X-bit labs

Started by: RtFusion | Date 04/17/08

Comments: 23 | Last Comment:  05/26/08

[1-16]

1. Wow, people are paying top-dollar for these boards and expect top-dollar performance.

Some people now get some serious slaps in the face. I imagine they demand explanations from the mobo partners and from nVidia themselves. I'd flip if my data was corrupted.

Well, I hope those people backed their data up on discs; data is the most expensive part on the computer as many people know.

Can you be sure the blame is only with nVidia for this problem :-)

I wouldn't put it past Intel to screw its only competition in the chipset business (Intel processors)

2. Nvidia needs to buy Via seriously and get their help with making better chipsets.

Not only the drivers suck now, they are also too hot, too power-hungry, and incompatible with many things like Samsung's F1 1 Terabyte hard drive, some sound cards and other devices.

VIA isn't any better. In fact, the latest PT900 chipset can't even support 1333MHz FSB core2 in STOCK speed, and have trouble with PCI Express 2.0 video cards. Worse yet, none of the chip sets support 2GiB DIMM _or_ DDR3.

3. Well at least they owned up to the problem.

Sure hope they find a solution because the the price level of their product, it has to be flawless, pure and simple.

4. Nothing surprising going by NV's previous appalling record of bad/bugy chipsets that they only own up to after thousands complain

And to make things worse they charge more than any other for their crap chipsets

5. I like Nvidia's GPUs but I never bought a motherboard powered by one of their chipsets. Intel's chipsets, while not perfect, are the gold standard for stability and performance even they've not always been feature-rich.

Looks like I'll be sticking with Intel chipsets for the foreseeable future.

6. Nvidia chipsets are flawed and they know it.
The only reason people went with Nvidia is because of their
stupid SLI.

If Nvidia did not tied their SLI to their chipsets, then you can see a majority of the users going with Intel or other chipsets instead of
SHIT-VIDIA .

BINGO!

7. Nvidia has start to become the next Creative.

"We'll fix it with out next product"!!!!! :P

8. Why am I not surprised? The reference 780i and 790i boards seem to have all kinds of driver, memory, and reliability issues. I'm glad I decided to go with a DFI X48 board. No SLI, but I'll take overclocking stability over SLI anyday.

9. learn how to write.

10. Same problems seem to have been occuring since the Nforce4 - I remember having a HDD keep bricking on me, blaming my overclock but eventually finding out through an Nv engineer on the phone that the Nf4 chipset caused problems with this specific model of drive - since both Nv and the HDD maker technically met the specs, neither was willing to step up to help fix the issue :(

11. Overclocking voids warrenty and is not supported by NVIDIA. so when users o/c, get data corruption, and then whine about it?
seems like trying to eat the cake and keeping it. but we all know the cake is a lie.

what cake?!
hell if you are going to get a top notch piece of hardware you should KNOW/EXPECT it to manage some little over squeezing. A 300$ motherboard should be, in ALL cases, surpassing the overclocking capabilities than ANY 150$ motherboard while keeping better stability of components.
Also, if nvidia/motherboard manufacturer company doesn't support overclocking, why do they enable it in their bioses?!
They all support overclocking, officially or unofficially, their bios settings told US all about that.

12. That truly sucks, i just bought one and i was planning on overclocking!

13. First off, overclocking is known to cause these types of problems. It comes with the territory. Data corruption during overclocking is not a bug, it is normal.

Second, Intel and Nvidia aren't exactly friends. They are bitter rivals. Intel isn't going to give Nvidia any details they don't have to about constructing chipsets. This is the reason Nvidias chipsets are lacking in the stability department. If you were Intel, what reason would you have to give Nvidia the secrets of your FSB? Money. And that is also why Nvidias chipsets are overpriced.

I would say the problem lies within the system of competition, rather than any sort of incompetence on Nvidias part. If Nvidia/Intel/AMD merged they would create alot better things than they do today... assuming they had honest, dedicated people at the top.

Quote from hedron:
"I would say the problem lies within the system of competition, rather than any sort of incompetence on Nvidias part. If Nvidia/Intel/AMD merged they would create alot better things than they do today... assuming they had honest, dedicated people at the top."

Did you learn your economics from China's formerly failed Chairman Mao's regime's Economic School of Socialism?

The common factor here is the CPU. Maybe if there were more competition on this front, we would have better overall computer tech.

Get a clue.

[1] overclocking doesn't cause problems, overclocking with stupidity and without knowing your safety zone, is what causes problems. Also, bad motherboards cause problems when overclocked, but that's all.

[2] well, if nvidia is going to keep SLi to itself and dedicated to their chipsets, then Intel can do the same with Nvidia, i see it a 1vs1 situation. That's of course if it was proven that the cause of these problems are intel related anyway.

[3] if nvidia/ati/intel/amd merged "with whatever meanings this word carries" that would screw US ALL on a customer level.

There was a time when there was only one phone company and the phone service here was great. Then the government came in and broke up AT&T and then phone service sucked. This isn't socialism, it is human nature to want to do better than others have done. It's just over-simplistic dialect to think that this is caused by competition. It is, in fact, your own marxist ideologues that come to the false conclusion that men only do things in desire for money.

14. "May experience" - are there successful overclocks being done on those chipsets?

Off the top of me head (ie - without going and checking how the data is being shipped around the bus and in what order and all the usual), maybe the data corruption is being caused by queing issues due to bottlenecks.

What if there were buffered DDR2 and 3s available for games/high-speed systems, or even slower but steadier error checking.

Say if the RAM is overclocked, but it's having to wait & hold that data due to bottlenecks - could the data held be suseptible to decay, as in - it begins to lose bits as it waits. Just thinking in terms of doing anything faster than usual, it tends to lead to more mistakes.

All that aside though, if the only difference between a successful overclock (same settings, same hw) and one with data corruption are these chipsets, then it must be something to do with them not being able to handle the bus speeds. It's kinda like DNA at that stage, you could drop/miss a bit and it makes the difference between coherence and gibberish.

15. The company answer refers to concerns andantech sent them; but their own review of a 790i board makes no mention of those issues....

it does imply though that you can't raise the FSB too high, and the multiplier is used instead.

16. A quick sweep in some forums gives you info on alot of people having these problems without overclocking.

[1-16]