Will you still get core i7 knowing this?

[xgman]

This may have been already discussed, but does this put a damper on your plans to upgrade to i7? Seems pretty scary for what is likely to be a big investment.

"Intel is adivising mobo-makers and memory vendors alike about their new X58+Core i7 combo and to abide to a strict 1.65V limit on memory voltages, beyond which there is the risk of burning out the CPU."

ouch!


http://www.theinquirer.net/gb/inquir...damage-nehalem

[MarlboroMan]

really bad news for overclockers

I will stay with the old skool line of processors, mobo and RAM, till Core i7 get out of "beta" stage and support 1600+ RAM

[THUMPer]

Apparently Asus is running 1.7v RAM in their Labs. What a bunch of crap. haha

[v0dka]

This reminds me of the 3.3v jumper on the S939 DFI. If you used that with high memory voltage, while at the same time using only stock voltage to the CPU, you damaged the mem controller.

Clearly, Intel's onboard memory controller is suffering from something similar.

This sucks, but there's no telling yet of how caerfull we should be with this. Maybe it's all within xx % safety range.

[RealTelstar]

Well, first is the inQ. Then, 1.7 could be enough for 2ghz ram.

[Solarfall]

what the hell??? only 1.65v thats terrible for overclocking !!!!!!!!!

[xgman]

Well, first is the inQ. Then, 1.7 could be enough for 2ghz ram.

The InQ just picked up the story from another site. The pic tells all though. I doubt I would be brave enough to put 1.7v through a $1000 oem cpu 24/7 under those rather dire warnings. That tells me that Intel knows that the i7 is sensitive as hell to mem controller voltage and will likely suffer damage unlike prior models. That gives me a good deal of pause for moving to this as the new overclocking platform which I normally be doing. I'm sure though that this too will have it's share of guinea pigs to see what the limits really are. In any case, this really puts a damper on my enthusiasm at least for now.

[B.E.E.F.]

This is terrible!

But won't this result in high-end memory producers making low(er) voltage DDR3?

[RealTelstar]

But won't this result in high-end memory producers making low(er) voltage DDR3?

I hope...

[Particle]

Might not be all bad then. Perhaps we'll start to see some high(ish)-perf DDR3L spec memory.

[Zeus]

1.65V? I bet my Diablo's won't even boot with that.

Bummer.

[ReverendMaynard]

Pretty sad, but this could work out in the end. Either way, it's not going to be a great start.

[Infa]

Well, what some of you are missing are, Default speed on the ram are 1333mhz, and DDR3 are "default" voltage 1.5 from scratch, so using cl6 or 7 will work at these low speeds.

[xgman]

I guess this also give more reason to go with the "xtreme" unlocked multiplier versions if possible.

[ownage]

W00t
Now I have no reason to buy i7. I'll keep my money in my pocket.

[jaredpace]

i won't be buying it. I still think a 4ghz penryn will be faster than a 3.2ghz nehalem in my precious video gamez

[Ket]

This is why I've said I have no interest in Nehalem In all honesty though, 1.8v is likely perfectly fine to use. Intel probably just trying to kill off some % of their RMAs due to dimwitted OCers.

[DMH]

Haha guys

Don't be so negative! I have used 1.8V+ memory voltage two days now and there is no problems.

If your CPU is really early ES, there might be problems. Those CPU's are not in public use, so this shouldn't be an issue.

Memory will overclock like before and in this board you can choose speeds up to 2000MHz(DDR3) (800,1066,133,1600,1800,2000)

I tested my sticks today at 1400MHz and 6-6-6-15 timings (1.8V)

Moderators can erase this thread

[lha]

OMG that is piece of crap then..

[Nasgul]

I think being naive and gullible at the same time just hit a new level.

[Mats]

I'd say C2 is good enough at least 6-12 months more (how unxtreme of me..). I don't want to join the i7 beta testing, too much money. The i7 will be fast though, no doubt.
It's not the same situation as when C2 showed up, this time it's more new technology involved and it will take time before everything works fine and DDR3 prices drops.
Dunno about the CPU's, but the first mobos will get old very fast. Name a mobo with a new socket that actually was really good . . I guess there aren't many.
It's not a problem for those who upgrade twice a year tho.

[miguelca]

xgman, thanks for posting this pic and starting this thread.
I want you to know that this is just my two cents and I thank everybody for reading my opinion.
In no way I want to start any kind of negative discussion or cause the impression that I'm smarter than anybody.
In this community called XS, I consider myself an apprentice.

This warning doesn't bother or discourage me at all.
When I bought my actual DDR3 memory, I didn't spend a fortune because I was already expecting that it wouldn't be compatible or adequate for use on the upcoming i7 platform.
I just wanted to learn how DDR3 behaves on a daily usage setup, but without taking the risk of throwing it away in a few months.

The chipset and architectural changes will be big on i7, and we should all be prepared to fight some long battles setting up memory in the first Nehalem desktop (and server) systems.
Most of the problems we had on AMD platforms will now "haunt" Intel.
Setups like "triple-channel" for example, will probably require stupidly high quality memory in order to be stable, and lots of tweaking for sure.
IMHO the whole DDR3 memory scene will (must) improve a lot in the following months.
Lower voltage and higher performance is the mandatory way to go for Intel, and since the Core architecture was released they're on the right path (again IMHO).
I wouldn't be surprised to find memory manufacturers soon releasing 3GB or 6GB kits (or maybe greater), with 2000+ MHz stock speed, functioning at a mere 1.4 or 1.5v.

Many people think voltage is the cure for every successful OC.
Of course that many components in fact need the extra power to be sucessfully overclocked.
In what concerns RAM, and from what I learned on my latest experiences while testing high-speed and high-capacity DDR2, I could see that most of the kits wouldn't OC better no matter what voltage I'd throw at them.
Some IC's simply don't react to more volts.
Adjusting skews, carefully setting up timings and learning how to deal with some settings that simply have different names from one MB brand to another, is much more important (and hard) than give more juice to the components.
Mid and high-end motherboards with P45 and X48 chipsets need a lot of tweaking in order to make memory perform well.
This will be probably even much harder on the X58.
But, no matter how hard the fight will be, I'm anxious to experiment with Intel's new babies!

Come on guys and let's be positive.
We're about to spend some huge amounts of money just to be cool and up-to-date.
Intel certainly is working very closely to the memory manufacturers, and they're probably trying to create the need for i7 certified RAM kits to emerge.
New architecture, new socket, new memory, new... God knows what.
Marketing BS soon to come.

At least let us hope that the performance improvements are worth the change.
If they do I'll have no problem buying a platform that requires me to be careful with the memory voltage I setup.

Well, enough BS from myself already.
Sorry for this huge post, that may not make any sense at all.

[stealthbomber]

DDR3 is still way too expensive, any C2D chip (even the E2xxx/E4xxx/E5xxx budget chips) @3GHz+ is good for at least another 12-18 months, if you have an overclocked kentfield/yorkfield might as well forget about i7 until at least the 2nd revision.


4GB was barely enough for my Vista64 install, adding another 2GB made it a lot better, I'd imagine Windows 7 would benefit with 6GB+ of ram.

I could buy 4x2GB PC8000 DDR2 modules for just a little over £100, a set of 3x2GB DDR3 modules still cost a small fortune!

[whe3ls]

cool im get another low vid q6600 anyway

[Dee]

DDR3 is still way too expensive, any C2D chip (even the E2xxx/E4xxx/E5xxx budget chips) @3GHz+ is good for at least another 12-18 months, if you have an overclocked kentfield/yorkfield might as well forget about i7 until at least the 2nd revision.

Yes, I agree.

The original point of overclocking was to push cheaper parts until they could compete with the top-end ones. Are we all forgetting this? I have more respect for someone who can get insane speeds from a low-end CPU than someone who jumps on the latest thing, has more money than sense and feels the need to impress people on the Internet.

i7 doesn't even interest me, nor does DDR3, and it'll be quite a while until I even care at all.