Abstract: The network switches and computer readable mediums of the present invention allocate port assignments based on load, that is, the amount of data being forwarded through each port in the group. The load balancing of the present invention is preferably dynamic, that is, packets from a given stream may be forwarded on different ports depending upon each port's current utilization. When a new port is selected to transmit a particular packet stream, it is done so that the packets cannot be forwarded out of order. This is preferably accomplished by ensuring passage of a period of time sufficient to allow all packets of a given stream to be forwarded by a port before a different port is allocated to transmit packets of the same stream. The invention may be used in a variety of different network environments and speeds, including 10Base-T, 100Base-T, and Gigabit Ethernet, and other network environments.

Abstract: The network switches and computer readable mediums of the present invention allocate port assignments based on load, that is, the amount of data being forwarded through each port in the group. The load balancing of the present invention is preferably dynamic, that is, packets from a given stream may be forwarded on different ports depending upon each port's current utilization. When a new port is selected to transmit a particular packet stream, it is done so that the packets cannot be forwarded out of order. This is preferably accomplished by ensuring passage of a period of time sufficient to allow all packets of a given stream to be forwarded by a port before a different port is allocated to transmit packets of the same stream. The invention may be used in a variety of different network environments and speeds, including 10Base-T, 100Base-T, and Gigabit Ethernet, and other network environments.

Abstract: Provided are methods, apparatuses and systems for balancing the load of data transmissions through a port aggregation. The methods and apparatuses of the present invention allocate port assignments based on load, that is, the amount of data being forwarded through each port in the group. The load balancing of the present invention is preferably dynamic, that is, packets from a given stream may be forwarded on different ports depending upon each port's current utilization. When a new port is selected to transmit a particular packet stream, it is done so that the packets cannot be forwarded out of order. This is preferably accomplished by ensuring passage of a period of time sufficient to allow all packets of a given stream to be forwarded by a port before a different port is allocated to transmit packets of the same stream. The invention may be used in a variety of different network environments and speeds, including 10Base-T, 100Base-T, and Gigabit Ethernet, and other network environments.

Abstract: Data is stored in a memory in a manner which eliminates dead time which occurs when the number of words in a page which are read-out are insufficient to provide enough time for simultaneously opening the next page. If the length of a frame being stored in memory is not an exact integral multiple of words in a page, a penultimate (or earlier) page is written with fewer words than the page can hold. This allows additional words to be placed into the last page, sufficient to provide every page used for storing a frame at least a number of words equal to the number of clock cycles needed for opening a next page.

Abstract: Provided are methods, apparatuses and systems for balancing the load of data transmissions through a port aggregation. The methods and apparatuses of the present invention allocate port assignments based on load, that is, the amount of data being forwarded through each port in the group. The load balancing of the present invention is preferably dynamic, that is, packets from a given stream may be forwarded on different ports depending upon each port's current utilization. When a new port is selected to transmit a particular packet stream, it is done so that the packets cannot be forwarded out of order. This is preferably accomplished by ensuring passage of a period of time sufficient to allow all packets of a given stream to be forwarded by a port before a different port is allocated to transmit packets of the same stream. The invention may be used in a variety of different network environments and speeds, including 10Base-T, 100Base-T, and Gigabit Ethernet, and other network environments.

Abstract: A memory system reducing or eliminating the effects of DRAM page-opening delays or row access delays is provided. The system uses DRAM and fast memory such as SRAM. SRAM is used to store the initial portions of data from data blocks and corresponding portions of DRAM are used to store the terminal portions of data from the data blocks. When access to a block of data is requested, DRAM row access procedures are initiated. During the delay period, while DRAM row access procedures are occurring, the initial portion of data from the requested block is read-out from SRAM. By about the time the initial data read-out from SRAM is completed, DRAM row access procedures are completed and the remaining portion of the data is read-out from DRAM.

Abstract: A method allows a designer to implement software for a wide variety of variant host architectures, without excessive usage of host memory, nor sacrificing the capabilities of high end versions of the variant architectures available. The method is based on providing an initialization module of the software to host memory. A portion of the initialization module determines the host architecture. Based on the determined host architecture, the unneeded portions of the initialization module are freed, and the needed portions are relocated into a contiguous memory space to minimize host memory usage. Any location dependent entries in the needed portions of the program are then updated based on the relocation. The initialization module includes a plurality of code blocks, each of which is optimized to a particular variant architecture. When the variant architecture of the host is identified, those code blocks which are optimized to the identified host are selected and the other code blocks are freed.

Abstract: Indication and interrupt signals generated by a network adapter representing asynchronous events are managed by a host system. The network adapter includes a first mask logic for selectively disabling the indication signals from being stored in a first memory location by the host writing to a first mask register. A second mask logic which is coupled to the first memory location also selectively disables the indication signals from being stored in a second memory location creating two levels of status information. The indication signals may also be disabled from being stored in the second memory location responsive to the host writing to a second mask register. The first memory location may be read from the host in order to determine whether a network event occurred during an interrupt service routine, while interrupt means generates an interrupt signal to the host responsive to the value in the second memory location.

Abstract: A method allows a designer to implement software for a wide variety of variant host architectures, without excessive usage of host memory, nor sacrificing the capabilities of high end versions of the variant architectures available. The method is based on providing an initialization module of the software to host memory. A portion of the initialization module determines the host architecture. Based on the determined host architecture, the unneeded portions of the initialization module are freed, and the needed portions are relocated into a contiguous memory space to minimize host memory usage. Any location dependent entries in the needed portions of the program are then updated based on the relocation. The initialization module includes a plurality of code blocks, each of which is optimized to a particular variant architecture. When the variant architecture of the host is identified, those code blocks which are optimized to the identified host are selected and the other code blocks are freed.