On Sat, 16 Mar 2002, Andy Ball wrote:

> Hello Andrew!
> 
>   AB> The 160Mb/sec SCSI doesn't run at 160MHz because it is wider
>     > than the 8 bits of the old sparcs...
> 
> That's one of the reasons, the other is double-transition clocking:
> data is transferred on both the rising and falling edges (of the
> clock?), effectively doubling the maximum possible throughput for any
> given clock speed.

Do you mean that the data is valid at the +ve clock transition and
(next byte) at the -ve transition? OK I know that the data is NRZ so I
suppose that this is practical, just requires a phase skew between
clock & data :-)

>    40 MHz x 2 transfers per clock x 2 bytes per transfer = 160 Mb/S
> 
>      ...theoretical peak, I imagine individual disks are still slow
> enough not to saturate an Ultra160 bus.

That's what cache's are for :-)

> 
>   AB> Could you imagine trying to get reasonable propogation
>     > characteristics at 160Mhz from unscreened ribbon cable?
> 
> No thanks! ;-)

Can you imagine what such signals do to the RF spectrum :-(

Regards

Geoff
-- 
    Geoff Blake	              Please reply to:		Using Linux on  
    Chelmsford		geoff (a) palaemon . co . uk	Intel & NetBSD 
    Essex UK		  geoff @ g8gnz . ampr.org	 on Sun Sparc

------------------------------------------------------------------------------

    This  E-mail  and any attachment(s) is strictly confidential and is 
    intended solely for the  addressee(s). If you are  not the intended
    recipient please notify <postmaster@palaemon.co.uk>  and the sender 
    by return and permanently delete the message. You may not disclose,
    forward or copy this E-mail or any attachment(s) to any third party 
    without the prior consent of the sender.

------------------------------------------------------------------------------