Are new G0 chips really too good OCers? - No!

[lopri]

Geez.. Why would anyone buy these new E6X50 chips? My E6600 from week 26th of 2006 does 3.60GHz with 1.31V (BIOS) - 1.27~1.29V (Windows). Yeah maybe X6850 is binned better and could do ~4.0GHz, but at what vCore? Is it worth it? Plus the lower multiplier and looser strap... The only thing going for them looks like lower temps to me.

Anyone who already has an E6x00 and wants to upgrade has only 1 choice - Q6600.

[lopri]

And I really hope this ES frenzy to come to an end. We all know what Intel did with E6600/E6700/X6800 when it was launched last year. As much as it's possible that the new stepping ES represent an improved overclocking, it's just as possible that it's another round of.. well you know what.

Besides, most folks have no access to ES chips and once retail chips become available verdicts will be out soon enough. Or at least wait for more legitimate reviews from reputable sites. (Intel even fooled them last year when it comes to overclocking)

[ziddey]

you have a very special e6600, and you're failing to see that most everyone else doesn't have as good a chip as that. in your case, definitely stick to your e6600.

as for es not being representative of the production models, it's already been well established. that said, from preliminary looks of production samples, e6850's are performing similarly to es models, and the lower models are suffering. personally, i had a e6750 l719a hit 8x400=3.2 with 1.37v actual under load. 8x420 wouldn't even stabilize with 1.5v, so it looks indeed like it hit a sharp wall somehow.

[sam95ta]

the e6850 really rocks!
e6850@3.8Ghz, fsb@423, ram sync, vcore@1.38v on air (cnps8700) with p5k-vm:

Great results.Can you please give us FPO/Batch #.

[pvhk]

Grat results.Can you please give us FPO/Batch #.

L719B016:


Max bench on air vcore @1.60v bios (1.49v under xp):

[MacClipper]

I am actually quite happy with my rather cool running E6750@3.6@1.45V stock Commando but just can't wait to grab the incoming Q6600 SLACR... if only I can find one.

Totally no reports of any retail G0 Q6600 benched over the www so far.

[crossg]

Totally no reports of any retail G0 Q6600 benched over the www so far.

Yeah I am sure there are a bunch of us waiting to see if the retail chips clock higher and run cooler. Both My E6600 and Q6600 are good Ocer's but to get the temps down on my Quad might be worth the jump to a G0.

[Forsaken1]

Just getting started.A chips do not appear to be all that bad.
FOP/BATCH:L718A190
Pack Date 6/27/07

[AndyM]

Are they Orthos stable for several hours at those speeds?

[kiwi]

I just found out that Xeon quads socket 775 3210 (8x) and 3220 (9x) are both cheaper than 6850. So if you prefer quad over dual core get xeon

[lee63]

Just getting started.A chips do not appear to be all that bad.
FOP/BATCH:L718A190
Pack Date 6/27/07

I have the same chip, I cant even get it stable @ 3.7 on a evga 680i A1 p30 bios

[newls1]

I have the same chip, I cant even get it stable @ 3.7 on a evga 680i A1 p30 bios

That might be the OCable difference between nvidia chipsets, and Intel chipsets

[Spanki]

Since they changed the max Tcase on these G0 stepping chips, all those Core temps are suspect. If you have CoreTemp 0.95, look under Options->Settings...



...click that "Show Delta to Tjunction temp" option... the temps reported will be how much headroom room you have left (it counts down as temps go higher) before the cpu throttles itself.

I'm not sure what value is being used for the new E6x50 chips, but earlier desktop quads used a 100C Tjunction, and it looks like they increased that by 11C for the new G0 stepping Q6600.

Older/current desktop C2Ds used 85C, but I'm guessing they bumped that up as well... maybe by 11C, but maybe less than that for the Duos. It would be nice if Intel actually said what was being used for each cpu/stepping.

[sofarfrome]

I just found out that Xeon quads socket 775 3210 (8x) and 3220 (9x) are both cheaper than 6850. So if you prefer quad over dual core get xeon

Please give a link to the site that has Xeon 3220 for less than the E6850.

[Brother Esau]

Yea no kidding

[turtle]

Please give a link to the site that has Xeon 3220 for less than the E6850.

Wait for it, it just so happened the C2D/Q cuts hit first. Xeons are planned for late July (29th?). It should be there in the next week.

He's right, word has it prices will be cut to $224 (cheapest quad) and $266 (same as Q6600) for the 3210/3220 respectively, and Xeons have a history of selling cheaper than their made-for-desktop counterparts...So they may end up even cheaper.

Overclocking the 3210 FSB to 400-450 would be the ultimate budget power rig.

[kiwi]

Please give a link to the site that has Xeon 3220 for less than the E6850.

I have it from my local supplier but you should see it very soon in a few days/weeks unless US stores rip you off and charge more for those xeons

[limb0]

over the last couple of days i've been playing with a e6550 on asus p5k and wondering whether to switch to this new "G0" steping from my trusty nine month old e6600 [9 x 400 on asus p5w]

when i saw the 7 multiplier I immediately thought of the e6300/6350.

it seemed to make sense seeing as the cpu sells for around the same price as them both.

the 6550 initially ran prime for a long time at 3.7ghz but later crashed and eventually i settled at 3.6ghz as the stable best "quickie overclock" result I could get.

i was considering two upgrades, A] the p5k [from p5w] and B] the 6550 from 6600.

naturally i LOVE the p5k but will probably pass on G0 stepping versus my antique e6600 because I still have hopes of hitting 4ghz on this fabulous new motherboard with my old processor, at least until the quads drop in price, whoot!!

[ziddey]

what's the batch code for your e6550

[palese]

Ok I am new to asus OCing so flame away if so desired But how the hell do you see the changes in cpu speed as your adjusting the FSB on the
P5K Deluxe? I asked in this thread as there is a lot of 6850 and P5K combos out there.

[lee63]

That might be the OCable difference between nvidia chipsets, and Intel chipsets

I think you might be right, but I need an SLi board

[Blauhung]

Are Q6600 G0 really Better than Q6600 B3 for overclocing?
how much?

see my earlyer post
http://www.xtremesystems.org/forums/...8&postcount=85

answer is that they are much better at stock speeds and a few have a chance to clock higher, but most likely you are probably getting a chip of lesser manufacturing quality that functions better at the stock clock speeds due to the fixes that came with the G step.

[Forsaken1]

Are they Orthos stable for several hours at those speeds?

No it is not orthos stable.Only had a hour to play this morning.Will try to get it orthos stable over 4 shortly.It does not look good.Maybe occt.Im on water with 1.58v.
Any volt mods for Gigabyte P35C-DS3R?

[lopri]

you have a very special e6600, and you're failing to see that most everyone else doesn't have as good a chip as that. in your case, definitely stick to your e6600.

as for es not being representative of the production models, it's already been well established. that said, from preliminary looks of production samples, e6850's are performing similarly to es models, and the lower models are suffering. personally, i had a e6750 l719a hit 8x400=3.2 with 1.37v actual under load. 8x420 wouldn't even stabilize with 1.5v, so it looks indeed like it hit a sharp wall somehow.

I agree the specific chip that I kept may be an exception vCore-wise, but I think hitting 3.50~3.60GHz with ~1.50V was pretty common with most E6x00 chips? (Before Intel started binning for Q6600s) Now we have these E6850s that hit 3.8~4.0GHz with 1.55V-ish vCore. Can't really see the point of upgrading, especially with their lower multipliers and loosened strap. I mean, outside professional benchers, I don't think the difference between 3.60GHz and 4.00GHz is something to go through the hassle of upgrade for.

Q6600, on the other hand, is a different story. It's unfortunate that there are only handful of apps that can take advantage of quad-cores, but for folks who use them, it's a god-send. Previously these folks had to rely on workstation boards with dual sockets for their productivity.

[Spanki]

... Can't really see the point of upgrading, especially with their lower multipliers and loosened strap.

Lower multipliers? I think you're viewing these new cpus wrong. If you compare price-points, you'll see the following:

E6850 - 9x multiplier, 3.00GHz replaces E6600 - 9x multiplier, 2.4GHz
E6750 - 8x multiplier, 2.66GHz replaces E64x0 - 8x multiplier, 2.13GHz
E6550 - 7x multiplier, 2.33GHz replaces E63x0 - 7x multiplier, 1.86GHz

...they're not lower multipliers, they're 'higher cpu speed' due to the higher FSB.