http://ps-2.kev009.com/ohland/8550/Under_The_Cover.html

I've had this one done months ago.

On Sun, 4 Jul 2021 12:24:23 -0500, Louis Ohland <ohland@charter.net>
wrote:

>http://ps-2.kev009.com/ohland/8550/Under_The_Cover.html
>
>I've had this one done months ago.

I always thought the 50 was just a crippled 70 and the 60 had all the
worst traits of any of them. A microchannel 286 never made sense when
you could just up clock an AT and use cheap cards in it.

Greg, I've had bi-curiosity of how the 8530-286 performs compared to the
8550 [maybe 50z?].

IBM though the 286 was sufficient [at the time] for run-of-the-mill
business apps [DOS based].

So... how does the AT bus 286 compare to the MC bus 286 for DOS and a
multi-tasking OS?

On 7/4/2021 14:07, gfretwell@aol.com wrote:
> On Sun, 4 Jul 2021 12:24:23 -0500, Louis Ohland <ohland@charter.net>
> wrote:
>
>> http://ps-2.kev009.com/ohland/8550/Under_The_Cover.html
>>
>> I've had this one done months ago.
>
> I always thought the 50 was just a crippled 70 and the 60 had all the
> worst traits of any of them. A microchannel 286 never made sense when
> you could just up clock an AT and use cheap cards in it.
>

On Sun, 4 Jul 2021 14:12:36 -0500, Louis Ohland <ohland@charter.net>
wrote:

>Greg, I've had bi-curiosity of how the 8530-286 performs compared to the
>8550 [maybe 50z?].
>
>IBM though the 286 was sufficient [at the time] for run-of-the-mill
>business apps [DOS based].
>
>So... how does the AT bus 286 compare to the MC bus 286 for DOS and a
>multi-tasking OS?
>
I had an AT 339 running 10mz with the Aust=n BIOS and I couldn't see
any difference between that and a 50. I never ran any benchmarks tho.

> >> http://ps-2.kev009.com/ohland/8550/Under_The_Cover.html
> >> I've had this one done months ago.

That's a different 8550-021 hard drive controller shown in the Byte article - pity the photos aren't higher resolution.

Between the 25 286/30 286 'Type 1' and 'Type 2' planars to the 8550-021 it's probably going to be real close - although additional memory over the first megabyte was on adapters on the bus. The 80287 was two-thirds of the CPU frequency (10MHz) on the 25 286 and 30 286, unless a 287XL was used. And the original hard drive interfaces will be close.

It is strange that a 25 286 / 30 286 had a better planar RAM ability, larger drives (the 30 and 45Mb), and easier ability to hit the 16Mb system RAM limit with aftermarket boards than the 8550-021. But the 50Z beats them all. Does a benchmark need to be done?

Dave, I've wondered how "bad" the ISA models were. Not too sure they
were "bad", I think IBM sold a butt-load of 'em.

What systems were roughly equivalent, 286-wise? Those should be the ones
that are compared and in some manner, "benchmarked"

What would be a reasonable test / benchmark app or utility?

Supposedly, ISA was OK for single-tasking, MCA was for multi-tasking. My
SWAG is they should be sordid the same for DOS, but OS/2 should start to
pull away with OS/2.

On 7/7/2021 00:53, IBMMuseum wrote:
> It is strange that a 25 286 / 30 286 had a better planar RAM ability, larger drives (the 30 and 45Mb), and easier ability to hit the 16Mb system RAM limit with aftermarket boards than the 8550-021. But the 50Z beats them all. Does a benchmark need to be done?

Interesting reading...

Cleaned up and added here:

https://www.ardent-tool.com/8550/Under_The_Cover.html

On 4.7.2021 19:24, Louis Ohland wrote:
> http://ps-2.kev009.com/ohland/8550/Under_The_Cover.html
>
> I've had this one done months ago.