Edward S Suffern

6505298, Jan 7, 2003

Oct 25, 1999

09/426434

Christopher D. Cerbini - Raleigh NC
Scott N. Dunham - Cary NC
Benjamin R. Grimes - Zebulon NC
Edward S. Suffern - Chapel Hill NC

International Business Machines Corporation - Armonk NY

G06F 900

713 1, 713 2, 713100, 709220, 709221, 709222

The present invention provides a method and system for providing a reset after an operating system (OS) hang condition in a computer system, the computer system including an interrupt handler not accessible by the OS. The method includes determining if an interrupt has been generated by a watchdog timer; monitoring for an OS hang condition by the interrupt handler if the interrupt has been generated and after it is known that the OS is operating; and resetting the OS if a device driver within the OS has not set a bit in a register, the bit for indicating that the OS is operating. The method and system in accordance with the present invention uses existing hardware and software within a computer system to reset the OS. The present invention uses a method by which a critical hardware watchdog periodically wakes a critical interrupt handler of the computer system. The critical interrupt handler determines if the OS is in a hang condition by polling a share hardware register that a device driver, running under the OS, will set periodically.

6529945, Mar 4, 2003

Jul 26, 1999

09/361079

George Michael Calhoun - Raleigh NC
David Charles Sips - Efland NC
Edward Stanley Suffern - Chapel Hill NC

International Business Machines Corporation - Armonk NY

G06F 15167

709213, 710 38, 710 40, 710 56, 36518533, 711173

A data buffer management system and method between two different types of systems, The data buffer management system employs circular buffer chaining, wherein multiple state machines service buffer updates from multiple sources, dual buffer structures are maintained, mapping is provided between two different control block structures, and full duplex operation is supported. The data buffer management system manages use of a data buffer memory. A transmit operation is performed by transferring and storing data from one of the two different systems to the data buffer memory The data in the data buffer memory is converted and read by another of the two different systems. A receive operation is performed by transferring and storing data from the other of the two different systems to the data buffer memory.

6647028, Nov 11, 2003

Sep 21, 1998

09/157743

Emmanuel Lancon - Nice, FR
Clement Yvon Gustave Poiraud - Cagnes sur Mer, FR
Edward Stanley Suffern - Chapel Hill NC
Spiros Joseph Teleoglou - Raleigh NC

Cisco Technology, Inc. - San Jose CA

H04J 304

370535, 370542, 370519, 370216

A method and apparatus for use where an incoming high speed data stream is first divided into N lower bandwidth channels and is then transmitted through the network to a receiver where it arrives with varying delays depending on the physical characteristics of the network. The high speed data stream is reconstructed by determining the different geographical delay parameters among the lower speed channels, by adjusting the transmission delays by alignment circuitry, and then by combining the lower speed signals into one high bandwidth channel. When one of the lower speed channels fails, the bandwidth is automatically re-adjusted using the remaining N-1 lines. When the failed channel is restored to an acceptable level of service, the communication protocol automatically re-adjusts the bandwidth to the full N channels. When the network indicates the loss of a line, the multiplexers at both ends of the link automatically compensate for the loss without any message exchange. When the line is restored, the multiplexers at both ends of the link automatically re-adjust the bandwidth.

6735635, May 11, 2004

May 18, 2000

09/574753

James William Feeney - Endicott NY
Jorge R. Rodriquez - Cary NC
Edward Stanley Suffern - Chapel Hill NC
Robert William Bartoldus - Raleigh NC

International Business Machines Corporation - Armonk NY

G06F 1516

709248, 709249, 709250, 709253, 370445, 370447, 370510

A method and system for adjusting a message preamble on a shared bus, wherein the message preamble includes N synchronization characters, and each of the synchronization characters is separated in time by a random delay interval. First, an activity status is determined for the shared bus in terms of the number of stations that are currently active on the bus. The number of synchronization characters is then adjusted according to the bus activity status. The activity status is also utilized as a dynamic adjustment parameter for the random delay interval that includes a fixed delay term, D, added to a randomly determined delay increment, d. In this manner the message preamble specification is optimized according to real-time network demands, such that latency and data collisions are minimized.

6816499, Nov 9, 2004

Jun 11, 1999

09/330969

Max Robert Povse - Apex NC
Larry Wayne Reynolds - Cary NC
Edward Stanley Suffern - Chapel Hill NC

International Business Machines Corporation - Armonk NY

H04L 1256

370421, 375257

A device that automatically configures the high speed (100 Mbps) port of a switch both physically and logically is provided. For physical configuration, the device determines the transmit pair conductor and the receive pair conductor interconnecting a switch port and Data Terminal Equipment (DTE). The switch port can be termed âPrimaryâ while the connected DTE is termed âSecondaryâ or vice versa. For logical configurations, the device makes the determination. The device includes a switching relay, a controller and impedance matching transformers. The controller switches the connections between the relay and the transformer so that the impedance is substantially constant even though the output from the high speed port is feeding different resistive transmission lines.

6880012, Apr 12, 2005

Aug 2, 2000

09/630912

Benjamin Russell Grimes - Zebulon NC, US
Jillian Lynn Kaufman - Raleigh NC, US
Edward Stanley Suffern - Chapel Hill NC, US

International Business Machines Corporation - Armonk NY

G06F015/16

709227, 709229, 709209

A system, method, and program for establishing modem communication between a master computer system and a plurality of slave computer systems coupled to a common serial communication channel are disclosed. An unique identification number (UIN) is assigned to each slave computer system. The slave system modems are initialized to a “receive mode. ” The master computer system directs a session request to the slave computer systems through the common serial communication channel. All slave computer systems receive and respond to the session request by changing each of the modems to the “answer mode. ” After all the slave computer systems respond, the master computer system requests communication with a particular slave computer system among the plurality of slave computer systems. The master computer system uses the UIN to establish communication with the particular slave computer system by sending a request from the master computer system specifying the UIN of the particular slave computer system. After the master computer system requests communication with the particular slave computer system, communication is maintained between the master computer system and only the particular slave computer system by maintaining the modem for the particular slave computer system in the “answer mode” and switching the modems for all other slave computer systems to the “receive mode.

7002911, Feb 21, 2006

Jul 15, 1999

09/353996

John Walter Linville - Durham NC, US
Brad Alan Makrucki - Durham NC, US
Edward Stanley Suffern - Chapel Hill NC, US
Jeffrey Robert Warren - Apex NC, US

International Business Machines Corporation - Armonk NY

H04L 12/24
H04L 12/56

370230, 370236

Data bandwidth on a congested link of a Carrier Sense Multiple Access with Collision Detection (CSMA/CD) network, is controlled by a Pause MAC Frame which carries information identifying devices causing the problem on the link. The Recipient of the Pause MAC Frame can use the information to deactivate selected devices on the congested link. As a consequence, the data bandwidth is reduced without shutting down the link.

7191347, Mar 13, 2007

Dec 31, 2002

10/248255

Brian E. Bigelow - Apex NC, US
Jospeh E. Bolan - Morrissville NC, US
Gregory W. Dake - Durham NC, US
Harold D. Hudson - Pittsboro NC, US
Edward S. Suffern - Chapel Hill NC, US

International Business Machines Corporation - Armonk NY

G06F 1/00

713300, 713330, 713340

The present invention facilitates the management of blade server information handling systems by providing a non-disruptive indication signal for power management to a Management Module or similar. By providing said functionality and method, the power available to power the management module at a particular instance, in relationship to the power state of the chassis may be determined, such that power may be configured in a non-disruptive manner for the management module. Said functionality and method also provides for both the safe and efficient powering of a server from cold start or the hot-plugging of a management module into an active, powered chassis.