680i Rejecting Quad-Core Penryns?

[Athildja-x64]

The eVGA board did not change the chipset, they changed traces on the board as well as other components the 680i chipset itself was and is a failure,
to state otherwise is to ignore facts.
"The EVGA AR version of the 680i had some problems with Quad core when Quads first came out. Then EVGA came out with the A1 version of the 680i which cured the problems. The perspective here is that the Quads worked perfectly fine on the AR board, you just didn't have overclocking head room. Even then, EVGA made the updated boards completely free and accessible to the customers who had purchased the AR boards. Even shipping both ways was covered and you didn't have to cross ship either. You got the new mobo and had 14 days to ship the AR back to them."

This is a crock,
I pointed out in a review of this board that when running a quad at stock speed the system slowed down due to overhead to the NB memory controller.
I also proved that the SATA controller had issues running SATA 150 drives especially Raptors.
Next in line was the issue with overvolting the memory and CPU randomly.
after that we have issues with the PATA contoller and certain (NEC) DVD-RWs
should I add more?
Also not everyone that bought this board got the new one for free,
I know people that bought them OEM and eVGA refused to replace them
offering only a BIOS flash
after the BIOS flash they reported that they could run their quads but only if they dissabled 2 cores.

To call me ignorant is also rather rude and ignorant yourself as you know nothing of me or my experience with this chipset.

[Loosecannon]

Originally Posted by Athildja-x64

should I add more?
.

Many 680i boards burned up memory too

[Athildja-x64]

"the 975 being one of the biggest disappointments in some time. "
I think you mean 965 as the 975 is one of the best selling and most compatible boards out.
It had full Crossfire support and worked with the 65nm CPUs right out of the gate on most boards requiring nothing but a BIOS flash for the others,
Interestingly enough.
While my Kentsfiled has nothing but problems in both of my 680i boards, it is running very happily in my P5W-DW 975 based mainboard with 6GB of memory.

[RA1D]

I luckily have not experienced the problems you've encountered with the 680i chipset. Maybe that's because I bought my board this past summer, awhile after the initial release of the chipset. Whatever it is, I believe there are positives and negatives with every chipset that comes out and everyone will have different experiences.

[Quakindude]

Originally Posted by Athildja-x64

The eVGA board did not change the chipset, they changed traces on the board as well as other components the 680i chipset itself was and is a failure,
to state otherwise is to ignore facts.

I didn't say the chipset had changed. You ignored what I actually posted.

Originally Posted by Athildja-x64

This is a crock,
I pointed out in a review of this board that when running a quad at stock speed the system slowed down due to overhead to the NB memory controller.

But yet, enthusiast communities the Internet over use the 680i as their test bed.

Originally Posted by Athildja-x64

I also proved that the SATA controller had issues running SATA 150 drives especially Raptors.

Again, there's plenty of people running Raptors in a RAID configuration on this chipset. Without issues. Show me a series of different review sites coming up with the same issue and then I'll see it as systemic. Until then, I'll continue to believe that my experiences with this board have been those of the typical user, and that your testing isn't indicative of the whole product line, but atypical of it.

Originally Posted by Athildja-x64

Next in line was the issue with overvolting the memory and CPU randomly.
after that we have issues with the PATA contoller and certain (NEC) DVD-RWs
should I add more?

There were memory issues with the 680i and running memory at 1066MHz and voltages greater than 2.2V. There's no doubt about that. But it was never determined that the 680i series "burned up memory" because of the chipset. That was determined to have been a problem with the memory itself not being able to handle voltages greater than 2.3V AFAIK. Personally, I've had a kit of DDR2-8500 running at 1150MHz and 2.2V for over six months without issue.

But over-volting the CPU and Ram on its own? If anything, the voltage droop is notable, not over-volting the CPU. Unless you were running an ECS 650/680i, and their BIOS dicked up the voltages so badly on anything I've ever used from them as to be laughable, over-volting hasn't been a problem resulting from the Nvidia chipset that I'm aware of. As a matter of fact, you're the first person I've heard mention this. It warrants more research if that is the case. My experiences have indicated on EVGA boards that even though there is voltage droop, the board is rather stable, fluctuating only slightly under load. The ECS boards I've looked over have shown voltage fluctuations ranging from .2 to 1.2V. So if you're intent is to run the CPU at 1.4V, and you take into account voltage droop and set the CPU Vcore to 1.5, you could get up to 1.6V+. On an ECS board.

Originally Posted by Athildja-x64

Also not everyone that bought this board got the new one for free,
I know people that bought them OEM and eVGA refused to replace them
offering only a BIOS flash
after the BIOS flash they reported that they could run their quads but only if they dissabled 2 cores.

Yet another great reason to buy the retail product. Ever experience Intel's customer service on an OEM CPU that has died? Less than laudable I can assure you. Also, if these folks bought the 680i SLI mobo when it first came out, the AR full more 680i, then they can still get that mobo replaced when they upgrade to a quad core. I swapped mine out with EVGA just last week. The EVGA forums are simply covered with my same experience.

Originally Posted by Athildja-x64

To call me ignorant is also rather rude and ignorant yourself as you know nothing of me or my experience with this chipset.

Actually, I was referring to the person who called the 680i's execution a failure, as indicated by what I actually wrote.

The Nvidia chipset being less than stellar in the RAID performance department is well documented. You can find posts by yours truly recommending that if RAID performance is your primary goal, to look elsewhere than the 680i. I call the board like I tested it, use it and know it.

Personally, I think calling a whole product line "Flawed" is irresponsible. There are certainly some quirks that come along with being a 680i owner, but "flawed" hasn't been one of them. There's been some all encompassing remarks made concerning the 680i chipset that, quite frankly, can't be supported by the masses of people who have this board in operation every day.

However, as with any popular motherboard, you will find those that have had a ton of issues. I respect their opinions and their experiences. I detest ASUS motherboards not due to their performance or reliability, but due to their historically horrid customer service. My mind can be changed though just as ECS can change my mind about their voltages and BIOS not being as good as they should be.

Didn't mean to piss you off by using the word ignorant. Actually, in thinking of your comments, I thought you were fairly closed minded about the chipset. Not ignorant. But I can see how you came to that impression and I apologize for not being clearer about it.

[Athildja-x64]

Asus also waited until the first group of 680i's were out and heavily tuned the boards to overcome some (but not all) of the limitations of the 680i

[Athildja-x64]

"Actually, I was referring to the person who called the 680i's execution a failure, as indicated by what I actually wrote."

You quoted me in the post where you said calling the chipset a filure was ignorant.

If you look over the Xtremesystem forums and the nVidia forums you will see post after post about all of the issues I have listed.
and the memory problems were shown to be with memory at stock voltages.
people would run them and have stability issues and then the system would not post.
when they would test the RAM out side of the 680i the would find it no longer worked there either.

When I tested the board I found that if you ran any heavily multi threaded application you would see serious performance losses versus other chipsets.
this was with the Quads and the Core 2 Duo

[Quakindude]

Well, i won't argue with you. Your experiences are just that.

There are many highly reputable organizations and publications that utilize the 680i for good reason. It has matured into an excellent enthusiast level motherboard with a dedicated support team. The 975 is known as a great CPU overclocker (a very well deserved reputation), but with some limitations on the RAM overclockability.

My point being, people like us who have a higher than average working knowledge of computer components owe it to the general population to be as open minded and fair as possible when dispensing information about computers. I think your statements about the 680i aren't representative of the actual experiences most people have had and continue to have. I do not question your experiences in the least though.

I've had an absolutely hard core bitch of a time with two different P35 motherboards using a Q6600, DDR2-8500 Dominator Ram and an Nvidia Video card. But I sure as hell won't be on here declaring the P35 chipset is flawed or a failure. There's too many others out there having a great time with that chipset. So I see my experiences with the P35 as indicative of an atypical experience.

[Athildja-x64]

Actually it can be said that the P35 has some issues with overclocking RAM.
Especially with a Quad.
this is mostly an issue with the board makers though as they are the ones tht design the trace layout on the board

Take, for example, the Asus CFX3200
this board was not a cery good board at launch but many many many sites claimed it was the best thing since sliced bread.
there were two sites (planetx64 and Driverheaven) that refused to release a review until Asus fixed issues with Ram timing and skew.
there was a flaw in the way the traces were run on the board for the DIMM slots.
this cuased some major RAM issues.
the chipset was good the board design was not.
here IMHO we have the oposite.
the chipset IS flawed but the mainboard manufacturers have tweaked the BIOS' traces, resistors, caps etc to get performance and stability out of them.
I know many many people that have had problems with the 590SLI and the 680SLI
more than I can count TBH.
In my opinion the 680i SLI chispet has serious flaws in design.

[Quakindude]

We have one thing to agree on then. The 590 chipset was shit IMHO.