IP Library Granted Patent US 7,246,217
Granted Patent B1
US 7,246,217 · App. 11/110,344 · Granted Jul 17, 2007

Interconnection arrangement of routers of processor boards in array of cabinets supporting secure physical partition

View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 7,246,217
App. No.
11/110,344
Granted
Jul 17, 2007
Kind
B1
Abstract

A multiple processor computing apparatus includes a physical interconnect structure that is flexibly configurable to support selective segregation of classified and unclassified users. The physical interconnect structure includes routers in service or compute processor boards distributed in an array of cabinets connected in series on each board and to respective routers in neighboring row cabinet boards with the routers in series connection coupled to routers in series connection in respective neighboring column cabinet boards. The array can include disconnect cabinets or respective routers in all boards in each cabinet connected in a toroid. The computing apparatus can include an emulator which permits applications from the same job to be launched on processors that use different operating systems.

Claims (31)

1. A computing apparatus, comprising:

a plurality of compute boards, each said compute board having mounted thereon a plurality of compute processors and a plurality of routers coupled to said plurality of compute processors;

a plurality of service boards, each said service board having mounted thereon a plurality of service processors and a plurality of routers, some of said routers coupled to said plurality of service processors;

said compute boards and said service boards each including a same number of said routers;

a plurality of groups of said compute boards respectively provided in a plurality of compute cabinets, and a plurality of groups of said service boards respectively provided, without any of said compute processors, in a plurality of service cabinets;

said compute cabinets and said service cabinets arranged in rows and columns of a generally rectangular array of cabinets;

within each said row of cabinets, each said router of each of a group of said cabinets in said row connected to a respectively corresponding router on a respectively corresponding board in each of two nearest neighbor cabinets in said row;

within each of said compute boards and each of said service boards, said routers thereof connected in a series configuration with one another;

within each of said columns, each said series configuration of routers of each of a group of said cabinets in said column connected in series with a respectively corresponding series configuration of routers on a respectively corresponding board in each of two nearest neighbor cabinets in said column; and

disconnect cabinets provided in said rectangular array of cabinets, each of said disconnect cabinets interposed between respective pairs of said nearest neighbor cabinets and including a switch structure for selectively making and breaking said connections between said routers of the associated nearest neighbor cabinets, said disconnect cabinets positioned relative to said compute cabinets and said service cabinets in said rectangular array such that, when said switch structures of a first group of said disconnect cabinets break the associated connections, some of said compute cabinets are disconnected from a first group of said service cabinets while remaining connected to a second group of said service cabinets, and said disconnect cabinets positioned such that when said switch structures of a second group of said disconnect cabinets break the associated connections, some of said compute cabinets are disconnected from said second group of service cabinets while remaining connected to said first group of service cabinets.

2. The apparatus of claim 1 , wherein each of said compute processors includes a lightweight kernel operating system tailored to support types of application code that run on said compute processors, each said compute processor also including a first interface for interfacing between said lightweight kernel operating system and a launch program that launches applications on said compute processor, each of said service processors including a general purpose operating system and a second interface for interfacing between said general purpose operating system and a launch program that launches applications on said service processor, said second interface emulating said first interface.

3. The apparatus of claim 1 , wherein, within each of said service cabinets and said compute cabinets, each said router of each said board is connected in a torus configuration with corresponding routers on the remaining boards.

4. The apparatus of claim 1 , wherein all of said compute cabinets are disconnected from said first group of service cabinets and remain connected to said second group of service cabinets when said first group of disconnect cabinets break the associated connections, and all of said compute cabinets are disconnected from said second group of service cabinets and remain connected to said first group of service cabinets when said second group of disconnect cabinets break the associated connections.

5. The apparatus of claim 1 , wherein a number of compute cabinets connected to said second group of service cabinets when said first group or disconnect cabinets break the associated connection is greater than a number of compute cabinets connected to said first group of service cabinets when said second group of disconnect cabinets break the associated connections.

6. The apparatus of claim 5 , wherein said disconnect cabinets are positioned to define a plurality of said columns within said rectangular array of cabinets.

7. The apparatus of claim 1 , wherein said disconnect cabinets are positioned to define a plurality of said columns within said rectangular array of cabinets.

8. The apparatus of claim 7 , wherein one of said columns of said disconnect cabinets is interposed between at least one column of said rectangular array that consists only of service cabinets and the remainder of said cabinets, and wherein another of said columns of disconnect cabinets is interposed between at least one further column in said array of cabinets that consists only of service cabinets and the remainder of said cabinets.

9. The apparatus of claim 1 , including user interfaces coupled to at least some of said service boards within said first and second groups of service cabinets.

10. The apparatus of claim 9 , including disk storage facilities coupled to at least some of said service boards within said first and second groups of service cabinets.

11. The apparatus of claim 1 , including disk storage facilities coupled to at least some of said service boards within said first and second groups of service cabinets.

12. A computing apparatus, comprising:

a plurality of compute boards, each said compute board having mounted thereon a plurality of compute processors and a plurality of routers coupled to said plurality of compute processors;

a plurality of service boards, each said service board having mounted thereon a plurality of service processors and a plurality of routers, some of said routers coupled to said plurality of service processors;

said compute boards and said service boards each including a same number of said routers;

a plurality of groups of said compute boards respectively provided in a plurality of compute cabinets, and a plurality of groups of said service boards respectively provided in a plurality of service cabinets;

said compute cabinets and said service cabinets arranged in rows and columns of a generally rectangular array of cabinets;

within each said row of cabinets, each said router of each of a group of said cabinets in said row connected to a respectively corresponding router on a respectively corresponding board in each of two nearest neighbor cabinets in said row;

within each of said compute boards and each of said service boards, said routers thereof connected in a series configuration with one another;

within each of said columns, each said series configuration of routers of each of a group of said cabinets in said column connected in series with a respectively corresponding series configuration of routers on a respectively corresponding board in each of two nearest neighbor cabinets in said column; and

within each of said service cabinets and said compute cabinets, each said router of each said board connected in a torus configuration with corresponding routers on the remaining boards.

13. The apparatus of claim 12 , wherein each of said compute processors includes a lightweight kernel operating system tailored to support types of application code that run on said compute processors, each said compute processor also including a first interface for interfacing between said lightweight kernel operating system and a launch program that launches applications on said compute processor, each said service processor including a general purpose operating system and a second interface for interfacing between said general purpose operating system and a launch program that launches applications on said service processor, said second interface emulating said first interface.

Assignments (3)
CHANGE OF NAME Recorded Sep 6, 2017
From: SANDIA CORPORATION
To: NATIONAL TECHNOLOGY & ENGINEERING SOLUTIONS OF SANDIA, LLC
Reel/Frame 043767/0054 →
CONFIRMATORY LICENSE Recorded Mar 29, 2007
From: SANDIA CORPORATION
To: ENERGY, U.S. DEPARTMENT OF
Reel/Frame 019093/0980 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded May 3, 2005
From: TOMKINS, JAMES L.; CAMP, WILLIAM J.
To: SANDIA CORPORATION, OPERATOR OF SANDIA NATIONAL LABORATORIES
Reel/Frame 016188/0336 →