on 12/16/99 1:30 PM, Jason Thorpe at thorpej@nas.nasa.gov wrote:


> Um... the IOASIC has a DMA channel for each of its devices that can do
> DMA.  It's simply a case of only supporting DMA SCSI on the IOASIC at
> the moment.
> 
> Actually, we do use other than one DMA channel; the LANCE on the IOASIC
> also does DMA (and we use it; otherwise we wouldn't be able to use the
> LANCE :-)
> 
> The Zilog serial ports can also do DMA, and there are IOASIC DMA channels
> for them (one transmit and receive per UART channel).
> 
> The 79c30 can do DMA, and there are IOASIC DMA channels for it, too.
> 
<SNIP>
> .
> .
> .
> 
> DECstations might be weird sometimes, but they suck far less than PCs :-)
   
   ^
    \_<  Understatement of the year :->



    Now that I have some free time, I'll try and figure out how the DEC's
are wired.
    I ran whetstone today and got 7.8 MIPS for the 5000/25 and 8.3 MIPS on
the 5000/200. Is this simply a matter of the faster TC bus in the 200? The
5000/25 certainly seems to be a much newer chip process, yet still slower
that the 200. They both have the same on chip cache and the same FPU don't
they?

    Chris
-- 
    "I use to be indecisive, now I'm not so sure..."