The Official Sandy Bridge-E Overclocking thread

[radrok]

I agree with what you said on another post that the chips are too expensive to try binning for a good one... with the same amount of money required to purchase a 3960X I could have double chance with two 3930K, in fact I picked a 3930K so I'm going to see if it is going to be a good overclocker and if not I could always buy a second one.
Anyway I'm eager to see your reviews

 

[LiveOrDie]

Are you guys using speed step i watched this video funny how they say that a H100 can hold a oc of 4.8-4.9Ghz .

http://www.youtube.com/watch?v=Kx2z07sFM2I&feature=related

 

[cadaveca]

Yeah, ASUS has been very helpful in showing how to overclock your system using their boards for some time now. JJ has quite a few helpful videos up.

 

[LiveOrDie]

Well im sitting on 4.5Ghz with 1.31 VCore but my temps look to high for my likings.

 

[cadaveca]

I'm at 4.6 @ 1.40v now. Your temps look like mine, but i dunno what you are using for cooling.

 

[LiveOrDie]

I'm at 4.6 @ 1.40v now. Your temps look like mine, but i dunno what you are using for cooling.

Im using a H100 with MX-4 i have 2 fans pushing air in from the top of the case because this case is to small inside to install 2 extra fans.

 

[cadaveca]

I'm using a Noctua NH-C14:

Of course, I'm saying your temps are similar to my 1.3v temps. When @ 1.4v, I'm just shy of thermal throttle @ 90c.

 

[MetalRacer]

Well im sitting on 4.5Ghz with 1.31 VCore but my temps look to high for my likings.
http://img72.imageshack.us/img72/1323/49416777.jpg

Try raising you BCLK to 125 and lowering your multi to 36, your temps should come down.

 

[LiveOrDie]

Try raising you BCLK to 125 and lowering your multi to 36, your temps should come down.

Thanks ill try that i was just following what asus said so i just left the BCLK.

 

[Feizy]

You could always turn of HT, that lowered my temps a good amount.

 

[LiveOrDie]

Any one else PCH temp high like this?

 

[cadaveca]

When overclocked, yes I do get close to those temps. Board doesn't matter. But all my boards have a heatpipe coming form VRM cooler to SB cooler, except one.


Just finished my preliminary testing over the past week. 3 boards, 3 days each. Now I get to work on the written protion of the reviews, as well as now I know which board to use for memory testing.

There's alot to play with, options-wise, with these CPUs. The heatload is killer though. I realyl need to get a decent watercooling kit...I assume you other guys with higher clocks are running watercooling?

 

[LiveOrDie]

When overclocked, yes I do get close to those temps. Board doesn't matter. But all my boards have a heatpipe coming form VRM cooler to SB cooler, except one.


Just finished my preliminary testing over the past week. 3 boards, 3 days each. Now I get to work on the written protion of the reviews, as well as now I know which board to use for memory testing.

There's alot to play with, options-wise, with these CPUs. The heatload is killer though. I realyl need to get a decent watercooling kit...I assume you other guys with higher clocks are running watercooling?

Yer my temp are high as well im trying to get a OC which runs cooler im testing on 4.2Ghz running LinX my temps still hit around 80-85c which is to high, i really need a push and pull settup on my H100 but there no room for it, ill be buying some better fans next week.

 

[Bundy]

When overclocked, yes I do get close to those temps. Board doesn't matter. But all my boards have a heatpipe coming form VRM cooler to SB cooler, except one.


Just finished my preliminary testing over the past week. 3 boards, 3 days each. Now I get to work on the written protion of the reviews, as well as now I know which board to use for memory testing.

There's alot to play with, options-wise, with these CPUs. The heatload is killer though. I realyl need to get a decent watercooling kit...I assume you other guys with higher clocks are running watercooling?

When do you think your reviews will be published? As you had posted in another thread, you are reviewing boards that are what I am considering. Been kinda holding off due to price and indecision.

Can you give a pre-hint? i.e do those extra PWM in the UD5 VRM help? pls, all the online sellers in AU shut down for Christmas and I was hoping I could buy before the break.

 

[cadaveca]

ASUS has the best BIOS so far. Gigabyte's BIOS is good, but needs a bit of work for memory clocking. Of course, that's air testing...may be very different under water.

 

[LiveOrDie]

I just reset my CPU with the pea method seems to have lower my temps a fare bit will retest some of my overclocks settings.

 

[Bundy]

ASUS has the best BIOS so far. Gigabyte's BIOS is good, but needs a bit of work for memory clocking. Of course, that's air testing...may be very different under water.

I had a feeling you might say that. Here, the ASUS boards are $100 more than GB. I'll have to dig deeper if I can afford this extra $ before years end.

 

[cadaveca]

I had a feeling you might say that. Here, the ASUS boards are $100 more than GB. I'll have to dig deeper if I can afford this extra $ before years end.

There's nothing wrong with any of the boards I have, really, but they all seem to have very differnt focuses. I haven't actually looked at the numbers yet(I never do until I'm nearly done with the review), but for ease of overclocking only, ASUS does make it easier.

Part of the issues with the Gigabyte board is loadline calibration, and cooling. Cooling of the VRM is critical when overclocking these chips, and the ASUS cooler is just simply physically larger, providing a bit more headroom, it seems.

Over the next couple of weeks I'll be writing the reviews up, and hopefully they'll be going live as I complete them. I still have yet to receive MSI and Biostar boards, and I do hope to get other ASUS and gigabyte boards over the upcoming months, so at the moment, my perspective is quite limited. Perhaps on of these other products will stand out over the others...time will tell.

 

[Bundy]

Now I have heard the C2 will be available late January, I think I'll wait. There will be a better chance of mature bios then.

 

[LiveOrDie]

Now I have heard the C2 will be available late January, I think I'll wait. There will be a better chance of mature bios then.

Apart for a feature which only businesses would use being fixed there no difference from C1 to C2, The time frame is so small between stepping you mite want to wait for D0 stepping if you want to see some improvement.

 

[EarthDog]

let me first say, Cad's reviews are top notch, so I understand the wait. That said, there are PLENTY of other quality reviews around for you to look at for the boards you want. I suggest googling instead of anxiously waiting...

I also suggest not waiting for specific steppings of the CPU. On another unrelated note, for most people, especially at this site, the bios' that are available today for most boards work juuuuuuuuuust fine.

I also have to agree that the Asus RIVE bios is the most mature. The UD7 with F4d isnt far behind though. I just reviewed the UD7. I had 4.949Ghz HARD wall with custom water (PA120.2 and MCR320). Under a terrible SS, I broke that and hit 5.2Ghz. This chip has more in the tank, just needs to be colder.

 

[cadaveca]

Yeah, the heat of these chips is killer to overclocking. At least it gives reason to use water and such.

First review I am working on right this second...hope to get it done between today and tomorrow. My reviews tend to have 6000+ words including code, and 200+ images that need to be editted, plus taking those pics, testing, and all that jazz. Takes some time.

After that, I got to get a memory review done, hopefully by Saturday, Sunday is my day, then the following week another two board reviews. I gotta go out of town, so I gotta get these done before I leave.

When it comes to clocking, memory support is key. It's not such a big deal for me, knowing how to adjust secondary timings, and not all BIOSes support XMP v1.3 correctly yet, either. Anyway, yeah, 6000 words is alot, better get back at it.

And yeha, there are lots of other reviewers. Not many do the component-level type of review I do though, but there ARE others.

 

[radrok]

And yeha, there are lots of other reviewers. Not many do the component-level type of review I do though, but there ARE others.

Also most reviews are just a copy-pasta of the pages with scores modified and different images.
We need deep analysis, that's what makes a good review different from a simple review

 

[cadaveca]

Meh. It's understandable that reviewers do what they do, trust me. I literally put 40 or so hours into every review. Not everyone has the time to do that, either due to having other responsibilities, or having many reviews to do, and deadlines. I've got deadlines too, but very few other responsibilities that take up my time.

I got my CPU like 10 days ago, already got all three boards tested, and the written part of the memory review done, nearly all my pictures for all four reviews editted, too. I needed to test all three boards so I could pick the best for the memory review, but i'm still stuck doing the stuff in the order they showed up at my door. I just want to get them done, but then we're talking 20k words I gotta write.

 

[EarthDog]

The memory portion seems to be CPU dependent with these chips. Raja has posted a way to bin these things as well by memory clockspeed of all things...seems they are somehow related.

There is another way to bin these chips as well. So far its proven on a RIVE, but I am praying that the method works on UD7/other boards...basically go 2c/2t raise voltage to 1.55-1.6v and adjust your PWM stuff. Start at 51x multi. It should go to windows. reboot, raise multi. Eventually you will start hanging at the windows spalsh screen (for w7 its those colored balls), if you hit reset and the board boots, that multi is good. If you cannot reset and have to POR (Power Off/On Reset) that multi is bad. If that method is good on teh UD7, I have 2 56x chips. Here is to hoping.

As far as reviews. I dont have that much time. And to be quite frank, our readers dont give a hoot about which chips are controlling what. Its high level features and overclocking performance that drives our group, so thats what we focus on. Only do our release reviews touch on architecture of the chipsets/chips and those details.