More Than Just a 'Paper Launch': 680i Motherboard Comparison, Part 1

[pschmid]

Over a month has passed since Nvidia announced the official release of its long-awaited 680i chipset to support Intel's Core 2 desktop processors. Part 1 of our comparison features a few early motherboard examples, as we continue to wait for the release of other products.

 

[djgandy]

The 680i got its biggest performance boost in games

A 2% fps increase over a Foxconn board that probably costs under half the price. Also since this is a gamers board, why such a lack of game testing?

Only reason i can see for buying this chipset is SLI.

 

[MarcelJV]

Agreed. If you need or use SLI this is really your only choice and if you use 8800GTX you need full 16x support on both PCIe16x slots for full use of the card.

For Xfire either 975 or RD600 (See DFI board) is you best performance choice.

For everyone else your best bet is 965 from performance/cost point of view.

This article was better than yesterdays article. I would have liked to see 965 chipset included in the tests though.

 

[lbax]

This article was better than yesterdays article. I would have liked to see 965 chipset included in the tests though.

I concur. The conclusions are supported by the data presented within the article, unlike yesterdays...

Yes, I would have liked to see best of breed 965 & 975 numbers as a basline for comparision.

 

[LiquidFlame]

whey didn't you guys include the eVGA 680i Motherboard

 

[little_scrapper]

whey didn't you guys include the eVGA 680i Motherboard

That was exactly the question I was going to ask. Especially since EVGA seems to be outselling everyone by a HUGE margin.

 

[Crashman]

EVGA and BFG use the same board.

 

[Baer]

Also eVGA is the most available board,
There is one more reason for choosing the S E and that is the solid capacitors. If you are looking for long life under more extreme conditions, especially for boards that are manufactured under RoHS specs (which reduce reliability) then the Striker will give you that reliability because of the more expensive but far more reliabile solid Caps.

 

[needless_kane]

I agree with the others who replied. The EVGA board has been out for about a month now and should have been included in this review. Was this board left out because of the problems? Seems EVGA is close to getting the board working with their P23 Bios.

Anybody know if Gigabyte is working on a 680i mobo? I've been an asus user myself, but with the cost of the 680i asus boards, I'm looking elsewheres for a board.

 

[Crashman]

Gigabyte announced its board a few weeks ago and it should be out the first week of Janaury.

The EVGA board appeared in Patrick and Achim's review and probably couldn't get to the reviewer in time. The BFG board tested is the same board!

 

[azrael-]

Two interesting questions (at least I think so) for the upcoming 680i review:

a) What would happen if one would use the NAM (firewall) from the nForce4 driver sets? I've heard rumors that the "hardware" is still in place, even though nVidia no longer advertises and supports the firewall feature.

b) The C55 northbridge apparently supports ECC memory and is supposed to be used in later nForce professional chipsets. Is this true?

Also, I'm quite interested how the ASUS P5N32-E SLI board will perform. A 680i board at about the price of the P5W DH Deluxe (at least around here). That's about 2/3 of the asking price for the Striker and most reference 680i boards. Can't be all bad, can it?

 

[yyrkoon]

Cooler Master's Hyper TX offered a significant heat reduction compared to our most recent previous game platform shootout.

Here, I think you could do without the "A", but otherwise, a very well written review. Glad to see you guys are making improvement on spelling/grammar, and if your native language is not English, its getting harder to tell otherwise

Glad to see someone doing a decent "preview" of the Abit IN9 32X-MAX, as it is the board that has been getting most of my most attention recently ( concerning researching new parts)

Keep up the good work.

 

[MarcelJV]

While you are checking out the Asus p5n32-E SLI you might want to review this link: http://www.anandtech.com/mb/showdoc.aspx?i=2894

This is the 650i chipset runing faster then the 680i in games and that includes SLI, even with 8800GTX in SLI!

 

[azrael-]

Have to admit that ASUS P5N-E SLI looks pretty interesting. Seems like it has everything you *really* need and a great price to match.

 

[MarcelJV]

If you have SLI then it might be the way to go. Hard to say at this point as we have no idea if DX10 games will cause the x8 bandwidth to be saturated and thus limit frame rates (in SLI mode). If they do not then this is the better price/performance board. Given that current games and GPUs can not saturate the x8 bandwidth (either PCIe or AGP) there really seems to not be a reason to go spend the extra cash for the x16 lanes.

I do not have SLI so I will not be looking at this board but am waiting for the 650i Ultra to see how it gets along over all. Currently I think the 965 chipset is the price/performance winner and this review has done nothing to change my mind. Mostly because 965 chipset was not covered, but the prior article did include it and it look to stay very close to 975 and 680i that it wins based on the price of the boards.

 

[Slayer616]

In the article you refrence "Nvidia reccomended settings" several times. Is there any link to where we can find this info or can you post it here??

I own a BFG board and the default "auto" settings on voltages in the BIOS come up as red (too high?) in Ntune. I have manually throttled them back to the green and the system seems to run stable, but I have no idea what is considred safe for the chipset/FSB etc.

 

[Crashman]

In the article you refrence "Nvidia reccomended settings" several times. Is there any link to where we can find this info or can you post it here??

http://www.nvidia.com/page/nforce_600i_tech_briefs.html

 

[Slayer616]

Thanks! Good stuff.

 

[clue69less]

Those benchmark charts are about as exciting as watching grass grow.

 

[Crashman]

Those benchmark charts are about as exciting as watching grass grow.

Suggestions? Perhaps packman style, with the winner eating the losers?

 

[clue69less]

Those benchmark charts are about as exciting as watching grass grow.

Suggestions? Perhaps packman style, with the winner eating the losers?

What I was getting at is the last plot shows the Asus with the most wins. It is the only plot that shows any real contrast in the data. Take the Sandra 2005 Arithmatic bench as an example. The Asus had the top score at 11938, but the bottom of the heap score by the Foxconn was 11812 - just a percent below the "winner". The Sandra Multimedia range was from 79160 for the Asus down to 78950 for the Foxconn. That's only 3 tenths of a percent range across seven mobo/RAM combos! In terms of real world performance, these benches looked essentially identical to my eyes.

I do realize that the memory bandwidth data has more spread but still, if I was trying to choose one of these boards, it would come down to the price to features ratio because there's not enough performance spread to stand out. I'd have to see a more exhaustive OC comparison before I could make heads or tails out of what these boards will do. The reviewer said that the Striker demolished the competition but I'd reserve that comparitor for the current Intel/AMD situation. Don't get me wrong, I'm glad Tom's put the data up - I would just have colored the conclusions very differently.

 

[max_normal]

100% agree, these results are in actual fact completely meaningless, to show any viable data they would need to be repeated a number of times (preferably on multiple different copies of the same mobo for each sample) the mean calculated, and subjected to a standard error test with error bars. Any alleged difference outside these error bars then needs to be proved by using a t-test to determine the p-value (probability of the alleged result occuring by chance).

Data this close that has not been analysed this way should be treated with a massive pinch of salt and ignored. Unless the difference between any test subjects in huge, there is no real evidence that there is any difference between the boards whatsoever. Buy whichever takes your fancy, the chances are that in your home rig it will be perform completely differently than the pointless one-off tests shown by the reviewers.

 

[Preecher]

Just some info I figured I would share. I read on a couple posts that motherboards using heatpipes may not cool properly if mounted inverted (some cases have the psu on the bottom and thus the motherboard is orientied upside down).

According to Asus tech support:

[Problem Description]
Can the Striker Extreme (680i chipset) be mounted "INVERTED" like in a Lani-Li Case?
Will the heatpipes cool correctly if inverted?

Response:

Dear ASUS Users:
Your query has been replied by ASUS technical personnel. The content of the reply is shown below. Time replied at : 2007-01-03 23:23:48

Subject Date
Re: Motherboard Striker Extreme 2007-01-04 02:16:27

this will cause the motherboard to overheat

Apparently the orientation is importaint for the heat pipes to function correctly.

Cheers!

 

[Phrozt]

Pretty good review... at least to test the waters for the 680i boards...

What I'd like to know is where that January review mentioned at the beginning of the article is :/.

 

[Crashman]

Heheh, all the boards got delayed. A better question is, where's that board with 4 PCI-Express X16 slots and a SoundBlaster X-Fi chip MSI announced back in November?