Operating System: Windows
As we know the 13 slot expansion provide 13 slot of PCI, that communicate thru the expansion interface so, this mean the one PCI ( expansion interface ) act as gateway to deliver data from the other PCI within the expansion to the host interface within the host system.
The question is how magma manage the bus bandwidth, is there any bottleneck because the expansion interface seems to shoulder the high data traffic from the rest of the slots.
The PCI bus arbiter generally handles the bus bandwidth issues. Does he mean to ask is there any bottle-necking because of high priority slots on the backplane? If so, then there can be some bottle-necking issues of a high priority slot is used enough to delay traffic that builds up in another slot(s). For example if slot #1 had priority and needed the bus three / seven cycles; with all the other cards/ slots (1-7) only needing the bus one / seven cycles. You would see a bottleneck because you have six slots requiring servicing every seven cycles; out of seven, six are required to service six slots. However, three of every seven would be required for your priority slot, this leaves a deficit of two cycles to service everything, so some of the low priority slots would not get the servicing they require. This can be fixed by moving the high traffic card to a low priority slot, but this may cause bottle-necking with that particular card, but won’t cause overall bottle-necking.