My take on Nvidia SLI

G

Guest

Guest
Archived from groups: alt.comp.periphs.videocards.nvidia,alt.comp.periphs.videocards.ati (More info?)

7 years ago, 3DFX followed up the Voodoo success with Voodoo 2 SLI.
One board cost about $5-600 at first and was out of reach for most
gamers. I got a hold of a pair of used V2 cheap a few months later
from a distressed owner who was in dire need of some money. That was
the first time I played games in all 1024x768’s glory and was wowed by
it. But I sold off one board a few days later because two boards were
still too rich in price for me.



It wasn’t until TNT2 (or TNT1 Ultra) that Nvidia overtook the speed
crown from V2 SLI. How ironically Nvidia now is the one who brings
SLI back?



Yet if you consider these two iterations of SLI, you’ll see what made
the V2 SLI a success is not presented to the Nvidia case.



Every Mobo had two pci slots back then but we haven’t seen one PCI
express boards out on the market yet.



The difference between playing 800x600 and 1024x768 (V2 SLI) is huge
but not between playing 1280 x1024 and 1600x1200 (Nvidia SLI). Unless
you have a 21” monitor, playing games on 1600x1200 res makes the icons
too small.



The V2 is purely a 3D gaming board that couldn’t do anything else.
This Nvidia board is a full fledge board and Nvidia expects people to
buy an extra board just to play some 3D games at a slightly higher
resolution.



My take is the Nvidia SLI is not gonna fly. People are not be easily
wowed by new boards anymore, unlike we did back 7 years ago when 3Dfx
was the king, Thomas Pabst was a rookie tech reviewer and Anand la
Shrimp just celebrated the 2nd year anniversary of his website.



Well, those were the days.
 
G

Guest

Guest
Archived from groups: alt.comp.periphs.videocards.nvidia,alt.comp.periphs.videocards.ati (More info?)

<nospam@nospam.com> wrote in message
news:harje09ehhbc9r2n7re7vjt5e5keujgv43@4ax.com...
>
> My take is the Nvidia SLI is not gonna fly. People are not be easily
> wowed by new boards anymore, unlike we did back 7 years ago when 3Dfx
> was the king, Thomas Pabst was a rookie tech reviewer and Anand la
> Shrimp just celebrated the 2nd year anniversary of his website.
>
>
>
> Well, those were the days.
>


Both ATI and nVidia are going to be coming out with SLI solutions, but I
have to agree with you that they aren't likely to catch on like the good
ol' Voodoo2. Price is one factor, but also the power consumption involved
and the amount of heat being produced by a higher-end system with high-end
video cards and hard drives would be tremendous. Now that summer has hit,
I've become keenly aware of just how much heat my current Pentium 4 system
with it's 9800 Pro video card is spitting out, and I sure as heck don't
want any more. My other concern with these new SLI solutions is the
proximity of the cards to each other. From pictures I've seen on the web,
the cards are awfully close to one another, and that is a major concern.

It's really little wonder why the Voodoo2 SLI will undoubtedly do better
than dual 6800 or X800 cards. Back in the days of the Voodoo2, the only
big consideration was the price. Now, in 2004, a lot has changed. And
how much of a limiting factor will the CPU be when people get these new
video cards working together?
 
G

Guest

Guest
Archived from groups: alt.comp.periphs.videocards.nvidia,alt.comp.periphs.videocards.ati (More info?)

nospam@nospam.com wrote:
> 7 years ago, 3DFX followed up the Voodoo success with Voodoo 2 SLI.
> One board cost about $5-600 at first and was out of reach for most
> gamers. I got a hold of a pair of used V2 cheap a few months later
> from a distressed owner who was in dire need of some money. That was
> the first time I played games in all 1024x768’s glory and was wowed by
> it. But I sold off one board a few days later because two boards were
> still too rich in price for me.
>
>
>
> It wasn’t until TNT2 (or TNT1 Ultra) that Nvidia overtook the speed
> crown from V2 SLI. How ironically Nvidia now is the one who brings
> SLI back?
>
>
>
> Yet if you consider these two iterations of SLI, you’ll see what made
> the V2 SLI a success is not presented to the Nvidia case.
>
>
>
> Every Mobo had two pci slots back then but we haven’t seen one PCI
> express boards out on the market yet.

When is the SLI thing due to be released? PCI-express is pretty new. It
will gain popularity.

It should be cheap and easy to put two PCI-E slots in, if there is a demand.

> The difference between playing 800x600 and 1024x768 (V2 SLI) is huge
> but not between playing 1280 x1024 and 1600x1200 (Nvidia SLI). Unless
> you have a 21?monitor, playing games on 1600x1200 res makes the icons
> too small.

Works fine on my 19" at 1600x1200@85Hz. :p

> The V2 is purely a 3D gaming board that couldn’t do anything else.
> This Nvidia board is a full fledge board and Nvidia expects people to
> buy an extra board just to play some 3D games at a slightly higher
> resolution.

You're not paying for the 2D stuff, it comes practically free considering
the rest of the 3d stuff and components on the board. And you're getting,
they suspect, ~80-90% improvement, not bad...

> My take is the Nvidia SLI is not gonna fly. People are not be easily
> wowed by new boards anymore, unlike we did back 7 years ago when 3Dfx
> was the king, Thomas Pabst was a rookie tech reviewer and Anand la
> Shrimp just celebrated the 2nd year anniversary of his website.

It should be quick. But the cost will be prohibitive for most.

Ben
--
A7N8X FAQ: www.ben.pope.name/a7n8x_faq.html
Questions by email will likely be ignored, please use the newsgroups.
I'm not just a number. To many, I'm known as a String...
 
G

Guest

Guest
Archived from groups: alt.comp.periphs.videocards.nvidia,alt.comp.periphs.videocards.ati (More info?)

Nvidia have the right to the acronym SLI but don't get confused into
thinking this is Scan Line Interleaving,as it isn't.

From what i've seen only one card is responsible for generating the image,
the other is just a second GPU which performs tasks allocated by the first
GPU. This is far more efficeint than the Voodoo solution as no graphics
calculations are done twice.

It also means that cards don't have to match so another upgrade path
becomes available. You can add a new faster card as the master and keep the
old one as the slave GPU.
 
G

Guest

Guest
Archived from groups: alt.comp.periphs.videocards.nvidia,alt.comp.periphs.videocards.ati (More info?)

[snip]
"John Russell" <john_e_russell@hotmail.com> wrote in message
news:40eea038$1_2@mk-nntp-2.news.uk.tiscali.com...
> Nvidia have the right to the acronym SLI but don't get confused into
> thinking this is Scan Line Interleaving,as it isn't.
[/snip]
According to NVIDIA for SLI = Scalable Link Interface

[snip]
>
> From what i've seen only one card is responsible for generating the image,
> the other is just a second GPU which performs tasks allocated by the first
> GPU. This is far more efficeint than the Voodoo solution as no graphics
> calculations are done twice.
>
[/snip]
Each GPU will start half and will change depending on the load. Software
will determine which load to distribute to which GPU.

[snip]
> It also means that cards don't have to match so another upgrade path
> becomes available. You can add a new faster card as the master and keep
the
> old one as the slave GPU.
>
[/snip]
You will need both identical card. Best is to have that same card from the
same manufacturer.

CapFusion,...
 
G

Guest

Guest
Archived from groups: alt.comp.periphs.videocards.nvidia,alt.comp.periphs.videocards.ati (More info?)

CapFusion wrote:

> [snip]
> "John Russell" <john_e_russell@hotmail.com> wrote in message
> news:40eea038$1_2@mk-nntp-2.news.uk.tiscali.com...
>> Nvidia have the right to the acronym SLI but don't get confused into
>> thinking this is Scan Line Interleaving,as it isn't.
> [/snip]
> According to NVIDIA for SLI = Scalable Link Interface
>
> [snip]
>>
>> From what i've seen only one card is responsible for generating the
>> image, the other is just a second GPU which performs tasks allocated by
>> the first
>> GPU. This is far more efficeint than the Voodoo solution as no graphics
>> calculations are done twice.
>>
> [/snip]
> Each GPU will start half and will change depending on the load. Software
> will determine which load to distribute to which GPU.
>
> [snip]
>> It also means that cards don't have to match so another upgrade path
>> becomes available. You can add a new faster card as the master and keep
> the
>> old one as the slave GPU.
>>
> [/snip]
> You will need both identical card. Best is to have that same card from the
> same manufacturer.

Since there are no boards on the market today or coming in the near future
with two PCI-Express 16x slots (read the fine print on the Alienware very
carefully then read the spec sheets on the components they use and you'll
find that it has one 16x and one 8x), if it depends on the boards being
identical then it's not going to work at all for a good long time.
>
> CapFusion,...

--
--John
Reply to jclarke at ae tee tee global dot net
(was jclarke at eye bee em dot net)
 
G

Guest

Guest
Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

"J. Clarke" <jclarke@nospam.invalid> wrote in message
news:ccmtg801ttb@news4.newsguy.com...
>
> Since there are no boards on the market today or coming in the near future
> with two PCI-Express 16x slots (read the fine print on the Alienware very
> carefully then read the spec sheets on the components they use and you'll
> find that it has one 16x and one 8x), if it depends on the boards being
> identical then it's not going to work at all for a good long time.

I do not get it. Why do you use Alienware as a reference for SLI?
Do you see Alienware have a MoBo with two 16X PCI-E? Naturally or atleast
currently they have only one 16X PCI-E on their top end PC.
Try this link, this may enlighted you abit -
http://www.nvidia.com/page/sli.html

Also can you show me that fine print you refer to? I do not seem to find it
since it so "fine printed". I will check my eye for a new subscription
later. During the meantime, can you point me to that link.

During the meantime, check this link -
http://www.hardwareanalysis.com/content/article/1728/
http://www.neoseeker.com/Articles/Hardware/Features/nvsli/2.html
http://www.firingsquad.com/hardware/nvidia_sli/page4.asp

If IIRC, you mention before regarding Alienware -
Do you remember this thread title -
Nvidia SLI, SLI's back with a vengeance
This subject have be discussed and it seem from previous post and reply that
you already understand and will inform about SLI with 6800 but also
regarding about Alienware PC.

cbqife01rql@news1.newsguy.com
**Quote**
Note by the way that the Alienware board is not going to support arbitrary
video boards--according to their press release it is going to be tied to
specific models and combines the video using a third board. In other words
they've done something nonstandard
**/Quote**
and this thread -
cc4idb01jb9@news3.newsguy.com

CapFusion,...
 
G

Guest

Guest
Archived from groups: alt.comp.periphs.videocards.nvidia,alt.comp.periphs.videocards.ati (More info?)

John Russell wrote:
> Nvidia have the right to the acronym SLI but don't get confused into
> thinking this is Scan Line Interleaving,as it isn't.
>
> From what i've seen only one card is responsible for generating the image,
> the other is just a second GPU which performs tasks allocated by the first
> GPU. This is far more efficeint than the Voodoo solution as no graphics
> calculations are done twice.

The Voodoo solution is entirely symmetrical, with each card performing half
the scan lines, hence double the speed.

nVidias solutions is asymmetric - master/slave, which is LESS efficient - it
will never be able to deliver twice the performance, they reckon up to about
90%.

> It also means that cards don't have to match so another upgrade path
> becomes available. You can add a new faster card as the master and keep
> the old one as the slave GPU.

"Secondly you'll need two identical, same brand and type, PCI-E GeForce 6800
graphics cards"
- http://www.hardwareanalysis.com/content/article/1728/

"SLI will only work on two of the same cards."
- http://www.neoseeker.com/Articles/Hardware/Features/nvsli/2.html

"Additionally, another requirement of SLI is that both cards must come from
the same manufacturer and be based on the same configuration"
- http://www.firingsquad.com/hardware/nvidia_sli/page4.asp

Ben
--
A7N8X FAQ: www.ben.pope.name/a7n8x_faq.html
Questions by email will likely be ignored, please use the newsgroups.
I'm not just a number. To many, I'm known as a String...
 

Andrew

Distinguished
Mar 31, 2004
2,439
0
19,780
Archived from groups: alt.comp.periphs.videocards.nvidia,alt.comp.periphs.videocards.ati (More info?)

On Fri, 9 Jul 2004 19:08:45 +0100, "Ben Pope" <spam@hotmail.com>
wrote:

>The Voodoo solution is entirely symmetrical, with each card performing half
>the scan lines, hence double the speed.

V2 SLI wasn't twice the speed of a single card, 10-50% speed boost was
more usual, or having the ability to do 1024x768.
--
Andrew. To email unscramble nrc@gurjevgrzrboivbhf.pbz & remove spamtrap.
Help make Usenet a better place: English is read downwards,
please don't top post. Trim messages to quote only relevant text.
Check groups.google.com before asking a question.
 
G

Guest

Guest
Archived from groups: alt.comp.periphs.videocards.nvidia,alt.comp.periphs.videocards.ati (More info?)

Andrew wrote:
> On Fri, 9 Jul 2004 19:08:45 +0100, "Ben Pope" <spam@hotmail.com>
> wrote:
>
>> The Voodoo solution is entirely symmetrical, with each card performing
>> half the scan lines, hence double the speed.
>
> V2 SLI wasn't twice the speed of a single card, 10-50% speed boost was
> more usual, or having the ability to do 1024x768.

Is that from a technology standpoint or is it due to the CPU being the
limiting factor?

"This shows that even a Pentium II 300 is not able feeding the Voodoo2 with
enough data at 640x480. It requires a CPU that's significantly faster than
this one."
- http://graphics.tomshardware.com/graphic/19980204/index.html

Whats the fastest CPU anybody has used to bench a Voodoo2 SLI rig?

Had a little look, but not found a site that compares a V2 SLI with a V2 on
a fast (1GHz or so) CPU.

My impression was that at the time, games were fill-rate (or CPU) limited,
and that the SLI would double fill-rate.

Ben
--
A7N8X FAQ: www.ben.pope.name/a7n8x_faq.html
Questions by email will likely be ignored, please use the newsgroups.
I'm not just a number. To many, I'm known as a String...
 
G

Guest

Guest
Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

CapFusion wrote:

>
> "J. Clarke" <jclarke@nospam.invalid> wrote in message
> news:ccmtg801ttb@news4.newsguy.com...
>>
>> Since there are no boards on the market today or coming in the near
>> future with two PCI-Express 16x slots (read the fine print on the
>> Alienware very carefully then read the spec sheets on the components they
>> use and you'll find that it has one 16x and one 8x), if it depends on the
>> boards being identical then it's not going to work at all for a good long
>> time.
>
> I do not get it. Why do you use Alienware as a reference for SLI?

They are the only outfit that has demonstrated a working product.

> Do you see Alienware have a MoBo with two 16X PCI-E? Naturally or atleast
> currently they have only one 16X PCI-E on their top end PC.

<http://www.alienware.com/alx_pages/main_content.aspx>

> Try this link, this may enlighted you abit -
> http://www.nvidia.com/page/sli.html

What of it?

> Also can you show me that fine print you refer to?

<http://www.alienware.com/alx_pages/main_content.aspx#faq> Specifically
"The X2 utilizes all available PCI-Express lanes provided by the Intel 7525
(Tumwater) chipset"

You can find the datasheets on that chipset on the Intel site.

> I do not seem to find
> it since it so "fine printed". I will check my eye for a new subscription
> later. During the meantime, can you point me to that link.
>
> During the meantime, check this link -
> http://www.hardwareanalysis.com/content/article/1728/

"First off you'll need a workstation motherboard featuring two PCI-E-x16
slots which will also use the more expensive Intel Xeon processors." They
do not identify such a board.

> http://www.neoseeker.com/Articles/Hardware/Features/nvsli/2.html

"Nvidia's announcement coincides with Intel's Tumwater chipset which is part
of the enabling technology behind SLI. Tumwater will provide two x16 PCI
Express ports making SLI possible. "

From <http://www.intel.com/pressroom/archive/releases/20040628comp.htm>,
"The Intel® E7525 chipset (formerly codenamed "Tumwater") integrates for
the first time several new technologies that eliminate system bottlenecks
by balancing performance between the processor, I/O and memory." Note that
this is the chipset that Alienware uses.

Now, I'll leave to to you to dig through the PDF data sheets that you can
reach from
<http://www.intel.com/design/chipsets/E7525/index.htm?iid=ipp_srvr_chpsts+e7525_wrkstn&>,
which show that the E7525 has _one_ PCI Express x16 interface and _one_ PCI
Express x8 interface configurable as two x4, so there is no circumstance
under which it will provide two x16 regardless of what neoseeker claims.
Unless of course Intel is lying about the capabilities of their own
product.


> http://www.firingsquad.com/hardware/nvidia_sli/page4.asp

Which is only relevant to the extent that it distinguishes between the
Alienware and nvidia approaches. It does not address the slot issue at
all.

> If IIRC, you mention before regarding Alienware -
> Do you remember this thread title -
> Nvidia SLI, SLI's back with a vengeance
> This subject have be discussed and it seem from previous post and reply
> that you already understand and will inform about SLI with 6800 but also
> regarding about Alienware PC.

Uh, could you try that again--I'm having trouble parsing that last sentence.

> cbqife01rql@news1.newsguy.com
> **Quote**
> Note by the way that the Alienware board is not going to support arbitrary
> video boards--according to their press release it is going to be tied to
> specific models and combines the video using a third board. In other
> words they've done something nonstandard
> **/Quote**
> and this thread -
> cc4idb01jb9@news3.newsguy.com
>
> CapFusion,...

--
--John
Reply to jclarke at ae tee tee global dot net
(was jclarke at eye bee em dot net)
 
G

Guest

Guest
Archived from groups: alt.comp.periphs.videocards.nvidia,alt.comp.periphs.videocards.ati (More info?)

"Ben Pope" <spam@hotmail.com> wrote in message
news:2l87d0Fa2kpqU1@uni-berlin.de...
>
> "This shows that even a Pentium II 300 is not able feeding the Voodoo2
with
> enough data at 640x480. It requires a CPU that's significantly faster
than
> this one."
> - http://graphics.tomshardware.com/graphic/19980204/index.html
>


That's the thing, the CPU on a system would forever be playing catch up to
a dual video card solution. Unless, of course, you hang onto those video
cards past their "Best Before" date. :)
 
G

Guest

Guest
Archived from groups: alt.comp.periphs.videocards.nvidia,alt.comp.periphs.videocards.ati (More info?)

NightSky 421 wrote:
> "Ben Pope" <spam@hotmail.com> wrote in message
> news:2l87d0Fa2kpqU1@uni-berlin.de...
>>
>> "This shows that even a Pentium II 300 is not able feeding the Voodoo2
>> with enough data at 640x480. It requires a CPU that's significantly
>> faster than this one."
>> - http://graphics.tomshardware.com/graphic/19980204/index.html
>>
>
>
> That's the thing, the CPU on a system would forever be playing catch up to
> a dual video card solution. Unless, of course, you hang onto those video
> cards past their "Best Before" date. :)

It has been said that HL2 will run at top whack on a 1GHz machine, provided
you have a good video card. Hence, with two Vid cards only a 2GHz machine
(probably less) would be required to take advantage of them. The thing with
modern graphics cards is that they offload so much of the work that the CPUs
used to do, that you don't need such a fast one.

Ben
--
A7N8X FAQ: www.ben.pope.name/a7n8x_faq.html
Questions by email will likely be ignored, please use the newsgroups.
I'm not just a number. To many, I'm known as a String...
 
G

Guest

Guest
Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

Scroll down if you want to review.

Sheezh, you compare to lemon and lime.
Alieanware have it own so-call SLI using multiple display by mean of VMH
Intel do not have any board with 2 16X slot out yet.
Nvidia is the forefront of this crusade by using Intel PCI-E. And again as
Intel, they do not have one public.

And yes, Alieanware indicate and you pointed out do have SLI but with a
twist. BUT it not the same as using what nvidia indicate using two 16X slot

I not sure you and I in a same page here. Alienware do not have 16X in-mind
using two same card which need two 16X PCI-E slot. Alienware simply using
VMH so it can be use with any VGA card right now. It almost like 3Dfx using
a cable patch to two card.

Go back to your old post regarding about Aleanware
cbqife01rql@news1.newsguy.com
cc4idb01jb9@news3.newsguy.com

I am not quite sure I understand this two statement. It seem to contridict
each other.
- Will I be able to use an ATI card in the same system as an Nvidia card?
- Compare Alienware's Video Array to Nvidia's SLI
Under differences > Flexibilites > 1.

CapFusion,...


"J. Clarke" <jclarke@nospam.invalid> wrote in message
news:ccsf2405i2@news4.newsguy.com...
> CapFusion wrote:
>
> >
> > "J. Clarke" <jclarke@nospam.invalid> wrote in message
> > news:ccmtg801ttb@news4.newsguy.com...
> >>
> >> Since there are no boards on the market today or coming in the near
> >> future with two PCI-Express 16x slots (read the fine print on the
> >> Alienware very carefully then read the spec sheets on the components
they
> >> use and you'll find that it has one 16x and one 8x), if it depends on
the
> >> boards being identical then it's not going to work at all for a good
long
> >> time.
> >
> > I do not get it. Why do you use Alienware as a reference for SLI?
>
> They are the only outfit that has demonstrated a working product.
>
> > Do you see Alienware have a MoBo with two 16X PCI-E? Naturally or
atleast
> > currently they have only one 16X PCI-E on their top end PC.
>
> <http://www.alienware.com/alx_pages/main_content.aspx>
>
> > Try this link, this may enlighted you abit -
> > http://www.nvidia.com/page/sli.html
>
> What of it?
>
> > Also can you show me that fine print you refer to?
>
> <http://www.alienware.com/alx_pages/main_content.aspx#faq> Specifically
> "The X2 utilizes all available PCI-Express lanes provided by the Intel
7525
> (Tumwater) chipset"
>
> You can find the datasheets on that chipset on the Intel site.
>
> > I do not seem to find
> > it since it so "fine printed". I will check my eye for a new
subscription
> > later. During the meantime, can you point me to that link.
> >
> > During the meantime, check this link -
> > http://www.hardwareanalysis.com/content/article/1728/
>
> "First off you'll need a workstation motherboard featuring two PCI-E-x16
> slots which will also use the more expensive Intel Xeon processors." They
> do not identify such a board.
>
> > http://www.neoseeker.com/Articles/Hardware/Features/nvsli/2.html
>
> "Nvidia's announcement coincides with Intel's Tumwater chipset which is
part
> of the enabling technology behind SLI. Tumwater will provide two x16 PCI
> Express ports making SLI possible. "
>
> From <http://www.intel.com/pressroom/archive/releases/20040628comp.htm>,
> "The Intel® E7525 chipset (formerly codenamed "Tumwater") integrates for
> the first time several new technologies that eliminate system bottlenecks
> by balancing performance between the processor, I/O and memory." Note
that
> this is the chipset that Alienware uses.
>
> Now, I'll leave to to you to dig through the PDF data sheets that you can
> reach from
>
<http://www.intel.com/design/chipsets/E7525/index.htm?iid=ipp_srvr_chpsts+e7
525_wrkstn&>,
> which show that the E7525 has _one_ PCI Express x16 interface and _one_
PCI
> Express x8 interface configurable as two x4, so there is no circumstance
> under which it will provide two x16 regardless of what neoseeker claims.
> Unless of course Intel is lying about the capabilities of their own
> product.
>
>
> > http://www.firingsquad.com/hardware/nvidia_sli/page4.asp
>
> Which is only relevant to the extent that it distinguishes between the
> Alienware and nvidia approaches. It does not address the slot issue at
> all.
>
> > If IIRC, you mention before regarding Alienware -
> > Do you remember this thread title -
> > Nvidia SLI, SLI's back with a vengeance
> > This subject have be discussed and it seem from previous post and reply
> > that you already understand and will inform about SLI with 6800 but also
> > regarding about Alienware PC.
>
> Uh, could you try that again--I'm having trouble parsing that last
sentence.
>
> > cbqife01rql@news1.newsguy.com
> > **Quote**
> > Note by the way that the Alienware board is not going to support
arbitrary
> > video boards--according to their press release it is going to be tied to
> > specific models and combines the video using a third board. In other
> > words they've done something nonstandard
> > **/Quote**
> > and this thread -
> > cc4idb01jb9@news3.newsguy.com
> >
> > CapFusion,...
>
> --
> --John
> Reply to jclarke at ae tee tee global dot net
> (was jclarke at eye bee em dot net)
 
G

Guest

Guest
Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

CapFusion wrote:

> Scroll down if you want to review.
>
> Sheezh, you compare to lemon and lime.
> Alieanware have it own so-call SLI using multiple display by mean of VMH
> Intel do not have any board with 2 16X slot out yet.
> Nvidia is the forefront of this crusade by using Intel PCI-E. And again as
> Intel, they do not have one public.
>
> And yes, Alieanware indicate and you pointed out do have SLI but with a
> twist. BUT it not the same as using what nvidia indicate using two 16X
> slot
>
> I not sure you and I in a same page here. Alienware do not have 16X
> in-mind using two same card which need two 16X PCI-E slot. Alienware
> simply using VMH so it can be use with any VGA card right now. It almost
> like 3Dfx using a cable patch to two card.

Uh, you can't buy the VMH "right now" so for all practical purposes it can't
do _anything_ "right now". As for "using it with any video card" something
has to tell the boards what part of the screen to display, and since the
drivers that come with the boards damned well aren't going to do that that
has to be an operating system extension of some sort that may or may not
work with different brands and models of board.

In any case, the only difference between the Alienware VMH and nvidia's new
"SLI" is that nvidia puts the hub on a bridgeboard and does a digital merge
instead of analog.

I have no idea what you mean by "Alienware do not have 16X
in-mind using two same card which need two 16X PCI-E slot.".
>
> Go back to your old post regarding about Aleanware
> cbqife01rql@news1.newsguy.com
> cc4idb01jb9@news3.newsguy.com
>
> I am not quite sure I understand this two statement. It seem to contridict
> each other.

I don't see any contradiction. Alienware has demonstrated a product with
two nvidia boards Geforce FX 5900 PCI Express boards. The only thing that
we can be _certain_ of with regard to their product is that it works well
enough to be used for a demonstration at a computer show using those
boards.

The second post you link is in regard to the availability of slots. There
is no chipset currently available that allows two x16 slots. That is a
fact, and does not contradict the fact that Alienware has demonstrated a
product.

> - Will I be able to use an ATI card in the same system as an Nvidia card?
> - Compare Alienware's Video Array to Nvidia's SLI
> Under differences > Flexibilites > 1.

Whether you will in fact be able to use an ATI card in the same system with
an nvidia I doubt that anybody knows at this point. Alienware may claim
that you can but they haven't demonstrated that you can so assume until you
see it work that their marketing guys are exaggerating the flexibility of
their system. And you certainly aren't going to be able to use an ATI with
an nvidia board using nvidia's system.



> CapFusion,...
>
>
> "J. Clarke" <jclarke@nospam.invalid> wrote in message
> news:ccsf2405i2@news4.newsguy.com...
>> CapFusion wrote:
>>
>> >
>> > "J. Clarke" <jclarke@nospam.invalid> wrote in message
>> > news:ccmtg801ttb@news4.newsguy.com...
>> >>
>> >> Since there are no boards on the market today or coming in the near
>> >> future with two PCI-Express 16x slots (read the fine print on the
>> >> Alienware very carefully then read the spec sheets on the components
> they
>> >> use and you'll find that it has one 16x and one 8x), if it depends on
> the
>> >> boards being identical then it's not going to work at all for a good
> long
>> >> time.
>> >
>> > I do not get it. Why do you use Alienware as a reference for SLI?
>>
>> They are the only outfit that has demonstrated a working product.
>>
>> > Do you see Alienware have a MoBo with two 16X PCI-E? Naturally or
> atleast
>> > currently they have only one 16X PCI-E on their top end PC.
>>
>> <http://www.alienware.com/alx_pages/main_content.aspx>
>>
>> > Try this link, this may enlighted you abit -
>> > http://www.nvidia.com/page/sli.html
>>
>> What of it?
>>
>> > Also can you show me that fine print you refer to?
>>
>> <http://www.alienware.com/alx_pages/main_content.aspx#faq> Specifically
>> "The X2 utilizes all available PCI-Express lanes provided by the Intel
> 7525
>> (Tumwater) chipset"
>>
>> You can find the datasheets on that chipset on the Intel site.
>>
>> > I do not seem to find
>> > it since it so "fine printed". I will check my eye for a new
> subscription
>> > later. During the meantime, can you point me to that link.
>> >
>> > During the meantime, check this link -
>> > http://www.hardwareanalysis.com/content/article/1728/
>>
>> "First off you'll need a workstation motherboard featuring two PCI-E-x16
>> slots which will also use the more expensive Intel Xeon processors."
>> They do not identify such a board.
>>
>> > http://www.neoseeker.com/Articles/Hardware/Features/nvsli/2.html
>>
>> "Nvidia's announcement coincides with Intel's Tumwater chipset which is
> part
>> of the enabling technology behind SLI. Tumwater will provide two x16 PCI
>> Express ports making SLI possible. "
>>
>> From <http://www.intel.com/pressroom/archive/releases/20040628comp.htm>,
>> "The Intel® E7525 chipset (formerly codenamed "Tumwater") integrates for
>> the first time several new technologies that eliminate system bottlenecks
>> by balancing performance between the processor, I/O and memory." Note
> that
>> this is the chipset that Alienware uses.
>>
>> Now, I'll leave to to you to dig through the PDF data sheets that you can
>> reach from
>>
>
<http://www.intel.com/design/chipsets/E7525/index.htm?iid=ipp_srvr_chpsts+e7
> 525_wrkstn&>,
>> which show that the E7525 has _one_ PCI Express x16 interface and _one_
> PCI
>> Express x8 interface configurable as two x4, so there is no circumstance
>> under which it will provide two x16 regardless of what neoseeker claims.
>> Unless of course Intel is lying about the capabilities of their own
>> product.
>>
>>
>> > http://www.firingsquad.com/hardware/nvidia_sli/page4.asp
>>
>> Which is only relevant to the extent that it distinguishes between the
>> Alienware and nvidia approaches. It does not address the slot issue at
>> all.
>>
>> > If IIRC, you mention before regarding Alienware -
>> > Do you remember this thread title -
>> > Nvidia SLI, SLI's back with a vengeance
>> > This subject have be discussed and it seem from previous post and reply
>> > that you already understand and will inform about SLI with 6800 but
>> > also regarding about Alienware PC.
>>
>> Uh, could you try that again--I'm having trouble parsing that last
> sentence.
>>
>> > cbqife01rql@news1.newsguy.com
>> > **Quote**
>> > Note by the way that the Alienware board is not going to support
> arbitrary
>> > video boards--according to their press release it is going to be tied
>> > to
>> > specific models and combines the video using a third board. In other
>> > words they've done something nonstandard
>> > **/Quote**
>> > and this thread -
>> > cc4idb01jb9@news3.newsguy.com
>> >
>> > CapFusion,...
>>
>> --
>> --John
>> Reply to jclarke at ae tee tee global dot net
>> (was jclarke at eye bee em dot net)

--
--John
Reply to jclarke at ae tee tee global dot net
(was jclarke at eye bee em dot net)
 
G

Guest

Guest
Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

"J. Clarke" <jclarke@nospam.invalid> wrote in message
news:ccvl0711av7@news4.newsguy.com...
[snip]
> In any case, the only difference between the Alienware VMH and nvidia's
new
> "SLI" is that nvidia puts the hub on a bridgeboard and does a digital
merge
> instead of analog.
[/snip]
This is where you and I have problem here, regarding SLI..
First I did not mention anything about Alienware as an example of SLI w/2
PCI-E slot until you indicating and point out that Alienware do support SLI.
All I know about alienware regarding their SLI is like 3Dfx way of doing -
using cable.

I also note there is not any public mainboard will have two PCI-E slot yet
include nvidia themself.

[snip]
>
> I have no idea what you mean by "Alienware do not have 16X
> in-mind using two same card which need two 16X PCI-E slot.".
[/snip]
That quote is from you. From your old post awhile back.

[snip]
> I don't see any contradiction. Alienware has demonstrated a product with
> two nvidia boards Geforce FX 5900 PCI Express boards. The only thing that
> we can be _certain_ of with regard to their product is that it works well
> enough to be used for a demonstration at a computer show using those
> boards.
>
[/snip]
You referring to 6800 I assumed.

[snip]
> The second post you link is in regard to the availability of slots. There
> is no chipset currently available that allows two x16 slots. That is a
> fact, and does not contradict the fact that Alienware has demonstrated a
> product.
>
[/snip]
I did not say there was a board that have two PCI-E slot, did I? I only
state that SLI need two 16X PCI-E slot according to nvidia. I thought this
thread is about nvidia SLI. You add Alienware to this thread just because
they have their own SLI. I did not know Alienware have this SLI but not the
same eventhough the result is the same.

nvidia use bridge or hub = use two 16X PCI-E
Alienware use cable = Mix slot.

All I hearing from your link to Alieanware, they seem to use one card on
PCI-E and other on a different slot and use VMH to do the SLI.

Alienware only demostrate that they can do SLI with mix slot instead of same
card with two 16X PCI-E like nvidia.
Alieanware method is propietary with patent pending. I am not sure if I can
purchase that board when release unless buy the whole PC.

[snip]
>
> Whether you will in fact be able to use an ATI card in the same system
with
> an nvidia I doubt that anybody knows at this point. Alienware may claim
> that you can but they haven't demonstrated that you can so assume until
you
> see it work that their marketing guys are exaggerating the flexibility of
> their system. And you certainly aren't going to be able to use an ATI
with
> an nvidia board using nvidia's system.
>
[/snip]
They say they can and then claim they can not. From the link you provided.

CapFusion,...
 
G

Guest

Guest
Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

CapFusion wrote:

>
> "J. Clarke" <jclarke@nospam.invalid> wrote in message
> news:ccvl0711av7@news4.newsguy.com...
> [snip]
>> In any case, the only difference between the Alienware VMH and nvidia's
> new
>> "SLI" is that nvidia puts the hub on a bridgeboard and does a digital
> merge
>> instead of analog.
> [/snip]
> This is where you and I have problem here, regarding SLI..
> First I did not mention anything about Alienware as an example of SLI w/2
> PCI-E slot until you indicating and point out that Alienware do support
> SLI. All I know about alienware regarding their SLI is like 3Dfx way of
> doing - using cable.
>
> I also note there is not any public mainboard will have two PCI-E slot yet
> include nvidia themself.

That's odd, newegg claims to have several in stock for immediate delivery.

> [snip]
>>
>> I have no idea what you mean by "Alienware do not have 16X
>> in-mind using two same card which need two 16X PCI-E slot.".
> [/snip]
> That quote is from you. From your old post awhile back.

Try again. That Pennsylvania Dutch sentence structure is yours, not mine.
Specifically from
<http://groups.google.com/groups?hl=en&lr=lang_en&ie=UTF-8&safe=off&threadm=Vr-dncOvx4-sq27dRVn-sw
40megapath.net&rnum=1&prev=/groups%3Fq%3D%2522Alienware%2Bdo%2Bnot%2Bhave
2B16X%2522%26hl%3Den%26lr%3Dlang_en%26ie%3DUTF-8%26safe%3Doff%26selm
3DVr-dncOvx4-sq27dRVn-sw%2540megapath.net%26rnum%3D1>

> [snip]
>> I don't see any contradiction. Alienware has demonstrated a product with
>> two nvidia boards Geforce FX 5900 PCI Express boards. The only thing
>> that we can be _certain_ of with regard to their product is that it works
>> well enough to be used for a demonstration at a computer show using those
>> boards.
>>
> [/snip]
> You referring to 6800 I assumed.

No, they demonstrated their product with 5900s.

> [snip]
>> The second post you link is in regard to the availability of slots.
>> There
>> is no chipset currently available that allows two x16 slots. That is a
>> fact, and does not contradict the fact that Alienware has demonstrated a
>> product.
>>
> [/snip]
> I did not say there was a board that have two PCI-E slot, did I? I only
> state that SLI need two 16X PCI-E slot according to nvidia. I thought this
> thread is about nvidia SLI. You add Alienware to this thread just because
> they have their own SLI. I did not know Alienware have this SLI but not
> the same eventhough the result is the same.
>
> nvidia use bridge or hub = use two 16X PCI-E
> Alienware use cable = Mix slot.

If there are no boards available then the whole discussion is moot. If you
can't buy the product then who CARES what benefits it provides?

> All I hearing from your link to Alieanware, they seem to use one card on
> PCI-E and other on a different slot and use VMH to do the SLI.

They use two boards in PCI-E slots. Read the material on their site--they
explain in some detail how it works.

> Alienware only demostrate that they can do SLI with mix slot instead of
> same card with two 16X PCI-E like nvidia.

Nvidia hasn't demonstrated _anything_.

> Alieanware method is propietary with patent pending. I am not sure if I
> can purchase that board when release unless buy the whole PC.

And nvidia's is not proprietary and they are not going to patent it?
Incidentally, if alienware's patent flies I suspect it will knock nvidia
out of the water.

> [snip]
>>
>> Whether you will in fact be able to use an ATI card in the same system
> with
>> an nvidia I doubt that anybody knows at this point. Alienware may claim
>> that you can but they haven't demonstrated that you can so assume until
> you
>> see it work that their marketing guys are exaggerating the flexibility of
>> their system. And you certainly aren't going to be able to use an ATI
> with
>> an nvidia board using nvidia's system.
>>
> [/snip]
> They say they can and then claim they can not. From the link you provided.

Look, if you are trying to make some point here would you please restate it,
as I can't figure out what you're on about anymore.


>
> CapFusion,...

--
--John
Reply to jclarke at ae tee tee global dot net
(was jclarke at eye bee em dot net)
 
G

Guest

Guest
Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

Yes..... you totally correct regarding SLI. I should have known better.
Thank for enlighten me regardling Alienware [SLI]. And while I am at it, I
will try to buy the correct fruit lemon instead of lime when I intend of
gettting lemon.
I will consider this thread as a lost clause for me.
The originail goal I believed I thought was to indicate SLI [nvidia version
not alienware]. I do not know why alieanware become to be the main focal
point. Yes, I know. Alieanware do have SLI as you mention before and now.
But this not regarding about alienware SLI. And yes. no mobo I know of at
this moment have two 16X PCI-E. I do not know if Intel or any othe chipset /
board maker will have one out [or anytime soon] soon.

The main idea regarding SLI for nvidia is to have the following:
2 16X PCI-E
2 same model PCI-E type card and best to have to from the same company.
With a Bridge to attach to both card. Not cable or any variation like 3Dfx.

If you still want alieanware include to this thread, just because it have
this SLI. By all mean, no one and especially me will stop you.
I do not see alieanware SLI the same as nvidia SLI. I believe you know the
reason.
Eventhough both SLI will get the same result.

Again, I believed this was my intend of this thread.

CapFusion,...


"J. Clarke" <jclarke@nospam.invalid> wrote in message
news:cd2b4a11cp4@news2.newsguy.com...
> CapFusion wrote:
> >>
> >> I have no idea what you mean by "Alienware do not have 16X
> >> in-mind using two same card which need two 16X PCI-E slot.".
> > [/snip]
> > That quote is from you. From your old post awhile back.
>
> Try again. That Pennsylvania Dutch sentence structure is yours, not mine.
> Specifically from
>
<http://groups.google.com/groups?hl=en&lr=lang_en&ie=UTF-8&safe=off&threadm=
Vr-dncOvx4-sq27dRVn-sw
> 40megapath.net&rnum=1&prev=/groups%3Fq%3D%2522Alienware%2Bdo%2Bnot%2Bhave
> 2B16X%2522%26hl%3Den%26lr%3Dlang_en%26ie%3DUTF-8%26safe%3Doff%26selm
> 3DVr-dncOvx4-sq27dRVn-sw%2540megapath.net%26rnum%3D1>
>
> > [snip]
> >> I don't see any contradiction. Alienware has demonstrated a product
with
> >> two nvidia boards Geforce FX 5900 PCI Express boards. The only thing
> >> that we can be _certain_ of with regard to their product is that it
works
> >> well enough to be used for a demonstration at a computer show using
those
> >> boards.
> >>
> > [/snip]
> > You referring to 6800 I assumed.
>
> No, they demonstrated their product with 5900s.
>
> > [snip]
> >> The second post you link is in regard to the availability of slots.
> >> There
> >> is no chipset currently available that allows two x16 slots. That is a
> >> fact, and does not contradict the fact that Alienware has demonstrated
a
> >> product.
> >>
> > [/snip]
> > I did not say there was a board that have two PCI-E slot, did I? I only
> > state that SLI need two 16X PCI-E slot according to nvidia. I thought
this
> > thread is about nvidia SLI. You add Alienware to this thread just
because
> > they have their own SLI. I did not know Alienware have this SLI but not
> > the same eventhough the result is the same.
> >
> > nvidia use bridge or hub = use two 16X PCI-E
> > Alienware use cable = Mix slot.
>
> If there are no boards available then the whole discussion is moot. If
you
> can't buy the product then who CARES what benefits it provides?
>
> > All I hearing from your link to Alieanware, they seem to use one card on
> > PCI-E and other on a different slot and use VMH to do the SLI.
>
> They use two boards in PCI-E slots. Read the material on their site--they
> explain in some detail how it works.
>
> > Alienware only demostrate that they can do SLI with mix slot instead of
> > same card with two 16X PCI-E like nvidia.
>
> Nvidia hasn't demonstrated _anything_.
>
> > Alieanware method is propietary with patent pending. I am not sure if I
> > can purchase that board when release unless buy the whole PC.
>
> And nvidia's is not proprietary and they are not going to patent it?
> Incidentally, if alienware's patent flies I suspect it will knock nvidia
> out of the water.
>
> > [snip]
> >>
> >> Whether you will in fact be able to use an ATI card in the same system
> > with
> >> an nvidia I doubt that anybody knows at this point. Alienware may
claim
> >> that you can but they haven't demonstrated that you can so assume until
> > you
> >> see it work that their marketing guys are exaggerating the flexibility
of
> >> their system. And you certainly aren't going to be able to use an ATI
> > with
> >> an nvidia board using nvidia's system.
> >>
> > [/snip]
> > They say they can and then claim they can not. From the link you
provided.
>
> Look, if you are trying to make some point here would you please restate
it,
> as I can't figure out what you're on about anymore.
>
>
> >
> > CapFusion,...
>
> --
> --John
> Reply to jclarke at ae tee tee global dot net
> (was jclarke at eye bee em dot net)
 
G

Guest

Guest
Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

CapFusion wrote:

> Yes..... you totally correct regarding SLI. I should have known better.
> Thank for enlighten me regardling Alienware [SLI]. And while I am at it, I
> will try to buy the correct fruit lemon instead of lime when I intend of
> gettting lemon.
> I will consider this thread as a lost clause for me.
> The originail goal I believed I thought was to indicate SLI [nvidia
> version not alienware]. I do not know why alieanware become to be the main
> focal point. Yes, I know. Alieanware do have SLI as you mention before and
> now. But this not regarding about alienware SLI. And yes. no mobo I know
> of at this moment have two 16X PCI-E. I do not know if Intel or any othe
> chipset / board maker will have one out [or anytime soon] soon.
>
> The main idea regarding SLI for nvidia is to have the following:
> 2 16X PCI-E
> 2 same model PCI-E type card and best to have to from the same company.
> With a Bridge to attach to both card. Not cable or any variation like
> 3Dfx.
>
> If you still want alieanware include to this thread, just because it have
> this SLI. By all mean, no one and especially me will stop you.
> I do not see alieanware SLI the same as nvidia SLI. I believe you know the
> reason.
> Eventhough both SLI will get the same result.
>
> Again, I believed this was my intend of this thread.

I guess we're clear then.

Personally my impression from reading what is on the Alienware site and what
is on the nvidia site is that both are doing basically the same thing in
slightly different ways. That being the case, I think that most of the
issues that affect the performance of one will affect the performance of
the other. They're both implementing partial-screen processing with a
combiner, alienware puts the combiner on a separate PCI board and connects
with a cable, nvidia does it with a bridgeboard and a feature connector,
alienware's combiner is analog, it's not clear what nvidia's will be, in
any case this is far downstream of the GPU, alienware claims that they'll
be able to mix GPUs and not need a special video driver, nvidia makes no
claims about mixing GPUs as far as I know and also claims no special
driver. The main thing is the partial-screen processing, assigning part of
the screen to one board and another part to the other.

However, it occurs to me that if the combiner is available then you can do
this with any kind of video boards, PCI, PCI-E, AGP, ISA, anything that XP
supports, just by setting it for dual monitors with appropriate resolutions
and feeding the signals into the combiner.

>
> CapFusion,...
>
>
> "J. Clarke" <jclarke@nospam.invalid> wrote in message
> news:cd2b4a11cp4@news2.newsguy.com...
>> CapFusion wrote:
>> >>
>> >> I have no idea what you mean by "Alienware do not have 16X
>> >> in-mind using two same card which need two 16X PCI-E slot.".
>> > [/snip]
>> > That quote is from you. From your old post awhile back.
>>
>> Try again. That Pennsylvania Dutch sentence structure is yours, not
>> mine. Specifically from
>>
>
<http://groups.google.com/groups?hl=en&lr=lang_en&ie=UTF-8&safe=off&threadm=
> Vr-dncOvx4-sq27dRVn-sw
>> 40megapath.net&rnum=1&prev=/groups%3Fq%3D%2522Alienware%2Bdo%2Bnot%2Bhave
>> 2B16X%2522%26hl%3Den%26lr%3Dlang_en%26ie%3DUTF-8%26safe%3Doff%26selm
>> 3DVr-dncOvx4-sq27dRVn-sw%2540megapath.net%26rnum%3D1>
>>
>> > [snip]
>> >> I don't see any contradiction. Alienware has demonstrated a product
> with
>> >> two nvidia boards Geforce FX 5900 PCI Express boards. The only thing
>> >> that we can be _certain_ of with regard to their product is that it
> works
>> >> well enough to be used for a demonstration at a computer show using
> those
>> >> boards.
>> >>
>> > [/snip]
>> > You referring to 6800 I assumed.
>>
>> No, they demonstrated their product with 5900s.
>>
>> > [snip]
>> >> The second post you link is in regard to the availability of slots.
>> >> There
>> >> is no chipset currently available that allows two x16 slots. That is
>> >> a fact, and does not contradict the fact that Alienware has
>> >> demonstrated
> a
>> >> product.
>> >>
>> > [/snip]
>> > I did not say there was a board that have two PCI-E slot, did I? I
>> > only state that SLI need two 16X PCI-E slot according to nvidia. I
>> > thought
> this
>> > thread is about nvidia SLI. You add Alienware to this thread just
> because
>> > they have their own SLI. I did not know Alienware have this SLI but not
>> > the same eventhough the result is the same.
>> >
>> > nvidia use bridge or hub = use two 16X PCI-E
>> > Alienware use cable = Mix slot.
>>
>> If there are no boards available then the whole discussion is moot. If
> you
>> can't buy the product then who CARES what benefits it provides?
>>
>> > All I hearing from your link to Alieanware, they seem to use one card
>> > on PCI-E and other on a different slot and use VMH to do the SLI.
>>
>> They use two boards in PCI-E slots. Read the material on their
>> site--they explain in some detail how it works.
>>
>> > Alienware only demostrate that they can do SLI with mix slot instead of
>> > same card with two 16X PCI-E like nvidia.
>>
>> Nvidia hasn't demonstrated _anything_.
>>
>> > Alieanware method is propietary with patent pending. I am not sure if I
>> > can purchase that board when release unless buy the whole PC.
>>
>> And nvidia's is not proprietary and they are not going to patent it?
>> Incidentally, if alienware's patent flies I suspect it will knock nvidia
>> out of the water.
>>
>> > [snip]
>> >>
>> >> Whether you will in fact be able to use an ATI card in the same system
>> > with
>> >> an nvidia I doubt that anybody knows at this point. Alienware may
> claim
>> >> that you can but they haven't demonstrated that you can so assume
>> >> until
>> > you
>> >> see it work that their marketing guys are exaggerating the flexibility
> of
>> >> their system. And you certainly aren't going to be able to use an ATI
>> > with
>> >> an nvidia board using nvidia's system.
>> >>
>> > [/snip]
>> > They say they can and then claim they can not. From the link you
> provided.
>>
>> Look, if you are trying to make some point here would you please restate
> it,
>> as I can't figure out what you're on about anymore.
>>
>>
>> >
>> > CapFusion,...
>>
>> --
>> --John
>> Reply to jclarke at ae tee tee global dot net
>> (was jclarke at eye bee em dot net)

--
--John
Reply to jclarke at ae tee tee global dot net
(was jclarke at eye bee em dot net)
 
G

Guest

Guest
Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

"J. Clarke" <jclarke@nospam.invalid> wrote in message
news:cd4ofv1s5v@news1.newsguy.com...
>
> I guess we're clear then.
>
> Personally my impression from reading what is on the Alienware site and
what
> is on the nvidia site is that both are doing basically the same thing in
> slightly different ways. That being the case, I think that most of the
> issues that affect the performance of one will affect the performance of
> the other. They're both implementing partial-screen processing with a
> combiner, alienware puts the combiner on a separate PCI board and connects
> with a cable, nvidia does it with a bridgeboard and a feature connector,
> alienware's combiner is analog, it's not clear what nvidia's will be, in
> any case this is far downstream of the GPU, alienware claims that they'll
> be able to mix GPUs and not need a special video driver, nvidia makes no
> claims about mixing GPUs as far as I know and also claims no special
> driver. The main thing is the partial-screen processing, assigning part
of
> the screen to one board and another part to the other.
>
> However, it occurs to me that if the combiner is available then you can do
> this with any kind of video boards, PCI, PCI-E, AGP, ISA, anything that XP
> supports, just by setting it for dual monitors with appropriate
resolutions
> and feeding the signals into the combiner.
>

I will end this thread as your last comment.

CapFusion,...