Minutes of Weekly Meeting, 2007-10-12

NOTE: Starting Monday, 10/29/2007 the meeting is being moved to Mondays at 8am ET.
Join a live meeting during ITC on Thursday, October, 25, 2007 from 10:30 – 12:30. (No phone provided)

Attendees:

Heiko Ehrenberg
Peter Horwood
Adam Ley
Bradford Van Treuren
Al Holliday
Scribes: Adam Ley, Heiko Ehrenberg, Bradford Van Treuren

Action Items:

  • Can ATCA backplane support a full star routing? (Al and Brad)
  • Where can a JTAG Switch Module (JSM) be placed within the system? (Al and Brad)
  • Do we need JTAG redundancy for ATCA since it was not implemented for MicroTCA? (Al and Brad)
  • Need a better overall understanding of ATCA architecture (Al will present an overview of ATCA presentation at next week’s meeting)
  • Adam proposed we cover the following at the next meeting:
    • Establish consensus on goals and constraints
    • What are we trying to achieve?
    • What restrictions are we faced with?
  • Establish whether TRST needs to be addressed as requirements in the ATCA specification if it is not going to be managed globally (All)
  • Distribute Gunnar Carlsson’s Use Case white paper for review and feedback (Brad)
  • Review original presentation sent out by Brad for these meetings – expecially the restrictions/constraints on slide 5 (All)
  • Distribute more slides focusing on goals and constraints. (Brad)

Issues to be discussed and considered in order to select a suitable architecture:

  • signal integrity
  • performance
  • protocol isolation / multi-vendor interoperability
  • signal voltage levels (microTCA: 3.3V; Patrick: we may need lower voltages, I/O level flexibility)
  • real estate available for JSM implementation and signal routing
  • ...

Corrections to 10/5/2007 meeting minutes:

  • The correct spelling of Adam’s name is Adam Ley NOT Adam Le.
  • The sentence stated as, “There was a discussion regarding the current gateway selection protocols and that all of the current protocols have patents behind them. It is a matter of whether the patents are going to be litigated or not.” In the middle of page 3 is incorrect. It should read:
    There was a discussion regarding whether the current gateway selection protocols would be useable in ATCA. Some of the protocols have known patents for them. The SCAN BRIDGE protocol is thought to be open since the protocol was presented in a paper at ITC in 1991 prior to the release of any patent. This openness is why Firecron and Lattice Semiconductor developed their own gateway devices using this protocol according to Peter Horwood. The group identified this topic as needing further investigation.
    (ITC Paper Reference: Dilip Bhavsar (Digital Equipment Corporation), An Architecture for Extending the IEEE Standard 1149.1 Test Access Port to System Backplanes, International Test Conference Proceedings, Paper 28.2, 1991.)

Al Holliday presented an overview of the ATCA system and answered questions from the group.

  • ATCA has been growing since 2002.
  • Al is the Alcatel-Lucent representative for the ATCA 300 family which is specialized for optical systems.
  • The overall makeup of blades in an ATCA system can be categorized as: Single board computers, Fabric boards, and Carrier/Mezzanine specialization boards housing DSPs, CPUs, NPUs, and FPGA filters.
  • 16 slots are possible for 600mm cabinets.
  • 14 slots are possible in 19 inch cabinets.
  • The interconnect signals in Zone 2 could be wired as a Star, Dual-Star, or full Mesh design.
  • The Zone 2 signals are implemented as lanes of SERDES connections that may be used in parallel to boost bandwidth between the slots.
  • There are 4 reserved pins on Zone 1 that are being proposed as the new JTAG interface pins.
  • Some people have implemented ad hoc Multi-drop JTAG using the Metallic Test Bus and Ring Test Bus signals from Zone 1 since the backplane already routes these signal as bussed interfaces to all slots.
  • During initial power up of a blade in a slot, only 10W max is to be consumed by the Board Management Controller (BMC) to begin the handshake requirements over the IPMI interface. The ShMC is responsible for informing each blade when they may power up to provide staged power conditioning to reduce the cost of the power supply required by the system.
  • The Advanced Mezzanine Cards (AMCs) are designed to be hot swappable when installed in ATCA Carrier blades or in a MicroTCA chassis.
  • Brad noted that he and Adam are presenting a paper on JTAG in MicroTCA as an ITC Lecture Series this year.
  • Adam asked, "If embedded boundary-scan is to be used, is the Shelf Management Controller the target location for it?"
  • Al explained how the Shelf Management Controller (ShMC) could be used to support the redundant JTAG switch under certain conditions.
  • Al said the ShMC is one but limited in real estate. The issue is the connector size required. The ShMC is a likely target if redundancy is required.
  • Al also explained that many people have added a third board by the ShMCs in the system to deal with alarm and ad hoc JTAG in the past. This is the way one of the Lucent systems implemented it.
  • Peter Horwood asked if the OpenIPMI capability was used for ATCA. Al replied that it is being used for ATCA.
  • Brad noted that the original specification for ATCA Carrier blades stated to use a serial architecture and required the automatic bridging of TDI to TDO for unpopulated slots. Brad commented that this shows how little the ATCA team knows about the issues regarding JTAG testing in a system.
  • The issue of TRST was discussed further. Peter emphasized that the module should handle TRST.
  • Peter also said, "Don’t mandate JTAG protocol on each board… Many boards have few to no Boundary-Scan devices so why require an interface on each board."
  • Peter further emphasized that the JTAG Switch Module concept could deal with signal level differences as needed.
  • Brad noted that many people successfully used multi-drop architectures effectively in proprietary shelf-oriented equipment where control over standardization on a single gateway protocol can be achieved. Cross protocol issues for open standards and chassis may make a multi-drop protocol difficult to realize.
  • Brad asked all to review Gunnar’s work on STAPL++ as posted on the SJTAG web site (http://www.dft.co.uk/SJTAG/).

Respectfully submitted,

Bradford Van Treuren
(Thanks to Adam and Heiko for additional note taking support)