Is the i7 920 D0 the new Q6600 G0?

[Gillbot]

The reason I ask this is.....

After reviewing many threads in CPUs and Overclocking, Cases & Cooling and Motherboards, it seems MANY are buying the i7 920 expecting to walk into a 4GHz overclock. We had the same scenario with the Q6600 G0 where people basically bought one expecting to walk into an easy 3.6-4.0GHz overclock on those. Unfortunately many found out later on, the later G0 batches were lucky to hit 3.4GHz.

Is all the internet hype around the i7 920 D0 giving people the wrong impression? I know, I never really pushed my i7 920 D0 at all. I ended up selling it to my brother and got it up to 4.0Ghz after battling with settings for quite some time. Even at that, it still runs VERY hot and IMHO, 3.6 seems to be a more reasonable clock.

There seems to be MANY threads asking for help getting the 920 up to 4Ghz and even fewer showing the 920 at 4Ghz. That leads me to believe the 920 is the next Q6600, where everyone buys them expecting a certain clock only to struggle and be disappointed.

 

[Markfw]

I got 3.8 easy, 3.9 after some work. 4.0 works for more than 5 minutes, but its not stable for me. I see 3.8-3.9 as the norm from the posts I have read.

 

[SunnyD]

I think people think this in general because of the reviews when a new model or stepping comes out. The first thing that's done is someone suicides the chip, and the immediate next review are the people who stretch and run as high/stable as they can. Once you get three or four of those kinds of posts on enthusiast forums like this, then people think the new model/stepping is golden.

Happened with the AMD Denebs too.

 

[exar333]

The reason I ask this is.....

After reviewing many threads in CPUs and Overclocking, Cases & Cooling and Motherboards, it seems MANY are buying the i7 920 expecting to walk into a 4GHz overclock. We had the same scenario with the Q6600 G0 where people basically bought one expecting to walk into an easy 3.6-4.0GHz overclock on those. Unfortunately many found out later on, the later G0 batches were lucky to hit 3.4GHz.

Is all the internet hype around the i7 920 D0 giving people the wrong impression? I know, I never really pushed my i7 920 D0 at all. I ended up selling it to my brother and got it up to 4.0Ghz after battling with settings for quite some time. Even at that, it still runs VERY hot and IMHO, 3.6 seems to be a more reasonable clock.

There seems to be MANY threads asking for help getting the 920 up to 4Ghz and even fewer showing the 920 at 4Ghz. That leads me to believe the 920 is the next Q6600, where everyone buys them expecting a certain clock only to struggle and be disappointed.

It has been for a year. My "ancient" C0 (also bought for $299 at launch) is chugging away at 3.8ghz happily. The D0 is a little happier and cooler at better voltages, but any 920 is great.

Without significantly upgrading my cooling, my rig is still pretty much as good as it gets a year later. Sure I could get a D0 and maybe run at 4.0-4.2 with the same cooling, but thats not a worthwhile upgrade IMHO unless you got a "dog" C0 originally.

 

[Gillbot]

I got 3.8 easy, 3.9 after some work. 4.0 works for more than 5 minutes, but its not stable for me. I see 3.8-3.9 as the norm from the posts I have read.

HT on or off? From what i've been reading, many are getting too hot with HT on.

 

[rolodomo]

Somehow I drew a mediocre D0 from a supposedly good batch made in 2009. Vid was 1.2785 and I had to increase vcore to get above 4 GHz. Nonetheless, my 24/7 OC is 4.2 GHz at 1.4V and heat buildup is not that bad at all (HT off). I'm on water, but a good air cooler would be in the same league.

I seldom post my OC setting because it seems so "average." I think most people would be interested in reading from someone who gets a D0 to 4.5 (or so) on good voltage.

Some of the posts I've been reading for OC help lately don't seem to have the right attitude. They don't want to read through the ton of information out there about overclocking the 920. That I don't understand, there's nothing that makes me feel better about a computer part than going on the net and finding tons of helpful threads about it.

Also, it is not a trivial matter to squeeze out that last .3 GHz on a fully featured 1366 bios. I can see how it would get really confusing to an inexperienced user.

 

[cubeless]

/\

everybody has a 'dog' unless it just pops up at 4ghz these days...

sheesh... you do a mediocre job of installing the hsf, you don't get there... you get any weak parts, you have to work hard or you don't get there...

and you would have to have a pretty busy app (or just be a bmk freak) to feel a seat of the pants difference between the 920 @ 3.8 and 4.0...

i just got through beating 4ghz out of my frys deal 860/t5xe/ocz obsidian combo... the last 200mhz probably wasn't worth the trouble... but i do this for fun... now it's on to the 965x4/ma790x-udp4...

 

[Gillbot]

It just seems like there are a million "help me get to 4.0+ or what to buy to hit 4.0+ on a 920" threads out there that they seem to have the same hype the Q6600 had.

 

[cubeless]

It just seems like there are a million "help me get to 4.0+ or what to buy to hit 4.0+ on a 920" threads out there that they seem to have the same hype the Q6600 had.

all the articles about 'effortless' overclocking... although i have to admit that getting a 920 to 4.0 was easier than getting the 860 or 9550... at least for the bits i used...

 

[exar333]

/\

everybody has a 'dog' unless it just pops up at 4ghz these days...

sheesh... you do a mediocre job of installing the hsf, you don't get there... you get any weak parts, you have to work hard or you don't get there...

and you would have to have a pretty busy app (or just be a bmk freak) to feel a seat of the pants difference between the 920 @ 3.8 and 4.0...

i just got through beating 4ghz out of my frys deal 860/t5xe/ocz obsidian combo... the last 200mhz probably wasn't worth the trouble... but i do this for fun... now it's on to the 965x4/ma790x-udp4...

As others have stated here on AT previously, I am guessing a decent share of these "stable" 4ghz chips are in fact, not so stable. While I consider 48 hours of linpack on all threads "stable", others may test for longer, shorter, or barely test at all ("It didn't crash while playing iTunes!").

Not to say that 4.0ghz is not often relatively easy to obain for a D0. I think a lot of people just don't want to admit their chip "only hits 3.6ghz" stable, so they run at a unstable speed, or turn off HT to hit a higher clockspeed. Turning off HT is completely valid, but I would rather keep it on rather than get an extra 100-200 mhz on my OC.

 

[dmens]

The 920 bin leakage variation has always been huge. Some of the 920 parts could have been sold as 975 if they used slightly less power, whereas others would need double the TDP to hit 975 speeds.

Almost all bloomfield parts can get to 4ghz, the power requirement varies widely. My home D0 runs at 4ghz on 1.18V (stable for over a year now), but I have a part that is completely off-the-chart on leakage.

D0 is only slightly faster than C0. Maybe 100MHz at 90C. There's no huge leap with D0. The previewers/reviewers probably got golden samples.

 

[Markfw]

HT on or off? From what i've been reading, many are getting too hot with HT on.

60-67c depending on abient, HT on, 100% load on all 4 cores. (F@H)

 

[Gillbot]

60-67c depending on abient, HT on, 100% load on all 4 cores. (F@H)

Nice, lucky chip.

 

[Markfw]

Nice, lucky chip.

nah....megahalems cooler !

 

[nitromullet]

As others have stated here on AT previously, I am guessing a decent share of these "stable" 4ghz chips are in fact, not so stable. While I consider 48 hours of linpack on all threads "stable", others may test for longer, shorter, or barely test at all ("It didn't crash while playing iTunes!").

Not to say that 4.0ghz is not often relatively easy to obain for a D0. I think a lot of people just don't want to admit their chip "only hits 3.6ghz" stable, so they run at a unstable speed, or turn off HT to hit a higher clockspeed. Turning off HT is completely valid, but I would rather keep it on rather than get an extra 100-200 mhz on my OC.

Agreed. I actual run mine at 3.5GHz because the temperature difference between 3.5/1333 (cpu/mem) and 3.6/1600 is significant. By running at 3.5, it allows me to run my system at a much more bearable fan volume. Plus, I can run at a higher multiplier, which means I see greater benefits from speed step when idling.

Granted, when I first got my i7 920 we were experiencing a heatwave in Seattle (don't laugh, it actually hits the high 90s on occasion), and since we don't have A/C it was sweltering hot indoors. I figured it was a solid move to OC and bench in pretty much the worst conditions I would possibly throw at the cpu. She's been rock stable, fast, and pretty quiet for an air cooled gaming rig.

 

[mejobloggs]

While I consider 48 hours of linpack on all threads "stable"...

Hmm serious. I'd be worried about frying my pc on stock settings with 48 hours of linpack

 

[Grooveriding]

I had little to no trouble getting my 920 D0 to 4.2, this is using and with HT on. 21x200 1.33V Cooling via Ultra-120 Extreme.

Granted it gets way too hot if I run prime95 etc.. about 80-85 on the cores, but that's only under that situation. Gaming, normal use, etc.. at most 60-65.

 

[MagickMan]

I've setup four 920 D0 rigs for people, built just for OCing, and all four hit a stable 4GHz without much effort at all.

 

[BTRY B 529th FA BN]

I think people are amazed they are seeing 4Ghz on their processors so they assume it will be stable 24/7. They do 30 min of stability testing and call it stable, atleast that's the majority of posts i see on the Extreme OC forum.

 

[Kantastic]

Nice, lucky chip.

F@H doesn't stress the GPU as much as OCCT/Prime95.

As for the original topic, most D0's hit 4GHz without a problem. The chip I had was OCCT (3 hours) & Prime95 (overnight) stable at 1.28v.

 

[Rubycon]

I think people are amazed they are seeing 4Ghz on their processors so they assume it will be stable 24/7. They do 30 min of stability testing and call it stable, atleast that's the majority of posts i see on the Extreme OC forum.

I have ways to test it and if it passes it's stable for sure. Many turnkey boxes (Dell) flat out FAIL too! Then again they're not doing the stuff WE do as well.

 

[Gillbot]

nah....megahalems cooler !

My brother is on water and he can't get there! Temps are holding him back with HT on.

 

[cbn]

The question in my mind is how quiet i7 920 would be at 3.8 GHz, 3.9 Ghz or 4.0 Ghz.

I would take lower noise over 100-200 Mhz any day of the week.

 

[cbn]

The 920 bin leakage variation has always been huge. Some of the 920 parts could have been sold as 975 if they used slightly less power, whereas others would need double the TDP to hit 975 speeds.

Almost all bloomfield parts can get to 4ghz, the power requirement varies widely. My home D0 runs at 4ghz on 1.18V (stable for over a year now), but I have a part that is completely off-the-chart on leakage.

D0 is only slightly faster than C0. Maybe 100MHz at 90C. There's no huge leap with D0. The previewers/reviewers probably got golden samples.

This makes me think more about the importance of binning.

 

[WhoBeDaPlaya]

I've gotten 4x 3849B018s and 2x 3849A925s to 4.0-4.1GHz without worries on EX58-UD3Rs with 1.152Vcore, HT and LLC enabled.
Hell, they all use the exact same voltages and settings.