Intel or Nvidia? nForce 680i Challenges Intel P965 and 975X

pschmid

Distinguished
Dec 7, 2005
333
0
18,780
Motherboards with Nvidia's nForce 680i chipset are now hitting the shelves, so we decided it was time for a chipset battle. The goal? To find out what is the best solution for your Core 2 Duo system.
 

Bass_Ackwards

Distinguished
Apr 22, 2006
14
0
18,510
"Our recommendation is simple: Go with the nForce 680i SLI if you want the best platform for Core 2"

It looked to me like 680i lost most of the benchmarks...Even overclocked.

I don't know did I miss something?

Edit: Ok...not overclocked, just with faster memory.
 

NotAPimecone

Distinguished
Nov 18, 2006
299
0
18,780
I've also come across this

According to their research, lowering your multiplier on a 965 (and possibly other intel chipsets like 975 as well) may have a huge impact on the northbridge core clock. Lowering the multiplier basically overclocks the NBCC, by a factor of Default/Set - in the case of your article, 11/7 = 1.57.

I, for example, run an E6400 on a P965 board at 425 FSB trouble-free. I've been stable at 450, but didn't like the high temps. Apparently RobsX2 runs an E6400 at 3.7GHz on a P5B Deluxe (also P965). That's 462 FSB, just like your 680i.

The upshot is that P965 is excellent for OC'ing low-multiplier CPUs like the E6300 and E6400, but if you need to lower your multiplier to achieve a high FSB, it looks like the 680i is the way to go.
 

JPFPsycho

Distinguished
Dec 19, 2006
1
0
18,510
I guess I'm a little confused too. All the numbers argue for anything but the 680i but then you recommend it? If the recommendation is a bet on future FSB speed improvements (Bearlake) and overclocking potential then I don't think the article includes the required verbage to argue for this. For example the two 16x PCI Express slots with LinkBoost seems to be a plus for the 680i but its benefits weren't evaluated?

Nonetheless, glad to see the additional information comparing these three chipsets.
 

pancito

Distinguished
Jun 9, 2006
68
0
18,630
This review is bogus.

I didn't see why the 680 is better at anything. except pay the bills for nvidia.

Can you compare Price/Performance Ratio. I know 965/975 chipsets. and the conclusion i have gotten is.

975
Lower FSB 400-430 Limit
Higher Multi CPU needed (e6600, e6700 even X6800)
Faster because of tighter timings

965
Higher FSB 450-500 Limit
Lower Multi CPU is OK (e6300, e6400)
Timings are the same as 975 to some point after that it changes strap and you have to OC you FSB even more to get decent times.

Either way the real deal is to mix and match components. What your memory sticks can do in 1 chipset probably can't do it in the other.
 

warezme

Distinguished
Dec 18, 2006
2,450
56
19,890
just built a new computer round a Evga 680iSLI, Evga 8800GTX, Corsair Dominator 2GB, 150GB Raptor and 2 500GB Seagates, plus Fatal1ty FPS audio in a water cooled Lian LI 1000B:

I have my C2D E6400 running 3.6Ghz full throttle is only 34C because I put a DangerDen Water block using dual 120mm raidator and 1250 Enheim pump that pumps like no ones business.

I haven't even started doing serious overclocks and its already putting out serious number.

680i ROCKS
 

pancito

Distinguished
Jun 9, 2006
68
0
18,630
just built a new computer round a Evga 680iSLI, Evga 8800GTX, Corsair Dominator 2GB, 150GB Raptor and 2 500GB Seagates, plus Fatal1ty FPS audio in a water cooled Lian LI 1000B:

I have my C2D E6400 running 3.6Ghz full throttle is only 34C because I put a DangerDen Water block using dual 120mm raidator and 1250 Enheim pump that pumps like no ones business.

I haven't even started doing serious overclocks and its already putting out serious number.

680i ROCKS

Not everyone can spend $2000 worth of equipment in a second. run cpu benchmarks you are going to end up with a sour taste on your moth when you either spend double as the next guy for a MoBo and get the same results.

Tweaking on the other hand that is impressive, we P5W guys or P5B Dlx guys are needing that and not getting it in favor of the ROG boards.

Do a plain level bencharks of all 3 even add the 650 chipset and you will get minor changes from 1 to the other.
 

czepiel

Distinguished
Dec 1, 2006
32
0
18,530
This review is bogus.

I didn't see why the 680 is better at anything. except pay the bills for nvidia.

I dont see how you see this review as bogus.

It compares and contrasts the differences between the three chipsets fine imo. I mean, yes, the 680i may allow you to overclock more with more ease but I cant see the benchmarks being reliable comparing overclocked components on boards. There are more factors involved with a successful overclock than the individual parts involved, although they do have an impact.

While the P965 may have performed on par or better in most of the benchmarks in the review I felt that it was made clear that the 680i would be the board taking the cake if you wanted to cram the best components and wanted max overclocking capability.
 

pancito

Distinguished
Jun 9, 2006
68
0
18,630
I dont see how you see this review as bogus.

While the P965 may have performed on par or better in most of the benchmarks in the review I felt that it was made clear that the 680i would be the board taking the cake if you wanted to cram the best components and wanted max overclocking capability.

I Don't understan your POV. Doing benchmarks it's simple get generic stuff for all 3 board trying to find a midle ground. Memory, HDD, Video, CPU etc.

Get Stock level numbers. Post, Overclock get fastest stable 24/7 config, post results.

I have a P5W DH, with a 975x and i had a 6400 wich did 425x8, i bought a e6600 guess what im still getting 425x9, so that makes my limit around 425 24/7 stable.

You will hit a wall sooner or later with each and every board. this review even if the numbers don't show the big advantage says GET the 680 BOARD. Tell that to all those poor people that spent 400 for a Strikout Xtreme and getting not much real world performance gains from their "old" 965 or 975 board.
 

czepiel

Distinguished
Dec 1, 2006
32
0
18,530
I Don't understan your POV. Doing benchmarks it's simple get generic stuff for all 3 board trying to find a midle ground. Memory, HDD, Video, CPU etc.

Get Stock level numbers. Post, Overclock get fastest stable 24/7 config, post results.

I have a P5W DH, with a 975x and i had a 6400 wich did 425x8, i bought a e6600 guess what im still getting 425x9, so that makes my limit around 425 24/7 stable.

You will hit a wall sooner or later with each and every board. this review even if the numbers don't show the big advantage says GET the 680 BOARD. .

Yes, everyboard will certainly hit a wall sooner or later that is a given. The review is not telling everyone to go buy a board with the 680i chipset. It's saying that if you want every bell, whistle, and the ability to truly have complete control over each component of your board - then buy the 680i. The Intel chipsets just cannot compare in the number of options that is provided with the 680i, but you pay for it.

Tell that to all those poor people that spent 400 for a Strikout Xtreme and getting not much real world performance gains from their "old" 965 or 975 board.

If someone had upgraded from a 965 or 975 chipset to the 680i expecting increases of leaps and bounds in stock performance, they didnt do their research, and owe their disappointment to themselves.
 

justjc

Distinguished
Jun 5, 2006
235
0
18,680
I can hardly wait till the RD600 chipset enters the race, as it will be cheaper than the nVidia solution, stay cooler, and still have at least the same overclockability.

To bad ATi won't make more Intel chipsets ;)
 

czepiel

Distinguished
Dec 1, 2006
32
0
18,530
I can hardly wait till the RD600 chipset enters the race, as it will be cheaper than the nVidia solution, stay cooler, and still have at least the same overclockability.

Possibly. But you have to take into account that without the RD600 out there is no real direct competition for the 680i, which I'd imagine accounts for some of the price difference. I'm sure it'll come down in sync with the release, unless the RD600 is similarly priced.
 

SuperFly03

Distinguished
Dec 2, 2004
2,514
0
20,790
I can hardly wait till the RD600 chipset enters the race, as it will be cheaper than the nVidia solution, stay cooler, and still have at least the same overclockability.

Possibly. But you have to take into account that without the RD600 out there is no real direct competition for the 680i, which I'd imagine accounts for some of the price difference. I'm sure it'll come down in sync with the release, unless the RD600 is similarly priced.

RD600 Discussion

Basic gist of it is the RD600 turns out to be a really solid chipset but not the "King of all C2's" that we had hoped it to be.

So far only 2 RD600 boards are in the works, one by DFI and ASUS just announced theirs. Look for Dario's and Ninja's posts as they can explain things better than I.

Needless to say, it looks like there is no silver bullet for the C2D.... just a matter of personal preference and budget.
 

torque79

Distinguished
Jun 14, 2006
440
0
18,780
I'll never understand why some people will pay so much for a motherboard. my last one cost me $75 canadian, and believe me CS:S and BF2 run BEAUTIFULLY on it.

I'd be very interested to see an entire article written about performance/$ differences between different examples of each component (motherboard, cpu, gpu, RAM). It would be interesting to see how much % difference each of these components can make (roughly) independent of the others. I'm betting the motherboard would be the smallest % contributor to your gaming experience by FAR.
 

warezme

Distinguished
Dec 18, 2006
2,450
56
19,890
Tweaking on the other hand that is impressive, we P5W guys or P5B Dlx guys are needing that and not getting it in favor of the ROG boards.

I was just lauding the benefits of the 680i as a good overclocker, and as I state in the end, I haven't even begun to do serious tweaking, like FSB combos, maybe CPU multiplier and mem combos, mem timings, etc. Just because I bought heavy hitting equipment does not mean I'm to lazy or dont' know what I'm doing. I've been tweaking and overclocking since before it was even mainstream and was still called hacking.

The 680i is just a great tool if you can afford it for this purpose. Once Intel Quads get some competition from AMD and drop their prices, I will most likely kick the OC E6400 out and put that in.
 

balister

Distinguished
Sep 6, 2006
403
0
18,790
This review is bogus.

I didn't see why the 680 is better at anything. except pay the bills for nvidia.

I dont see how you see this review as bogus.

It compares and contrasts the differences between the three chipsets fine imo. I mean, yes, the 680i may allow you to overclock more with more ease but I cant see the benchmarks being reliable comparing overclocked components on boards. There are more factors involved with a successful overclock than the individual parts involved, although they do have an impact.

While the P965 may have performed on par or better in most of the benchmarks in the review I felt that it was made clear that the 680i would be the board taking the cake if you wanted to cram the best components and wanted max overclocking capability.

I think where some of us are looking at the article and going "huh?" is with the conclusions. The article writer is saying that the 680i is better, but the benchmarks do not bear that out. The 680i only won the memory benchmarks, but lost all others (either it barely lost or lost significantly) to the Intel chipsets. Now, while the 680i is going to do SLI much better and overclock better, not everyone is going to overclock or run SLI. The author of the article barely touches on this when he makes his recommendations to go with the 680i. The real conclusion that should have been written is this:
If you plan to do a range of overclocking or use SLI, then the 680i is the chipset to choose. If instead you plan to use a single video card and do moderate to no overclocking, the Intel 9x5 chipsets would be a better choice with the 965 being geared more to more extreme overclocking of lower multiplier Core2 processors and the 975 being geared to lesser overclocking of higher multiplier Core2 processors.
 

aramarc

Distinguished
Jul 3, 2006
11
0
18,510
Hmmm.... No mention of the well publicized 680i stability and corruption issues? I would think the article would at least address them a bit.
 

lbax

Distinguished
May 15, 2006
112
0
18,680
Add me to the list of people scratching their heads. The data presented in this review is not supportive of the conclusions. It appears that the conclusions were written before the data was pluged into the process....
I hope that this is a one time Holiday deadline glitch.
 

MarcelJV

Distinguished
May 30, 2006
26
0
18,530
Yes and this is the very board with all the Sata data corruption issues.

I love this line.

Be sure to reduce the processor multiplier before overclocking your system speed that much! At the default x11 multiplier of the Core 2 Extreme 2.93 GHz, the processor would have to run at over 5 GHz - which it cannot do.

Sure what ever you say. Check out this link where this very processor is running at 5.5 Ghz.

http://www.xtremesystems.org/forums/showthread.php?t=59753
scroll to the Intel Conroe section or follow this link.

http://www.iamxtreme.net/coolaler/WR/x6800/wr_1m_5526_9000S_cpuz

The first link will also verify that the P965 chipset is capable of 567.04 mhz FSB.

In all fairness though, one should follow the above statement when overclocking the FSB to find the max the board will support as your process may not support the resulting speed.

I found the article more then confusing in the layout, page 2 seemed to have nothing to do with the topic at hand. The conclusion is equally confusing as the 680i chipset seems to be only for the Extreme gamer who over clocks and who spends the money to make the board over clock well. While the recommendation for the chipset is not wrong the article should have supported it. If the author showed overclocking results and then used the included conclusion then you can say it was accurate, but the facts presented do not support the conclusion. I think the chipset has merit, well once the issues with the reference board are worked out, but at the current price point, I would suggest that the best results, either overclocking or in stock performance, goes to P965.

The 680i chipset gives the overclocking result that the Nvidia users were hoping to find in the 590 chipset. So we finally have a competative chipset from Nvidia assuming you have the cash to pay for it.

Also the information about Intel Matrix Raid technology is wrong. In reference to intel Raid technology we get the usual incorrect 0,1, 0+1, 5 when the P965/975 website show this text.

With a 2nd hard drive added, provides quicker access to digital photo, video and data files with RAID 0, 5, and 10, and greater data protection against a hard disk drive failure with RAID 1, 5, and 10.

This is nothing new as every where you see this shown incorrectly.
 

Eurasianman

Distinguished
Jul 20, 2006
883
0
19,010
Sorry, not impressed.

As someone else once stated, "benchmarks are just numbers and bragging rights".

IMHO, it's up to the user to decide if the product is the best or not. For example, for me, the 975x chipset is the best I've ever worked with. But perhaps for someone else, it was a nightmare. I love how they write reviews on CPUs and motherboards and rarely ever mention to the reader that every board is different regardless that they're the same model. My motherboard has a lower wall than others. Then again, maybe it's the CPU? Who the h3ll knows? Who should care as long as everything works good enough for them?

My point is, IMO, this review was not very well written.
 

WeaponX

Distinguished
Dec 20, 2006
3
0
18,510
The NVIDIA chipset clearly offers 2 lanes at 16x, but the video card in the review was an ATI. Most of the benchmarks were the same, so I don't see much advantage to going with NVIDIA. I want to see benchmarks from each board with SLI.
 

lbax

Distinguished
May 15, 2006
112
0
18,680
I think some of our readers should consider trying their hand at an article and submitting it for consideration. ;)

I assume that most of us would do a very poor job of it. However, it appears that a significant number of readers of this article recognize that it does not meet the historical quality standard that we have grown to expect from Tom’s. Remember that most of us do not purport to be professional technical journalists. However, we can & will judge a review or article by our standards.

Furthermore, if we withhold positive or negative feedback, we fail to contribute to the improvement of the review process.