Additionally, in accordance with methods and systems consistent with the present invention, a number of network topologies are provided that are designed to facilitate recon?guration.

Peercy, M. et al., “Distributed Algorithms for Shortest—Path, Deadlock—Free Routing and Broadcasting in Arbitrarily Faulty Hypercubes,” International Symposium on Fault Tol erant Computing Systems (FTCS), US, Los Alamitos, IEEE Comp.

As used herein, the term “node” refers to any device capable of communicating, such as a computer, router, switch, network processor, or symmetric multiprocessor.

Additionally, in accor dance with methods and systems consistent with the present invention, a number of network topologies are provided that are designed to facilitate recon?guration.

FIG. 15 depicts a flowchart of the steps performed-when reconfiguring one of the network topologies in accordance with methods and systems consistent with the present inven- tion.

Descriptions of parallel computing systems are found I in US Pat. Nos. 4,636,948, assigned to IBM; 4,514,807 to Tatsuo; 4,814,980 and 4,730,322 to California Insti tute of Technology; 4,811,214 to 4,543,630 to Teradata; 4,748,660 to Jeumont-Schneider; and 4,598,400 to Thinking Machines Corp.

These patents describe vari ous architectures for parallel processing that represent 5,181,017 earlier development of routing systems similar to the subject invention.

If the received byte is a LINK-CLOSE command the processing continues in function block 9190 where the path through the crossbar switch is closed and the state becomes IDLE.

The delay between END-OF-MESSAGE and ACKNOWLEDGE or LINK-CLOSE should be as short as possible since the routing mechanism holds a path back to the source node open until the ?nal LINK CLOSE is seen.

The method as recited in claim 1, wherein in step (a) said header portion includes at least one bit position corresponding to each output link transmitter in said one node, said method comprising the further steps of: responding to a bit group state to establish said con nection to a corresponding output link transmitter, except if said corresponding output link transmitter manifests busy or inoperative signals to indicate a non-available status; and transmitting to said preceding node a link-close signal in response to said manifestation of said busy or inoperative signals.

In one embodiment, the distance that the edge connection request message travels is established by the portal computer to be approximately twi e the estimated diameter of the broadcast channel.

The broadcast technique can then repeatedly invoke the XDR routines to retri ve the messages and use the success or failure of each invocation to determine whether ano er block of 1,024 bytes needs to be retrieved from the operating system.

In block 1303, the IOITILIDC invokes the Figure 14 is a flow diagram illustrating the processing Lpf the external connect callback routine to notify the application program and then returns.

Coincident with these developments and the uptake in adoption of these applications, Clip2 has seen a steady increase in the number of responsive hosts active ‘at any given time on the network, rising from a typical figure of 500 in October to more than 1500 in early January 2001.

Response times vary considerably for any given mes- sage, depending on load and on how ‘far away” one is from the token holder when there is a need to speak Frequently low-end electronic operator interface (EOI) units will be found on I/0 networks, basically replacing simple push button, pilot light and meters.

Such applications often involve many par- ticipants and typically require a specific form of mul- ticast communication called dissemination in which a single sender must reliably transmit data to multiple receivers in a timely fashion.

The control tree hierarchy employs restricted nooks with suppression and an expanding ring search to distribute the functions of state management and error recovery among many members, thereby allowing scalability to large numbers of receivers.

Third, multiple Tack in- tervals during the retransmission interval provide suffi- cient opportunity for a domain manager to recover from transient network load in its part of the subtree without unnecessarily applying backpressure to the sender.

After some preliminary evaluation of different setting for N, our empirical results indicated that N = 3 provides suffi- cient time for local domains to recover without delaying acks unnecessarily or consuming too much buffer space.

Computation capability in intermediate nodes is usumed limited so that all decisions made enroute should be simple and could not rely, for example, on generating random numbers or on tables that grow with the size of the network.

Attachment 11d: Statewide Illinois Library Catalog record for IEEE Transactions on Computers

Attachment 11e: List of publications citing Document 11 identified by IEEE Xplore Digital Library

Attachment 11e: List of publications citing Document 11 identified by IEEE Xplore Digital Library

Attachment 12a: Copy of Document 12 from the University of Illinois at Urbana-Champaign Library ACTIVISION, EA, TAKE-TWO, 2K, ROCKSTAR, Ex. 1004, p. 378 of 787

Attachment 12a: Copy of Document 12 from the University of Illinois at Urbana-Champaign Library ACTIVISION, EA, TAKE-TWO, 2K, ROCKSTAR, Ex. 1004, p. 379 of 787