Archived from groups: alt.sys.pc-clone.gateway2000 (
More info?)
Well, 99 times out a 100, when a computer shows up here for service and it needs
drivers (owner lost the driver CD, etc), I install the drivers from the chip
manufacturer's web site and all is well. This is true across all name brands
I've had to service: Gateway, HP, IBM, Dell, Compaq (still a separate brand
name), eMachines (ditto). Some of them, such as eMachines (before the Gateway
acquisition/merger/whatayacallit) had such abysmally bad web sites that one
could not find drivers with a flashlight and bloodhounds. Others, such as HP
with its Pavilion models, provide only the drivers for the operating system that
was shipped with the model, and often not even that! Still other name brand
computer web sites may have the needed drivers, but they are painfully difficult
to navigate. So I got in the habit of using chipset mfrs drivers and never
looked back. With scenarios such as those presented by the HP Pavilions and
eMachines, how would one get the drivers if not from the chipset mfrs web site?
Chip manufacturers may produce a dumbed-down OEM version of a card or chip, but
the costs of supporting the OEM chip would double if it had functionality that
was not present in the mainstream chip. The costs accure in the software area:
writing drivers, debugging and testing the drivers, fixing defects in drivers,
posting updated drivers. Chip manufacturers may sometimes produce dumbed-down
OEM chips, but they are not dumb. They do not want to double their software
development and support costs for themselves. (Compaq did some entirely bizarre
chips way back when, having chips manufactured with different register files
than retail chips, necessitating non-standard drivers. Once Compaq had a taste
for the mess they'd made for themselves, they stopped doing this. Real fast.
The person who made the decision to do this probably got whomped over the head
for the stupid decision.)
Now I must admit to being enough of a chiphead that it is very easy for me to
identify the major chips on all the boards inside a computer.
As my response relates to this newsgroup, EVERY Gateway computer or
Gateway-branded part I have touched in the last 10+ years going back to 386s
used absolutely standard chipsets with no required difference in the driver
software. Even the lamentable Gatrox PCI card uses standard Matrox drivers,
albeit with a crippled RAMDAC which is only peripherally visible to the drivers.
20 years in the biz? I can top that. A cinch. I got my first paying job in
the computer biz in 1962 as an undergraduate student programming in assembly
language on a GE-225 computer. Stick to facts. Years of survival and
adaptation to rapid change in the industry have nothing to do with the topic at
hand.
As for drivers from the Micro$oft update web site, we are 100% in agreement that
installing them is like playing Russian roulette... Ben Myers
On Sun, 10 Apr 2005 05:41:29 GMT, Jed Taylor <aspc-clone-gateway@offthedial.com>
wrote:
>You can take exception to anything you want. The motivation, however,
>is to save costs. OEMs like GW and Dell buy so much at a time that
>they can afford to get a vendor to create a "dumbed-down" version of a
>card that lacks higher-end features the vast majority of buyers will
>never use.
>
>There's a reason OEM's post drivers for their components on their
>websites instead of just telling people to go visit XYZ's. It's that
>these drivers are written for the OEM version, not the retail version.
>
>You do what you want. I've been in the biz for over 20 years, so I'm
>painfully familiar with it's history and evolution.
>
>OTOH, people who download drivers from M$ Update, or load retail
>drivers for OEM cards, make up a share my business, so I guess I
>should just keep quiet and collect the money.
>
>
>
>
>On Thu, 07 Apr 2005 13:09:58 GMT, ben_myers_spam_me_not @ charter.net
>(Ben Myers) wrote:
>
>>Sorry, but I have to take strong exception to the statement that only drivers
>>from the name-brand computer manufacturer's site should be used. Long
>>experience here has shown the opposite.
>>
>>At this time (and going back maybe 6, 7, or more years), there is zero
>>motivation for a name brand computer company to use "special" hardware. It
>>costs more to manufacture special than to manufacture standard off the shelf.
>>It costs oodles more to support special hardware in terms of special drivers and
>>a special supply of spare parts, not to mention tech support, such that it is.
>>
>>There is no difference whatsoever between chipsets used in OEM hardware and
>>retail hardware. About all you will find these days in differences are smallish
>>ones like graphics cards with less memory (e.g. a 32MB card instead of 64MB or
>>128MB) or a graphics chip with less power (a slower nVidia GForce chip for
>>example). Drivers are written so that they autodetect and use the exact memory
>>size on the card they find when the system boots up. Driver software does not
>>care whether the graphics chip is a slower or a faster one. nVidia's single
>>graphics driver set support all the versions of its chips. ATI and Matrox (Is
>>Matrox still in the league?) are very similar. Intel's Extreme Graphics on the
>>motherboard are the same.
>>
>>The same reasoning goes for motherboard chipsets, audio, network, USB and
>>whatever else you want. The computer industry has truly become more of a
>>commodity industry than any other as far as the electronic guts of a computer
>>are concerned. Differences among name brands exist in the styling of the
>>computer chassis, customization of the BIOS splash screen (Dell and IBM give
>>their BIOSes a specific look-and-feel user interface), and choice of
>>manufacturers. But most of the parts are interchangable unless there are
>>physical differences in size, shape, connectors, etc. as with motherboards and
>>power supplies.
>>
>>A number of years ago, there was the "Gatrox" fiasco, caused by Gateway's
>>stubborn insistence on Matrox manufacturing a PCI graphics card with a cheaper,
>>less capable RAMDAC chip. Back then, the RAMDAC chip imposed limits on the
>>refresh rates and/or color depths available with a graphics card. But the
>>standard off-the-shelf Matrox drivers worked 100% perfectly with the Gatrox
>>card. (Nowadays, the RAMDAC functions are built into the graphics chip itself.)
>>
>>Trust me on this one. I've ripped apart and put back together thousands of name
>>brand computers (all brands) in the last 10 years. I read and decipher all the
>>inscrutable markings on integrated circuits... Ben Myers
>>
>>On Thu, 07 Apr 2005 07:57:47 GMT, Jed Taylor <aspc-clone-gateway@offthedial.com>
>>wrote:
>>
>>>I never use M$ Update hardware drivers. They've always corrupted
>>>machines I support.
>>>
>>>OEMs like GW and Dell use non-retail versions of components, so it's
>>>best to use drivers from those sites. Sometime retail drivers work,
>>>but there's no guarantee since they might look for features the OEM
>>>version doesn't have (to keep the costs down).
>>>
>>>If you've added a retail card, then use the vendor's site.
>>>
>>>
>>>
>
>
>
>----------------
>
>"Jim, you think he's with Jesus now? We only have 30 seconds." - Larry King to "Passion" star Jim Caviezel, on the Pope.