🚀 go-pugleaf

In article <1993Jan23.231824.24755@rose.com> gyl.midroni@rose.com (gyl midroni) writes:
	[DELETIONS]
>    It seems to me that local bus disk access will be a speed booster 
>for the reason of faster data communicatio with the CPU / memory.  It 
>is four times faster to get data off the disk at 33 MHz off the local 
>bus connection to the CPU and memory than at 8 MHz via the ISA bus. 
>As simple as that... no?   Why would this impact system performance 

	Seems like the local bus might deliver 33 Mhz x 4 bytes delivered
in parallel for 132 million bytes/sec, and the 8 MHz ISA 16-bit bus slot
can deliver about 5.33 million bytes/sec (assuming 2 bytes delivered
in parallel every 3 clock cycles). So the local bus could be about
25 times faster than the ISA bus. That is great for rapidly stuffing huge 
quantities of image information into electronic-quick video memory on a local 
bus video card. It might not be so great for the mechanical-slow ( relatively
speaking ) rotating machinery of a hard drive managed by a local bus disk
controller. 

	Even the fastest single drives can only spin the bytes under
the drive's head at around 3 million bytes/sec.

	So, isn't a 5.33 million bytes/sec 16-bit slot on the ISA bus pretty
much up to the task of delivering and retrieving data to and from a
single hard drive that is only capable of spinning the data off or onto
the platters at a 3 million bytes/sec rate?
 
	And isn't it expensive overkill to provide a 132 million bytes/sec
truck (local bus) to haul 3 million bytes/sec freight ( a fast hard drive's
maximum, spin-limited data transfer rate)?  Seems like that expensive local 
bus truck will be sitting idle almost 98% of the time. Even the cheaper
ISA bus truck will be idle almost 44% of the time. This assumes that
DMA transfers from disk to memory can burst the data at the bus rate and then
free up the memory address bus until the next truckload of bytes can be spun 
up off the disk. (Can a local bus card use DMA, as I am assuming?)

	Perhaps in a heavy, multitasking environment there may be some
merit to the ability of the local bus to more than double (98/44 = 2.2) the
amount of time relative to an ISA bus that the memory address bus is freed up.
However, whether or not the ongoing multiple tasks can even make use of the 
freed time depends on whether or not any of the tasks need to exchange data 
with RAM memory.

	I guess it seems to me that I can't necessarily conclude that
the fantastic data rate capability of a local bus disk controller is going
to give me any noticeable performance improvement over a 16-bit ISA bus
disk controller -- especially if I am running a typical single-user
system without heavy multitasking requirements. Certainly, I am not going
to improve the prolonged data transfer rate onto or off the disk -- 
3 million bytes/sec is as fast as data physically can be sucked up or spewed
out.

	I am not an expert in any of this and I am ready to learn 
why I might be wrong in any of the comments above. I am simply trying to 
understand what the real performance impact is likely to be of some quite 
expensive, heavily promoted, and increasingly popular decisions regarding 
hardware for the machine my wife keeps her recipes on :-).


_____________________________________________________________________________
Bob Davis, UofALA alum \\ INTERNET: sonny@trantor.harris-atd.com  |  _   _  |
Harris Corporation, ESS \\    UUCP: ...!uunet!x102a!trantor!sonny |_| |_| | |
Advanced Technology Dept.\\ AETHER: K4VNO          |==============|_/\/\/\|_|
PO Box 37, MS 16/1912     \\ VOICE: (407) 727-5886 | I SPEAK ONLY | |_| |_| |
Melbourne, FL 32902 USA    \\  FAX: (407) 729-3363 | FOR MYSELF.  |_________|