At least 100 million AMD systems have been produced already (at least
those mated to VIA chipsets).

http://www.stockhouse.com/news/news.asp?newsid=2657664

Yousuf Khan

On 23 Feb 2005 07:29:50 -0800, "YKhan" <(E-Mail Removed)> wrote:

>At least 100 million AMD systems have been produced already (at least
>those mated to VIA chipsets).
>
>http://www.stockhouse.com/news/news.asp?newsid=2657664

AMD has been selling somewhere on the order of 20-40 million chips a
year for the past 7 years now, so I would say that 100M chips is
definitely a low-ball figure for total AMD processor sales.

Also, my understanding of it is that the 100M VIA/AMD chipsets are
strictly for Athlon (+ Duron, AthlonXP, Sempron) and Athlon64/Opteron
processors. I'm guessing that the earlier chipsets for the K5 and K6
chips would not have been counted as "AMD systems" since the chipsets
could just as well have been used for Intel processors. The same is
even more true for 486 and earlier processors.

-------------
Tony Hill
hilla <underscore> 20 <at> yahoo <dot> ca

Tony Hill wrote:
> Also, my understanding of it is that the 100M VIA/AMD chipsets are
> strictly for Athlon (+ Duron, AthlonXP, Sempron) and Athlon64/Opteron
> processors. I'm guessing that the earlier chipsets for the K5 and K6
> chips would not have been counted as "AMD systems" since the chipsets
> could just as well have been used for Intel processors. The same is
> even more true for 486 and earlier processors.

That would be my guess as well, since it's hard to figure out whether
those 486 to Pentium-class chipsets were being mated to what. The
article did mention that VIA has had a cooperation with AMD since the
486 days.

I guess in the 486 days, AMD and Cyrix were the only ones that took the
486's FSB upto 40Mhz, since Intel stopped at 33Mhz. Only VIA was
producing 486 chipsets that went to those levels, so maybe they can
assume those went to non-Intel systems.

In the K6/Pentium/6x86 days, Socket 7 development was stopped at 66Mhz
by Intel, but I think AMD & Cyrix took it out to 100Mhz. Still no
guarantee it was specifically for AMD as it could've been for Cyrix as
well, it was just highly likely not for Intel.

Yousuf Khan

On 24 Feb 2005 07:27:45 -0800, "YKhan" <(E-Mail Removed)> wrote:

>Tony Hill wrote:
>> Also, my understanding of it is that the 100M VIA/AMD chipsets are
>> strictly for Athlon (+ Duron, AthlonXP, Sempron) and Athlon64/Opteron
>> processors. I'm guessing that the earlier chipsets for the K5 and K6
>> chips would not have been counted as "AMD systems" since the chipsets
>> could just as well have been used for Intel processors. The same is
>> even more true for 486 and earlier processors.
>
>That would be my guess as well, since it's hard to figure out whether
>those 486 to Pentium-class chipsets were being mated to what. The
>article did mention that VIA has had a cooperation with AMD since the
>486 days.
>
>I guess in the 486 days, AMD and Cyrix were the only ones that took the
>486's FSB upto 40Mhz, since Intel stopped at 33Mhz. Only VIA was
>producing 486 chipsets that went to those levels, so maybe they can
>assume those went to non-Intel systems.

I don't recall seeing a VIA chipset back then for 486. Their earliest
chipsets, with their name on them that I recall, were the VP1 and VPX which
were for Pentium compatible CPUs. I had a VP2 (aka AMD-640) mbrd with a
Cyrix 6x86 which was a Pentium class system. More well known as 486
chipsets were OPTi and UMC, IIRC. Before the 6x86, my first major upgrade
was swapping out a POS Gigabyte "586" mbrd with i486/33 for a Shuttle mbrd
with a "UMC 8881F/8886AF and 8663AF" chipset and a Cyrix 5x86/120.

It's hard to tell what the details were but I believe there's a suspicion
that VIA branded chipsets grew out of technology/people transfers to/from
OPTi and/or UMC.

>In the K6/Pentium/6x86 days, Socket 7 development was stopped at 66Mhz
>by Intel, but I think AMD & Cyrix took it out to 100Mhz. Still no
>guarantee it was specifically for AMD as it could've been for Cyrix as
>well, it was just highly likely not for Intel.

Yes, that's strictly true, but by the time Cyrix introduced the MII/100
they were pretty much dead in the water and then Nat Semi ****ed it all up.
Funny, but I now see that Nat Semi has "acknowledged" that their niche is
analog.:-)

--
Rgds, George Macdonald

George Macdonald wrote:
> I don't recall seeing a VIA chipset back then for 486. Their earliest
> chipsets, with their name on them that I recall, were the VP1 and VPX which
> were for Pentium compatible CPUs. I had a VP2 (aka AMD-640) mbrd with a
> Cyrix 6x86 which was a Pentium class system. More well known as 486
> chipsets were OPTi and UMC, IIRC. Before the 6x86, my first major upgrade
> was swapping out a POS Gigabyte "586" mbrd with i486/33 for a Shuttle mbrd
> with a "UMC 8881F/8886AF and 8663AF" chipset and a Cyrix 5x86/120.

Back in those days, I can recall the big chipset houses were ALI, SIS,
and OPTI. UMC might have had a business in it too, but I didn't see too
many of them. Similarly, I would gather VIA was also around, just not a
big player yet.

The really earliest chipset I can think of was from Chips &
Technologies, they had a chipset business since the 286 days, I can
recall. Then they just fell by the wayside, after creating the whole
business.

Yousuf Khan

On Thu, 24 Feb 2005 23:41:18 -0500, Yousuf Khan <(E-Mail Removed)> wrote:

>George Macdonald wrote:
>> I don't recall seeing a VIA chipset back then for 486. Their earliest
>> chipsets, with their name on them that I recall, were the VP1 and VPX which
>> were for Pentium compatible CPUs. I had a VP2 (aka AMD-640) mbrd with a
>> Cyrix 6x86 which was a Pentium class system. More well known as 486
>> chipsets were OPTi and UMC, IIRC. Before the 6x86, my first major upgrade
>> was swapping out a POS Gigabyte "586" mbrd with i486/33 for a Shuttle mbrd
>> with a "UMC 8881F/8886AF and 8663AF" chipset and a Cyrix 5x86/120.
>
>Back in those days, I can recall the big chipset houses were ALI, SIS,
>and OPTI. UMC might have had a business in it too, but I didn't see too
>many of them. Similarly, I would gather VIA was also around, just not a
>big player yet.

Yeah I think UMC came in more towards the end of the 486 with the 486
enhanced mbrds with PCI; the Shuttle HOT-433, which used the UMC chipset,
actually worked quite well for me though I don't think it did any Bus
Mastering at all. Did Intel ever make a 486 PCI mbrd? ISTR no?? VIA
started in Silicon Valley, before moving to Taiwan, more as a design house
AFAIK - it wouldn't surprise me to learn that VIA designs were used in
particularly the UMC chipset... but possibly some of the others. They used
the same nomenclature as VIA ended up using: 82C5xx etc. etc.

>The really earliest chipset I can think of was from Chips &
>Technologies, they had a chipset business since the 286 days, I can
>recall. Then they just fell by the wayside, after creating the whole
>business.

The C&T was a PITA for many 3rd part cards - we were using Definicon 68020
add-in cards and they wouldn't work in C&T boards - something to do with D
flip-flops being used for latching the upper address lines, IIRC.

--
Rgds, George Macdonald