Man Vs. Machine: Four Automatic Overclocking Techs, Compared

Page 2 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
Status
Not open for further replies.
[citation][nom]hyteck9[/nom]Cangelini, thank you for the reply. I think you have confirmed my interpretations. So if the CPU is the "control" in this experiment, and we know it takes 1.35V to make the CPU do what we want, then all this overclocking "buzz" is really about pushing the board's stability isn't it? (Like the SBM Gigabyte board missing a few heatsinks issue.) So how can a consumer identify a "low-push" board? Is it the chipset? the BIOS? integrated video? It sure isn't price, because I know that Gigabyte board from the SMB isn't on the cheap side. Tell us what to look for![/citation]

Well, it's almost never as easy as looking at the board, physically, and telling. That's why we try to be as thorough as possible in reviewing motherboards, categorizing them by chipset, and then building round-ups based on price ranges.

If you're not already familiar with some of the work we do in that regard, please take a look here: http://www.tomshardware.com/reviews/Components,1/Motherboards,2/

Going down the list, you can identify the stories most interesting to you, and then dig into the specifics of each board, including how well they each overclock. In many cases, Thomas will weigh in on why one platform might behave a certain way, too!

All the best,
Chris
 
I have been using Gigabyte's Smart QuickBoost on Twin Turbo for several days now, effectively overclocking my i7 2600K to 4.2Ghz. I've experienced zero problems, although I haven't yet tested it on full load using a utility like Prime95. I've played some games, though, and so far so good. IMO, QuickBoost makes overclocking easy and relatively safe for those of us who lack the knowledge (or the time) to OC ourselves.
 
I have to say, all this trashing of auto-overclocking really stinks of elitism. Tom's presented a thorough overview and analysis of several auto-OC utilities, with benchmarks, and they have indicated in the comments that they stress tested the setups. How can people say that these options for achieving an overclock are useless? Sure, learning to do it yourself is empowering and the knowledge is useful, but not all of us have the time to get into manual overclocking.
 
I don't know about other mobo utilities, but Gigabyte Easytune6 is garbage. I tried using the preset to OC my i5 750 (stock 2.66GHz) to 2.8GHz and the CPU idle temp went from 30C to 60C (intel stock cooler). Recently, I tried using it to change the ram speed, and upon rebooting, found only 2 of my 4 RAM modules were enabled (8GB total, but only 4GB enabled). I had to remove the CMOS battery for a few minutes to reset it. Guess what utility just got uninstalled?
 


I'm using Gigabyte's QuickBoost, and my CPU consistently idles at 23C.
 
[citation][nom]okcookienc[/nom]I'm using Gigabyte's QuickBoost, and my CPU consistently idles at 23C.[/citation]Wow, in a 23C room? With a case temperature of 30C, right? And using air cooling?

I believe your thermal monitor is producing an erroneous reading. Physics anyone?
 
would have been better if you guys used 3 or 5 different samples of processors.
that would give a better idea what kind of real world result people could/should expect rather than running a hand picked sample.

I understand time constraints and deadlines and such...
 
Guys, please remove WinZip from the benchmark suite. I really don't think anyone in their right mind would use that when 7-zip is freely available and decimates even the threaded WinRar in terms of compression speed. The purpose of the benchmark suite is to demonstrate performance in real world usage scenarios. WinZip no longer fits that description.
 
Nice article, but what is the point of overclocking a i7 2600K? And most people shelling out that much money on a processor probably know how to manually overclock. Personally I would have picked an i3 2100 or the i5 2500K for this test.
 
Hi,

Ive just had a computer made for me it has the i7 2600K processor and the builder has clocked it to 4gh. Now I know nothing about over clockling or anything tecnical with computers. Im getting about 47C - 51C sitting on windows desktop screen and when I play Witcher 2 on ultra graphics and 1920 x 1080 res the prossesor goes up to 97C. This to me seems extremely high. Ive just bought 3 more fans to bllow air into the cse (Antec300). I think the i7 2600K has the standard heatsink and fan on it. My graphics card is the Nvidia GTX 590. It has no problem running Witcher 2 and theres great framerate with everything on ultra settings its just the temp of the processor that worries me. Its the AXTU extreme tuning abillity thats showing me these temps when I alt tab out of the game to look at it (but then the temp goes straight down again when alt tab out of the game).

Any advice would be most appreciated
 
[citation][nom]Chrismvts[/nom]Hi,Ive just had a computer made for me it has the i7 2600K processor and the builder has clocked it to 4gh. Now I know nothing about over clockling or anything tecnical with computers. Im getting about 47C - 51C sitting on windows desktop screen and when I play Witcher 2 on ultra graphics and 1920 x 1080 res the prossesor goes up to 97C. This to me seems extremely high. Ive just bought 3 more fans to bllow air into the cse (Antec300). I think the i7 2600K has the standard heatsink and fan on it. My graphics card is the Nvidia GTX 590. It has no problem running Witcher 2 and theres great framerate with everything on ultra settings its just the temp of the processor that worries me. Its the AXTU extreme tuning abillity thats showing me these temps when I alt tab out of the game to look at it (but then the temp goes straight down again when alt tab out of the game).Any advice would be most appreciated[/citation]I'm not sure if those temperatures are correct but if they are, it's probably because the cooler is insufficient. Realtemp give me reliable temperature readings and as for the cooler, nearly any 120mm tower-style cooler will work well with your case if you mount it cross-flow so that the rear exhaust fan sucks heat away from the back of the cooler.
 
[citation][nom]Chrismvts[/nom]... Any advice would be most appreciated[/citation]

As Crashman says, you need a better cooler. The range is vast of course,
but heatsink reviews should reveal some sensible options for performance
vs. price. I've been using Thermalright U120 Extreme RevCs (TRUE for short)
because of their low price now, combined with Gelid Wing Blue 120mm fans
(the only decent replacement I could find for the previously popular
Scythe SFF21F, the gelid is actually quieter). However, if you can't
find a TRUE, the are numerous equivalent models in terms of price/performance,
such as the later TRUE Black (somewhat better, bit higher cost), the
Venomous-X, Noctua NH-D12/14, Cogage True Spirit, etc. The list is vast.
Indeed, at only 4GHz, it's unlikely you'd need something as good as these
examples, though they would mean lower temps & hence longer life.

Once you get rid of the stock heatsink, you should see much better temps.

Use the excellent instructions on the Arctic Silver website for applying
thermal paste.

Ian.

 
My experience with Asus OC Tuner has not been very favorable in the past so I uninstalled it. I just use it to monitor my Asus MB voltages and temps. My MB is a P5N-D with an Intel 6850 using Win7. In fact after I read this article I downloaded and installed the latest version from Asus's site. From that point on it turned into a night mare!! I though I was going to have to start from scratch with having to reload my OS etc..
After a couple of hours of trying to do restores etc. I was able to bring it back from the dead. I now have a note on my monitor to remind me NEVER use Asus OC on my system.
 
[citation][nom]usafang67[/nom]My MB is a P5N-D...[/citation]And it still works? All of my P5N's died within weeks of setting them up and were never great to begin with, it's an apples-to-oranges comparison.
 
[citation][nom]Chrismvts[/nom]Hi,Ive just had a computer made for me it has the i7 2600K processor and the builder has clocked it to 4gh. Im getting about 47C - 51C sitting on windows desktop screen and when I play Witcher 2 on ultra graphics and 1920 x 1080 res the prossesor goes up to 97C. This to me seems extremely high. Ive just bought 3 more fans to bllow air into the cse (Antec300). I think the i7 2600K has the standard heatsink and fan on it. My graphics card is the Nvidia GTX 590. [/citation]

The stock cooler is okay, but hardly effective at keeping SNB at a reasonable temp for longevity, even less effective if you're overclocking it, for that 4GHz you should only be seeing temps in the upper 60s(if that even) with a good air cooler like the Hyper212+/EVO. My 2600k at 4.7Ghz@1.368-1.382vcore never goes above 75c and rarely ever goes higher than 69c with my Hydro H80 under load and idles at 36c.

You'll also want to check what sort of voltage is being pushed through the chip(can use CPU-Z for that). 4Ghz shouldn't require more than 1.3vcore, many of the SNB i5/i7 chips can do as much as 4.5Ghz at stock voltage and minimal BIOS settings changes with out missing a beat.
 
Manual overclocking is like a fine art where everything is perfected. High clock speeds are achieved through carefully moving things up, and once the desired clock speed is met we reduce the voltage to as low as we can get it. Automatic overclocking usually adds unnecessary additional voltage.
 
I have an Asrock z77 Extreme4 with an IB 3570K, and I agree with the results found here. The settings for the auto-OC are within 0.04V of what I could achieve manually. It doesn't get much better than that.
 
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom