NuBus Video Cards - Slot ID install config. I there really a performance differential?

[Trash80toHP_Mini]

I've heard much over the years about there being an optimum NuBus Slot location for installation of a VidCard to maximize performance. However, I've never actually seen this specified in my many bound PaperDocs or PDF versions of  Apple's Macintosh Family Hardware References.

From the Macintosh IIfx Special Features guide:

NuBus expansion slots

The six NuBus expansion slots are designed for a variety of expansion cards. You use one
of these slots for the video card to which the monitor of your Macintosh IIfx is
connected.

Use the documentation supplied with the card you plan to install, or follow the
instructions in Setting Up Your Macintosh IIfx or in Chapter 9 of the Macintosh Reference
to install a card in one of these slots.

One would think that if ever there were to be an explaination of such a performance differential for install location of a VidCard, Apple would have definitely documented it for the IIfx.

So what's really the story here? Is such a specification to be found in Inside Macintosh? That's about the only set of Docs with which I'm totally unfamiliar. I'm definitely not a programmer  .  .  .  unless you count Basic. :I

 

[Gorgonops]

That sounds suspiciously like an urban legend to me. I mean, it's possible it's based on some germ of truth that turns up in strange edge cases, like if you have multiple video cards there might be some advantage to rearranging them so the fastest one is in the lowest slot and thereby becomes the "primary" or something, but outside of that I can't think of a reason.

 

[Floofies]

I know the PCI standard has something like this, dunno about NuBus though.

 

[Gorgonops]

... the only thing like this I can think of relating to standard PCI would be in strange cases where some of your slots are "direct" to the primary system controller and others are hanging off a PCI bridge. (Which can effectively limit those slots to the aggregate performance of a single card on the direct slots.) Of course, I say "strange case" but this actually perfectly describes how the PCI slots in a B&W G3 or Yikes G4 work. (There's also the oddity there that the slot you're supposed to put the video card in runs at a sort of non-standard 66mhz vs. 33mhz for the bridged slots.)

 

[trag]

For the early PCI Macs there's an Apple note about something I can't remember. Ah, here's the text from back when my brain wasn't pudding:

It is true (sometimes) that the top slot on each bandit may be best performing. This is becasue of a software artifact either in the ROM or the OS by Apple and has to do with support for 32 byte cache line transfers. If your card doesn't need these transfers, it won't care whether it's in the top slot or not. This issue is documented in Apple's Tech Note TN_1008.pdf, "Understanding PCI Bus Performance":

===========================

As an example, if two cards (card x and card y) have addresses mapped into segment 8, one at 0x80800000 and another at 0x80801000, the first call to SetProcessorCacheMode from the driver of card x to make a cacheable address space in segment 8 will work. A second call, say from the driver of card y, to modify the cache setting in segment 8 will not work nor will it report an error. This scenario will most likely result in a lower than expected performance for card y, because card y address space is actually cache inhibited which disables PCI transactions of 32-byte cache lines. If the two cards are mapped into different segments, such as 8 and A, then they both can modify the cache settings withintheir perspective segments. This limitation will be relaxed in the future.

============================