[comp.sys.nsc.32k] DB1600 Development Board questions.

rflukes@violet.waterloo.edu (Richard F. Lukes) (08/04/88)

I am using a DB16000 development board which uses an NS16032 CPU.
I have been told by some people that this CPU had some `bugs' which
were corrected in the 32016. Is this true? Or was the renaming of
the 16032 to the 32016 just a marketing decision?

I was thinking of purchasing a 32032 Designer's kit, and using the
MMU and FPU chips in the DB16000. But if the 16032 has bugs, I may
have to purchase the 32016 Designer's kit. Some information to end
my confusion would be greatly appreciated.

Thanks in advance,

			Richard F. Lukes
			Computer Science
			University of Waterloo
			Waterloo, Ontario
			Canada

daver@nsc.nsc.com (Dave Raulino) (08/04/88)

In article <8113@watdragon.waterloo.edu> rflukes@violet.waterloo.edu (Richard F. Lukes) writes:
>I am using a DB16000 development board which uses an NS16032 CPU.
>I have been told by some people that this CPU had some `bugs' which
>were corrected in the 32016. Is this true? Or was the renaming of
>the 16032 to the 32016 just a marketing decision?
>
>I was thinking of purchasing a 32032 Designer's kit, and using the
>MMU and FPU chips in the DB16000. But if the 16032 has bugs, I may
>have to purchase the 32016 Designer's kit. Some information to end
>my confusion would be greatly appreciated.
>
>
>			Richard F. Lukes

I think that the decision to rename the NS16032 to the NS32016 was a 
marketing decision based on the fact that the the internal architecture
of the chip was full 32 bits.  It was felt that Series 16000 might imply
a 16 bit architecture, whereas Series 32000 implies 32 bits.  As you may
then have guessed, the 16 in 32016 and second 32 in 32032 imply that the
external data bus is 16 and 32 bits, respectively.  (For you smart alecks
out there, no, the 32081 does not have an 81-bit data bus, the 8X series
numbers are for slave processors, i.e., 32081 for FPU, 32082 for MMU).

With regards to bugs, the early Series 16000 devices would not run Unix
properly without major workarounds (they probably had other bugs as well).
I haven't seen a DB16000 in a while.  It was obsoleted by the DB32016.
I would recommend using all of the latest revisions of the devices.  
National does have a University Program, which should be able to help
you in aquiring Designer Kits or devices at most reasonable prices.
Contact Linda Dimmick at (408) 721-7295 for more info.

By the way, I'm a member of a group at National called the Technical
Support Engineering Center (TSEC).  We supply after sales support
to people working with Series 32000 products.  We're devided into
two sections:  one group supports chip design-in and related problems,
and the other (my group) supports developments system, software, and
demo board products.  Call us between 8 AM and 5 PM PST (a machine is 
on line the rest of the day, and we do return our calls) at our
800 numbers:

                  (800) 538-1866  outside California
                  (800) 672-1811  inside CA
                  (800) 223-3248  from Canada

Tell the person who answers the phone what product you want to talk
about, and they'll route you to the appropriate engineer.  

Disclaimer:  I'm not in Marketing (I used to be in Marketing but that's
another story).

-- 
Thank you for your continued support --

                     Dave R.

aeusemrs@csuna.UUCP (Mike Stump) (08/05/88)

In article <8113@watdragon.waterloo.edu> rflukes@violet.waterloo.edu (Richard F. Lukes) writes:
>I am using a DB16000 development board which uses an NS16032 CPU.
>I have been told by some people that this CPU had some `bugs' which
>were corrected in the 32016. Is this true?

Richard, please, if you want help determining something like
this, then include what rev level chips you are using...
Yes, there are problems with some of the early chips...
Just post all the numbers on the backs of each chip, and
maybe someone can help you...
-- 
Mike Stump, Cal State Univ, Northridge Comp Sci Department
uucp: {sdcrdcf, hplabs, ttidca, psivax, csustan}!csun!csuna!aeusemrs

agnew@trwrc.UUCP (Robert A. Agnew) (08/09/88)

In article <5399@nsc.nsc.com>, daver@nsc.nsc.com (Dave Raulino) writes:
> I think that the decision to rename the NS16032 to the NS32016 was a 
> marketing decision based on the fact that the the internal architecture
> of the chip was full 32 bits.  It was felt that Series 16000 might imply
> a 16 bit architecture, whereas Series 32000 implies 32 bits.  As you may

I heard it on the grapevine that TI forced NS to change the name to 32000
before they would agree to second source it! Seems to make sense, but it 
gets very confusing when you talk to TI about 32000 series, they immediately
assume you mean 32020 etc.