltages and frequencies in order to maximize system performance under a given power budget (usually set by the operating system). This paper presents a performance and power analysis methodology, featuring a simulation model for multi-core systems that can be easily reconfigured for different scenarios and a PM infrastructure for the exploration and analysis of PM algorithms. Two algorithms have been implemented: one for discrete and one for continuous power modes based on non-linear programming. Extensive experiments are reported, illustrating the effect of power management both at the core and the chip level.

A two-layer library-based approach to synthesis of analog systems from VHDL-AMS specifications

This paper presents a synthesis methodology for analog systems described using VHDL-AMS language. Synthesis produces net-lists of analog components that are selected from a library, and sized so that specified objectives (like AC response, signal to noise ratio, dynamic range, area) are optimized. The gap between abstract specifications and implementations is bridged using a two-layered methodology. The first layer is architecture generation. The second layer is component synthesis and constraint transformation. Architecture generation employs the branch-and-bound algorithm to create architectural alternatives for a system. Component synthesis and constraint transformation use a directed interval based genetic algorithm that operates on parameter ranges. The performance estimation engine embeds technology process parameters, SPICE models for basic circuits, and symbolic composition equations for basic structural configurations. The paper discusses the VHDL-AMS subset for synthesis. The subset offers the composition semantics. As a result, specifications offer sufficient insight into the system structure to allow automated architecture generation. To justify the flexibility of the methodology, the paper presents results for three case studies, a signal conditioning system, a filter, and an analog to digital converter. Experiments show that constraint-satisfying designs can be synthesized in a short time, at a low cost, and without requesting broad knowledge on analog circuits.

Hierarchical constraint transformation using directed interval search for analog system synthesis

In this paper, we present a hierarchical approach for constraint transformation. The important features of this are: a genetic algorithm (GA) based search engine that computes design parameter ranges, a hierarchically organized characterization mechanism based on the concept of directed intervals that assists the search engine and an analog performance estimator. Experiments were conducted comparing the hierarchical approach with a flat bottom-up one. The results obtained demonstrate the effectiveness of the former approach. Experimental results highlighting the impact of using the characterization information within the constraint transformation process are also presented.

SEAS: a system for early analysis of SoCs

Systems-on-chip (SoC) continue to be very complex to design and verify, despite extensive component reuse. Although reusable components are predesigned and preverified, when they are assembled in an SoC there is no guarantee that the whole system will behave as expected from a performance, cost and integration point of view. In many cases this is because of faulty early design decisions regarding the architecture, core selection, floorplanning, etc. This paper presents a system for early analysis of SoCs which helps designers make early decisions regarding performance, area, timing and power; and allows them to quickly evaluate cross-domain effects, such as the effect that an architectural decision may have on the performance and chip area.

Transaction-level modeling for architectural and power analysis of PowerPC and CoreConnect-based systems

Transaction-Level models have emerged as an efficient way of modeling systems-on-chip, with acceptable simulation speed and modeling accuracy. Nevertheless, the high complexity of current architectures and bus protocols make it very challenging to develop and verify such models. This paper presents the transaction-level models developed at IBM for PowerPC and CoreConnect-based systems. These models can be simulated in a SystemC environment for functional verification and power estimation. Detailed transaction-based power models were developed. Comparisons between the simulated models and real hardware resulted in errors below 15% in timing accuracy, and below 11% in power estimation compared against gate-level power. These results demonstrate the efficiency of our transaction-level models for early analysis and design space exploration.

Performance modeling for early analysis of multi-core systems

Performance analysis of microprocessors is a critical step in defining the microarchitecture, prior to register-transfer-level (RTL) design. In complex chip multiprocessor systems, including multiple cores, caches and busses, this problem is compounded by complex performance interactions between cores, caches and interconnections, as well as by tight interdependencies between performance, power and physical characteristics of the design (i.e., floorplan). Although there are many point tools for the analysis of performance, or power, or floorplan of complex systems-on-chip (SoCs), there are surprisingly few works on an integrated tool that is capable of analyzing these various system characteristics simultaneously and allow the user to explore different design configurations and their effect on performance, power, size and thermal aspects. This paper describes an integrated tool for early analysis of performance, power, physical and thermal characteristics of multi-core systems. It includes cycle-accurate, transaction-level SystemC-based performance models of POWER processors and system components (i.e., caches, buses). Power models, for power computation, physical models for floorplanning and packaging models for thermal analysis are also included. The tool allows the user to build different systems by selecting components from a library and connecting them together in a visual environment. Using these models, users can simulate and dynamically analyze the performance, power and thermal aspects of multi-core systems.

A genetic approach to simultaneous parameter space exploration and constraint transformation in analog synthesis

In this paper, we present a genetic algorithm based constraint transformation method. A unique feature of this is performing system level parameter space exploration to generate component design parameter ranges. We discuss a theoretical formulation and also show the methods application in pruning the search space of an analog circuit synthesis tool. Finally experimental results showing the impact of parameter space exploration on circuit synthesis are presented.