• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

**LETS SEE YOUR PILEDRIVER OVERCLOCKS - LET ME START WITH 5GHz+!!**

My 8350 at 5.04GHz will run one pass of x264 then locks up part way through second pass. Does not seem to be throttling but temps were getting high, 63C core temps at 1.48-1.5V. Probably need to reduce vcore. 5Ghz seems fine in windows though, the lockup came when using mwsnap for capturing screen with cpu-z, hwmonitor, and x264 also running. Will try again later.
 
I have been experimenting a bit and have found that the only way to keep socket temp below 62C is to run with a reasonably low voltage (1.3-1.35V).

If you aren't interested in 100% stability (eg. Prime95) then I suppose the sky is the limit as they do run fairly cool under normal usage, stress all of the cores though and even with only ~1.4V it's like trying to contain a star going nova. :p

4.2ghz at 1.32V at the moment with FX8320, it's only going to be running Linux so I'm not too fussed about gaming performance.
 
Last edited:
Superpi is fine for comparing like with like. It's no good for AMD vs Intel.

Think of Superpi like a dragstrip race: what vehicle can go from start to 1 mile the quickest?

If you compare Intel vs AMD that could be like a 2 wheeled motorbike vs a 4WD SUV vs a dragster. All of them may have a top speed of say 180Mph / 5Ghz but one will be able to hit the 1 mile mark the quickest.

Superpi would be good for measuring the difference between piledriver at stock and at 5Ghz - you can see the percentage performance increase. It would also be good for measuring the difference between bulldozer and piledriver at the same clock speed. But for AMD vs Intel it's not the right test to perform.

I still rely on Superpi for determining what platform to work on. For the stuff I do I need a setup that can hit the 1 mile mark the quickest. Encoding, crystalmark, FPS mean nothing to me.

It's coming up to two years since I got this Sandybridge rig. Looking at what is available from Intel now there only seems to be a ~15% gain over the 2500K so there's not much point in updating.

I don't think I could go back to AMD. Piledriver would have to hit 10Ghz to match the Superpi speeds of Intel.
 
Frank_Rizzo said:
I still rely on Superpi for determining what platform to work on. For the stuff I do I need a setup that can hit the 1 mile mark the quickest. Encoding, crystalmark, FPS mean nothing to me.
Click to expand...

This is an AMD overclocking thread not one for some AMD vs Intel thing.

Moreover,superPi uses ancient x87 instructions which are irrelevant to most people and almost all modern applications. It is so irrelevant that BOTH Intel and AMD tell people to use SSE at least.
 
Last edited:
CAT-THE-FIFTH said:
This is an AMD overclocking thread not one for some AMD vs Intel thing.
Click to expand...

Yes. Fully aware of that. I was responding to the earlier posts, which compared the 18s AMD with 8s for Intel. My point is that you can not do that much in the same way you can not compare a motorbike 0-1 mile with a 4WD SUV 0-1 mile.

And even though Superpi is x87 it can still be used for a like for like test. If someone runs Superpi on a Piledriver at stock and someone else runs Superpi on a Piledriver @ 5.2Ghz then that is a perfectly valid metric to use for comparison. Similarly Superpi for Bulldozer vs Piledriver.

It is relevant to this thread.
 
That's just stupid. if it can't be used as a yard stick for everything that's available and how modern software will perform on hardware it makes it even more irrelevant. We don't need this old x87 app to tell us what modern apps can tell us, and a whole load more. Jeez, what is it with everyone and this ancient piece of software? Move on already, this is 2012 for goodness sake. :rolleyes:
 
I got a 6300 and it seems a great clocker but im being held back by my board!

I have a GA-990FX-UD5, with probably the worst vdroop ever with no llc setting for the v1.0 boards!

Currently prime95 for 4 hours @ 4.3Ghz with 1.475v in bios (1.395v while prime running).

I think im going to struggle to get 4.5Ghz but thats my possible goal. I have it on a scythe mugen 2 cooler, which is doing okay (could be better!)

Temps are around 50-53 on cores, 52 on cpu (not sure on temps yet...seems odd) vrm's around 60.
 
Hate to be pedantic but isn't that a 3% increase? a win but hardly a convincing win especially when its a 4/8 V 8/8 configuration in a multi-thread heavy benchmark even taking the extra 200MHz into account.
 
Rroff said:
Hate to be pedantic but isn't that a 3% increase? a win but hardly a convincing win especially when its a 4/8 V 8/8 configuration in a multi-thread heavy benchmark even taking the extra 200MHz into account.
Click to expand...

Is that for me? its about 3 or 4% yes, but i'm not sure what your saying about the 200Mhz, the 2600K is clocked 200Mhz higher at 5.2Ghz vs 5Ghz on the FX-8350.

Its quite clearly faster in this situation, i'm simply saying it goes to show what it can do.

And its not 8/8, its actually 8/4. :)

you have to admit, that ain't bad, that ain't bad at all.
 
Last edited:
True but overly simplified the AMD cores have 2 seperate logic blocks per real core whereas intel have essentially an advanced scheduler - on paper the AMD solution should give it a much bigger edge.

It ain't bad but I think its a bit far calling it a convincing win atleast to my mind a convincing win is in the region of 15-20% gains.
 
Rroff said:
True but overly simplified the AMD cores have 2 seperate logic blocks per real core whereas intel have essentially an advanced scheduler - on paper the AMD solution should give it a much bigger edge.

It ain't bad but I think its a bit far calling it a convincing win atleast to my mind a convincing win is in the region of 15-20% gains.
Click to expand...

Well, the 3570K is only 10% ahead of the FX-8350 in those few games where Intel is faster so that works both ways.
 
You must log in or register to reply here.
Back
Top Bottom