Analog/mixed-signal IP design flow for SoC applications

System-on-chip (SoC) with reuse of intellectual property (IP) is gaining acceptance as the preferred style for integrated circuit (IC) designs. Increasing demand for analog/mixed-signal (AMS) cores on SoCs is creating a need for new design methodologies and tools that facilitate the creation and...

Full description

Bibliographic Details
Main Author: Hamour, Marwa Mukhtar
Language:English
Published: 2009
Online Access:http://hdl.handle.net/2429/15319
id ndltd-LACETR-oai-collectionscanada.gc.ca-BVAU.2429-15319
record_format oai_dc
spelling ndltd-LACETR-oai-collectionscanada.gc.ca-BVAU.2429-153192014-03-14T15:48:17Z Analog/mixed-signal IP design flow for SoC applications Hamour, Marwa Mukhtar System-on-chip (SoC) with reuse of intellectual property (IP) is gaining acceptance as the preferred style for integrated circuit (IC) designs. Increasing demand for analog/mixed-signal (AMS) cores on SoCs is creating a need for new design methodologies and tools that facilitate the creation and integration of reusable AMS IP. In this work, a practical definition of AMS IP and an associated design-for-reuse process is proposed. This thesis begins by investigating the issues of reusable AMS IP using a phase-locked loop (PLL) as the design vehicle. In the first pass, a PLL is designed without reusable IP. In the process of designing this PLL, the issues of how the IP should be defined for improved reusability and ease of design is addressed. Then a suitable design methodology is proposed to use the IP in a design flow. Firm IP is suggested as the most appropriate format to deliver the AMS IP library components. Unlike hard IP, this form allows ease of migration of IP from foundry-to-foundry, customer-tocustomer and application-to-application. Then, a design methodology is developed for constraintdriven top-down design and model-driven bottom-up characterization of the cores. Once the design topology is selected, an IP hardening flow is described to produce a physical layout. To validate the approach, the primary specifications of the first PLL were then changed and a second architecture was generated using the proposed design methodology. The time needed to design the second PLL is shown to be greatly reduced (by a factor of four) for comparable performance. 2009-11-20T01:17:58Z 2009-11-20T01:17:58Z 2003 2009-11-20T01:17:58Z 2004-05 Electronic Thesis or Dissertation http://hdl.handle.net/2429/15319 eng UBC Retrospective Theses Digitization Project [http://www.library.ubc.ca/archives/retro_theses/]
collection NDLTD
language English
sources NDLTD
description System-on-chip (SoC) with reuse of intellectual property (IP) is gaining acceptance as the preferred style for integrated circuit (IC) designs. Increasing demand for analog/mixed-signal (AMS) cores on SoCs is creating a need for new design methodologies and tools that facilitate the creation and integration of reusable AMS IP. In this work, a practical definition of AMS IP and an associated design-for-reuse process is proposed. This thesis begins by investigating the issues of reusable AMS IP using a phase-locked loop (PLL) as the design vehicle. In the first pass, a PLL is designed without reusable IP. In the process of designing this PLL, the issues of how the IP should be defined for improved reusability and ease of design is addressed. Then a suitable design methodology is proposed to use the IP in a design flow. Firm IP is suggested as the most appropriate format to deliver the AMS IP library components. Unlike hard IP, this form allows ease of migration of IP from foundry-to-foundry, customer-tocustomer and application-to-application. Then, a design methodology is developed for constraintdriven top-down design and model-driven bottom-up characterization of the cores. Once the design topology is selected, an IP hardening flow is described to produce a physical layout. To validate the approach, the primary specifications of the first PLL were then changed and a second architecture was generated using the proposed design methodology. The time needed to design the second PLL is shown to be greatly reduced (by a factor of four) for comparable performance.
author Hamour, Marwa Mukhtar
spellingShingle Hamour, Marwa Mukhtar
Analog/mixed-signal IP design flow for SoC applications
author_facet Hamour, Marwa Mukhtar
author_sort Hamour, Marwa Mukhtar
title Analog/mixed-signal IP design flow for SoC applications
title_short Analog/mixed-signal IP design flow for SoC applications
title_full Analog/mixed-signal IP design flow for SoC applications
title_fullStr Analog/mixed-signal IP design flow for SoC applications
title_full_unstemmed Analog/mixed-signal IP design flow for SoC applications
title_sort analog/mixed-signal ip design flow for soc applications
publishDate 2009
url http://hdl.handle.net/2429/15319
work_keys_str_mv AT hamourmarwamukhtar analogmixedsignalipdesignflowforsocapplications
_version_ 1716653262946435072