US2008089322A1PendingUtilityA1

Storage medium for delay optimization for scheduling in bufferless crossbar switches

Assignee: IBMPriority: Jan 31, 2005Filed: Nov 30, 2007Published: Apr 17, 2008
Est. expiryJan 31, 2025(expired)· nominal 20-yr term from priority
H04Q 2213/1302H04Q 2213/1304H04Q 3/5455H04Q 2213/13242H04Q 2213/13003H04Q 2213/1331
54
PatentIndex Score
0
Cited by
0
References
0
Claims

Abstract

A computer readable storage medium includes instructions that, when executed by a computer, implement a method for delay optimization scheduling in bufferless crossbar switches, the method including: transmitting, by each line card of a plurality of line cards logically organized into rows and columns, scheduling information to a horizontal control broadcast network and a vertical control broadcast network; receiving, by each line card, a plurality of requests from the horizontal and the vertical control broadcast networks connected to that line card, the requests being from the other line cards in a same row and a same column as that line card; performing, by each line card, a two-phase distribution of the requests; determining, by each line card, a partial schedule for that line card based on the scheduling information and the requests; and sending, by each line card, data according to the partial schedule through a bufferless crossbar switch.

Claims

exact text as granted — not AI-modified
1 . A computer readable storage medium comprising instructions that, when executed by a computer, implement a method for delay optimization scheduling in bufferless crossbar switches, the method comprising: 
 transmitting, by each line card of a plurality of line cards logically organized into rows and columns, scheduling information to a horizontal control broadcast network and a vertical control broadcast network;    receiving, by each line card, a plurality of requests from the horizontal and the vertical control broadcast networks connected to that line card, the requests being from the other line cards in a same row and a same column as that line card;    performing, by each line card, a two-phase distribution of the requests;    determining, by each line card, a partial schedule for that line card based on the scheduling information and the requests; and    sending, by each line card, data according to the partial schedule through a bufferless crossbar switch.    
   
   
       2 . The storage medium of  claim 1 , wherein each line card transmits the information approximately simultaneously to the horizontal and the vertical control broadcast networks connected to that line card.  
   
   
       3 . The storage medium of  claim 1 , wherein the bufferless crossbar switch is a replicated distributed responseless crossbar switch.  
   
   
       4 . The storage medium of  claim 1 , wherein each line card performs the two-phase distribution of the requests by the following: 
 aggregating, by each line card, the requests from the other line cards in the same row as that line card into a single message;    forwarding, by each line card, the single message to the other line cards in the same column as that line card;    aggregating, by each line card, the requests from the other line cards in the same column as that line card into a single message; and    forwarding, by each line card, the single message to the other line cards in the same row as that line card.    
   
   
       5 . The storage medium of  claim 1 , wherein error checking is performed before determining the partial schedule.

Join the waitlist — get patent alerts

Track US2008089322A1 — get alerts on status changes and closely related new filings.

We store only your email — no account needed. See our privacy policy.