X56xx locked Uncore has been unlocked

[Sam_oslo]

The locked x20 Uncore needs a lot of VVT, and has always been the limiting factor for OCing my X5650. The latest BIOS-F7f for GA-X58A-UD7 (which adds Intel's new micro code patch) has unlocked it

I've just flashed to this new BIOS, and haven't tested the stability yet, but I couldn't even log into w7 @4GHZ with VVT<1.38v. Now i can sett the Uncore to 16x and it seams really promising. Going to test further and report back.

X5650 @4.31GHZ(196x22) 1.34v , VVT=1.30v

[tool_462]

Darn, don't think the latest beta BIOS for UD5 unlocks it Good to know it can be done though! Hopefully soon.

Any higher max stable OC?

[Movieman]

Interesting but Sam,whats' with that SP1m time?
Man that is BAD!

[Sam_oslo]

Interesting but Sam,whats' with that SP1m time?
Man that is BAD!

Yes indeed but that's the hyperPI1M-time. Still is a bad time, but nothing is optimized. I think the MEM was acing strange under that run, and I need to work more on it. It was just a quick and first run.

Darn, don't think the latest beta BIOS for UD5 unlocks it Good to know it can be done though! Hopefully soon.

Any higher max stable OC?

I've always had faith this would happen one day. I haven't tried max yet, but got right into this with VVT=1.32v . Voltages are not optimized, still testing.

[Movieman]

Sam;
Ahh ok on SP1m..
I'm happy for you pal,and only you understand what I mean.

Oh, and fixed your title..
Not 65xx but 56xx...

[Sam_oslo]

Sam;
Ahh ok on SP1m..
I'm happy for you pal,and only you understand what I mean.

Oh, and fixed your title..
Not 65xx but 56xx...

I know what you mean pal .

I guess I got too excited for sharing the good news . A superPi1M-run just for you for fixing the tittle

[Movieman]

Very nice!

[=[PULSAR]=]

What cooling do you have on your 5650's now oslo?

[Sam_oslo]

What cooling do you have on your 5650's now oslo?

On air, Noctua NHD14 using the stock fans a NF-P12 + a NF-P14 push/pull.
Just a quick Prime-run to test the temp, and it looks promising, but I have to test and optimize more.

X5650 @4.2GHz(210x20) 1.32v , VVT=1.30v

[=[PULSAR]=]

Looks like 4.4ghz should be easily attainable on water

[Sam_oslo]

Looks like 4.4ghz should be easily attainable on water

I believe so , and maybe even more. The VVT-limitation is gone and the only limitation seams to be vCore/cooling now.
These fast tests shows it can run SuperPi32M @4.4GHz(220x20) 1.38v and VVT=1.34v . It seams to be stable too, but I didn't have time to test it throughly. I guess the same VVT=1.34v could do @4.84GHz(220x22) too, but it needs more vCore and one needs really good cooling for that.

[=[PULSAR]=]

After my 980x dying with 1.4vcore and stock vtt I'll probably keep my limited to 1.35vcore on water.

[Sam_oslo]

Yeah, I see what you mean. I personally will be really happy with anything over @4GHz for 24/7 use. This 6C/12T has been a monster even @3.8GHz.

It is fun to find the limits too, but The 24/7 OC is more about finding the sweet spot, than the edge. It is not everyday one discovers @4MHz extra OC and then it is fun to play around for short runs.

[Blauhung]

After my 980x dying with 1.4vcore and stock vtt I'll probably keep my limited to 1.35vcore on water.

hehe, I'm glad I backed mine off them. was running 4.5GHz @ 1.425v and 1.35Vtt. Now I'm down to 4.32GHz and 1.375Vcore, probably can lower the voltage too.

[jcool]

Great news, now we need this unlock on other boards as well.

[Sam_oslo]

Yeah, this is really good news. This happen when they added Intel's new microcode patch for 32nm to the BIOS-F7f for GA-UD7, so I guess it will happen to other MBs too, soon hopefully.

There is more good news. According to CDIPD HM the CPU is using only 77.5W after disabling the turbo , it was reporting 113W with turbo.

I'll try to get this, or something close to it 24/7 stable, with VVT=1.32v

[cdolphin]

I am dubious of those figures, the TDP is 120W I thought?

[Movieman]

I am dubious of those figures, the TDP is 120W I thought?

TDP I beleive on a X5650 is 95w
130W on the top X5680

[Sam_oslo]

Yeah, the TDP for X5650 is 95w, at stock with turbo.
Turning off the turbo will always reduce it, but 77.5W with vCore=1.28v(load) is actually too good to believe, I hope and believe it is right tho.

[Sam_oslo]

Have been testing the TDP under prime-load, and interestingly it is less with x19 multi than x20, at the same frequency and same vCore.

4.1GHz (216x19) 1.28v , CDIPD HM reporting =77.66W
4.1GHz (205x20) 1.28v , CDIPD HM reporting =95.00W

Just couple of short runs to see the TDP under prime-load:

4.1GHz (216x19) 1.28v , 77.66W


4.1GHZ(205x20) 1.28v , 95.00W

[Sam_oslo]

And BCLK221 with VVT=1.34v , so far so good

[jcool]

You know, I'm still running an old EX58-Extreme with my X5680. There is a test bios (F13c) from April 20th that says "Update CPU Microcode"... think it may unlock my uncore mult as well?
Not that I'd need it, of course.. still, curious!

[Movieman]

You know, I'm still running an old EX58-Extreme with my X5680. There is a test bios (F13c) from April 20th that says "Update CPU Microcode"... think it may unlock my uncore mult as well?
Not that I'd need it, of course.. still, curious!

Go fo rit..and I wonder if the newest bios will work on the GB X 58 UD4 I have here with a X5677 in it..

[Sam_oslo]

You know, I'm still running an old EX58-Extreme with my X5680. There is a test bios (F13c) from April 20th that says "Update CPU Microcode"... think it may unlock my uncore mult as well?
Not that I'd need it, of course.. still, curious!

Yeah, give it a shot. Hopefully they got it.

On UD7, they said the microcode was added before, but somehow it wasn't exactly successful. They got it right with BIOS-F7f (released 16.April...) finally, and it says "Update CPU Microcode"... too. Your BIOS date is indication that it will unlock the Uncore, hopefully.
The new microcode makes other improvements to 32nm too, so it is a good idea to get it, even if your X5680 doesn't need unlocked Uncore.

[Sam_oslo]

Have been testing with x16 Uncore until now, but since the Uncore is unloked and 32nm has introduced the Uncore>=1.5xRAM rule, so why not trying x15 Uncore?
It got right into BCLK221 with VVT=1.30v

RAM x8
Uncore x15
I'm still testing and don't know who stable it really is, but so far so good:

@4.2GHZ(221x19) 1.31v, VVT=1.30v