Nvidia went SLI because they can't compete?

[WaltC]

It's also expensive as hell,

OK

Hmmm....about the heat...


Ugh, no.


I'm beginning to wonder if I'm the only one who read the text...

This article has been a long time in the making with numerous false starts due to software and platform issues. We've been through numerous different SLI connectors, platform BIOS's, platform and graphics drivers, and even different graphics boards in the case of the 6800 Ultra - initially there were often stability issues and frequent crashes. However, as the platform has matured a little and various updates have come about things have begun to become much more cohesive and hopefully products that are in the retail channels now will already be making use of these updates so that end users shouldn't experience quite as many issues - that said, the heat issues we suffered with the 6600 GT's in SLI mode is something that may need to be watched out for and plenty of airflow through the case may be useful.

And...

All the boards on test here are high performance boards on their own in the first place and as such will have fairly large power draws and have a fairly large heat output - doubling will obviously amplify those needs.

And...

The 6600 GT, however, has by far the highest temperatures - although the 6600 GT has fewer pipelines than the other chips it is running at a much higher clockspeed, and the MSI board is running overclocked from the reference speed which may be producing more heat than usual - what we did notice in this instance was that with continued use in this enclosed case rendering issues occurred in the SLI configuration which appear to be heat related as they would disappear with the case side off.

So the answer here appears to be "OK," too, doesn't it?

...louder 'n hell...

Again, no. Jesus...

And again...

Again as we would expect, using two boards increases the noise levels in relation to just one, with the levels going up between 1% and 5% in SLI mode; however, bear in mind that actual noise levels are not increased linearly and is a logarithmic scale in terms of audibility. Not unexpectedly the 6800 Ultra's in SLI are the noisiest, and they are very audible in SLI mode - also note that they there is no increase in noise level when they are used in SLI as for some reason as soon as they are coupled together both of the 6800 Ultra's fans appear to spin up to full speed without any variation so in this configuration they are constantly noisy. Again the 6600 GT's are producing some surprising results in this area being that they are challenging the 6800 GT's in noise output in SLI mode, but again this may relate to their clockspeeds and temperature output and the fans having to spin fast in order to cool them effectively.

Looks like another "OK," to me. (Come on--can't be that hard to actually *read* the review, can it?)

 

[trinibwoy]

What's up with the specific reference to the 6600GT heat levels - it is obviously anomalous since the Ultras are coming in lower.

Do any SLI boards currently have enough width between slots to support two cards with silencers? Dave did say that the Asus board had more width between slots than usual.

 

[digitalwanderer]

Do any SLI boards currently have enough width between slots to support two cards with silencers? Dave did say that the Asus board had more width between slots than usual.

There were some pictures up at HEXUS the other day of a dual 6800 Ultra set-up with coolers on 'em that looked bigger than silencers.

They looked like each card took up 2-1/2 slots, so yeah I think there will be some mobos designed for some serious cooling in mind.

 

[WaltC]

What's up with the specific reference to the 6600GT heat levels - it is obviously anomalous since the Ultras are coming in lower.

This is the last time I will baby sit...

I'll quote this again and I encourage you to read it this time as it seems to me it answers your question on the nose...

The 6600 GT, however, has by far the highest temperatures - although the 6600 GT has fewer pipelines than the other chips it is running at a much higher clockspeed, and the MSI board is running overclocked from the reference speed which may be producing more heat than usual - what we did notice in this instance was that with continued use in this enclosed case rendering issues occurred in the SLI configuration which appear to be heat related as they would disappear with the case side off.

That seems to illustrate what is happening rather well, doesn't it? At its higher clockspeed it is producing more heat, is what it says to me.

Do any SLI boards currently have enough width between slots to support two cards with silencers? Dave did say that the Asus board had more width between slots than usual.

Well, I suppose that spending even *more money* might be one way to address the noise issue--but, that's a matter of *how many hoops* the customer is willing to jump through, isn't it?

 

[trinibwoy]

Now this is the kind of system SLI was meant for. Dell can keep their single card setups

 

[trinibwoy]

Well, I suppose that spending even *more money* might be one way to address the noise issue--but, that's a matter of *how many hoops* the customer is willing to jump through, isn't it?

One man's hoop is another man's hobby. Just admit that you're an old man that can't bother will this newfangled stuff

Regarding the 6600GT's being so hot, shouldn't the Ultra's and GT's have similar heat output since their transistor counts are a lot higher. I recall the 9600XT having much higher clocks than the 9800PRO - did it also show higher heat output?

 

[WaltC]

Now this is the kind of system SLI was meant for. Dell can keep their single card setups

Heh-Heh... I've got Mickey D's email right here...tell you what, would you mind if I forwarded your post and comment on to him and signed it like this:

"Love and kisses, Mickey baby,

JHH and the whole nVidia gang"....!

(Nah--I won't do that...)

 

[WaltC]

One man's hoop is another man's hobby. Just admit that you're an old man that can't bother will this newfangled stuff

Regarding the 6600GT's being so hot, shouldn't the Ultra's and GT's have similar heat output since their transistor counts are a lot higher. I recall the 9600XT having much higher clocks than the 9800PRO - did it also show higher heat output?

Well, it *is* true that I'm an ancient crustacean ('scuse me while I scrape off a barnacle or two), but I have to admit that I haven't quite figured out the appeal of listening to a gurgling water pump just yet... Not to mention going into anaphylactic shock upon smelling burning plastic while suddenly realizing I can't hear the gurgling anymore!... I know, I must be missing *something* however, so I'll keep it in mind...

As to your question--it all depends upon the chip. If it's being overvolted and overclocked (past its design specs--which only nVidia would know)--then you might expect such a chip to run much hotter than it normally would if it was run at its intended voltages and/or clocks. This threshold would be different for each company's chips so what holds true for one company's chips would not necessarily apply to another's at all. Just guessing, is all.

I can't recall at the moment--is nV into low-k yet, or not?

 

[digitalwanderer]

I can't recall at the moment--is nV into low-k yet, or not?

You're mean!

 

[Geeforcer]

It's also expensive as hell,

OK

Hmmm....about the heat...


Ugh, no.


I'm beginning to wonder if I'm the only one who read the text...

This article has been a long time in the making with numerous false starts due to software and platform issues. We've been through numerous different SLI connectors, platform BIOS's, platform and graphics drivers, and even different graphics boards in the case of the 6800 Ultra - initially there were often stability issues and frequent crashes. However, as the platform has matured a little and various updates have come about things have begun to become much more cohesive and hopefully products that are in the retail channels now will already be making use of these updates so that end users shouldn't experience quite as many issues - that said, the heat issues we suffered with the 6600 GT's in SLI mode is something that may need to be watched out for and plenty of airflow through the case may be useful.

And...

All the boards on test here are high performance boards on their own in the first place and as such will have fairly large power draws and have a fairly large heat output - doubling will obviously amplify those needs.

And...

The 6600 GT, however, has by far the highest temperatures - although the 6600 GT has fewer pipelines than the other chips it is running at a much higher clockspeed, and the MSI board is running overclocked from the reference speed which may be producing more heat than usual - what we did notice in this instance was that with continued use in this enclosed case rendering issues occurred in the SLI configuration which appear to be heat related as they would disappear with the case side off.

So the answer here appears to be "OK," too, doesn't it?

No, of course it doesn't. All of them reference an anomaly with 6600 GT, while the rest of the cart blows your assertion regarding "hotter then hell" out of the water, with SLI configurations only a few degrees higher then non-SLI systems. In fact, the temperature inside that particular test case when equipped with 6800 Ultra SLI was less then one measured using a single X600 XT board. I guess X600XT is "hotter then hell too"? It's ridiculous that anyone would overlook the fact that A) SLI configuration, as tested by Dave, cause only a small temperature increase as opposed to a single card and B) the ambient temperature inside the system with 2 6800 Ultras was lower then in a different system with a single X600. Instead of simply admitting that your accretion doesn't seem to bear out, you latch onto 6600GT results, which are out of line with the rest of the field. Where are you going with this?

...louder 'n hell...

Again, no. Jesus...

And again...

Again as we would expect, using two boards increases the noise levels in relation to just one, with the levels going up between 1% and 5% in SLI mode; however, bear in mind that actual noise levels are not increased linearly and is a logarithmic scale in terms of audibility. Not unexpectedly the 6800 Ultra's in SLI are the noisiest, and they are very audible in SLI mode - also note that they there is no increase in noise level when they are used in SLI as for some reason as soon as they are coupled together both of the 6800 Ultra's fans appear to spin up to full speed without any variation so in this configuration they are constantly noisy. Again the 6600 GT's are producing some surprising results in this area being that they are challenging the 6800 GT's in noise output in SLI mode, but again this may relate to their clockspeeds and temperature output and the fans having to spin fast in order to cool them effectively.

Looks like another "OK," to me. (Come on--can't be that hard to actually *read* the review, can it?)

Look at the graph. The difference between single and double 6800 Ultra at full load in one decibel, indistinguishable for all intends and purposes. Are you tying to dispute that?

 

[lopri]

Remember this though, if someone wants to buy an ATI or Nvidia dual solution, they won't give a rat's ar$e what you say. Why? Well, just ask anyone who has an SLI system.

You have no clue about how respected Walt's opinion is around here then.

Sig worthy

 

[lopri]

Can I go ahead and make it as my sig? I thought i'd ask both of you.

I'm a nice person.

lop

 

[lopri]

No, actually it's not clear. You cannot drive two cars at once, however you can use two videocards at once. If you're going to draw an analogy, please make a passing attempt for it to actually be, oh... I don't know... analogous.

This made my day. Thanks flf for a good laugh.

lop

 

[ondaedg]

Remember this though, if someone wants to buy an ATI or Nvidia dual solution, they won't give a rat's ar$e what you say. Why? Well, just ask anyone who has an SLI system.

You have no clue about how respected Walt's opinion is around here then.

Buh bye.

Smartest thing you've posted at this site.

Are you kidding me?

Lol!

Dig, you get someone who will stand in your corner to obsess side by side with you and you feel free enough to make a statement like that. Amazing stuff man. Amazing.

 

[ondaedg]

Is there a competition to say absolutely nothing in 20,000 words or more, or is Walt trying to set some type of record?

I have not read one thing of his that was either true or made sense... Someone help me here.

I am still trying to understand his dual hummer H2 to SLI analogy.....

 

[ondaedg]

Anyone understand how Walt can turn that cost analogy into a comparison of 2 Hummer H2s? Once again Walt, you have lost me. Congratulations.

Thank you... I do confess that I didn't intend to lose you, so I'll try again...

nV SLI = How many nV 3d cards? (be quick now...)

A) One 3d card
B) Two 3d cards

Hummer purchase = how many Hummers? One Hummer, or two? (Heh...)

Cleared up for you?...

No, actually it's not clear. You cannot drive two cars at once, however you can use two videocards at once. If you're going to draw an analogy, please make a passing attempt for it to actually be, oh... I don't know... analogous.

Ok so I wasn't the only who didn't understand his analogy. Thank you!

 

[WaltC]

No, of course it doesn't. All of them reference an anomaly with 6600 GT, while the rest of the cart blows your assertion regarding "hotter then hell" out of the water, with SLI configurations only a few degrees higher then non-SLI systems. In fact, the temperature inside that particular test case when equipped with 6800 Ultra SLI was less then one measured using a single X600 XT board. I guess X600XT is "hotter then hell too"? It's ridiculous that anyone would overlook the fact that A) SLI configuration, as tested by Dave, cause only a small temperature increase as opposed to a single card and B) the ambient temperature inside the system with 2 6800 Ultras was lower then in a different system with a single X600. Instead of simply admitting that your accretion doesn't seem to bear out, you latch onto 6600GT results, which are out of line with the rest of the field. Where are you going with this?

You're funny--test results which you don't like you call "anomalies"--and of course the test results which support your pre-determined point of view I would imagine you'd call "representative." Sorry, but I missed the part of the review in which Dave said the heat issues were mere "anomalies" which we could all pretend didn't exist in reality and which we should ignore. Mind quoting where he said that as I surely didn't read it?...:/

BTW, when you have to remove the side cover to get the system to render without visible anamolies--that means it's hotter 'n hell... Maybe you disagree but that's my opinion.

Also, it may have escaped your attention but I said SLI is hotter 'n hell--I never said any specific single card in normal operation is "hotter 'n hell", did I? I mean, why should I have said such a thing since the review was about SLI and absolutely nothing else, right?

Look at the graph. The difference between single and double 6800 Ultra at full load in one decibel, indistinguishable for all intends and purposes. Are you tying to dispute that?

OK, again we have you saying that when Dave said "logarithmic scale" and "so in this configuration [SLI] they are constantly noisy" and "Again the 6600 GT's are producing some surprising results in this area being that they are challenging the 6800 GT's in noise output in SLI mode, but again this may relate to their clockspeeds and temperature output and the fans having to spin fast in order to cool them effectively" he wasn't serious--so I guess he was just joking?

Tell you what--while you read the graph and ignore what Dave wrote in the review about the graph, I'll read the graph and then I'll read what Dave wrote about *the numbers he put into the graph.* I'm surprised you didn't call these "anomalies" as well...

How in the name of all that is rational might you conclude that *TWO* 6800GT's running their fans at full speed might only be as loud as ONE 6800GT running its fan at full speed...? Heh... The human capacity for self-deceit never ceases to amuse and amaze, and of course what you say about Dave's graph and what Dave says about Dave's graph simply do not match, do they? Since Dave did the graph and the testing I'd suggest you substitute his interpretation of the numbers in it for your own...

 

[WaltC]

Ok so I wasn't the only who didn't understand his analogy. Thank you!

Heh... Yes, and not only did you not understand the analogy, you also don't understand that it wasn't my analogy to begin with. But you have a point, as there doesn't seem to be very much in this thread which you do understand, unfortunately.

Wow... I sure stepped on some sensitive SLI toes here, didn't I? Sorry, guys, that SLI (regardless of who makes it) doesn't exactly trip my trigger, but that's just the way it is in my case. Just because I don't care for it doesn't mean you can't fall all over it, right? By all means dig into SLI if you want--you have my official symp--er, blessings...

 

[WaltC]

I can't recall at the moment--is nV into low-k yet, or not?

You're mean!

Heh... Unintended, really.

He did ask about the temperature differentials as he perceives what they ought to be, right? It's a decent hypothesis that might answer his question, I think. As I said, on my part it's just a guess.

 

[Sxotty]

Umm edit? 3 posts ina row, you go!