Phenom II 920 goodness thread

[zif33rs]

Good to see nice clocks from PhII 920!
It looks like 2.5-2.7GHz NB will be a common thing with these.

Am I right that you're running Vista x64 zif33rs?

Yes..running Vista 64bit..and it could probably do with a fresh install..this has all my 9950be kludge on it.

I'm assuming he meant in super pi.

What type is the G. Skill if it's pi black you should be able to get pretty tight timings.

Yea...its these. Got them folding im my 6400be rig atm..have to yank em out.

http://www.newegg.com/Product/Produc...82E16820231209

A couple more SPi runs.

1M

[zif33rs]

and 32M..almost sub 20 minutes...

[Titan7171]

voltage seems a little high for 3.7 but great oc none the less.

[Sparky]

voltage seems a little high for 3.7 but great oc none the less.

1.45V high for 3.7GHz? C'mon man, my q6600 is taking 1.475V to run 3.2GHz so I don't think it is high at all

[demonkevy666]

1.45V high for 3.7GHz? C'mon man, my q6600 is taking 1.475V to run 3.2GHz so I don't think it is high at all

what it's highest temp on air at those voltages do you know ? ( I know your water now I bett it was hot as hell on air)

I think lap mine phenom II down to copper when I get one.

[Daveburt714]

zif33rs: That looks GREAT man! Gratz...

Does the HT Ref clock seem to go higher using the PhII? I pushed mine to the limit on the Asus 79-T and could only get 250 with stability (was just testing for limits). If anything is keeping you from going higher that may be it.

Personally I only up the HTRef to tweak the memory and use multis for everything else, but those are some pretty impressive results for a locked chip!!

Can't wait for the 940's...

[Stuen4y]

1.45V high for 3.7GHz? C'mon man, my q6600 is taking 1.475V to run 3.2GHz so I don't think it is high at all

Q6600 is 65nm .
P.S. @zif33rs, that are good clocks now, good job

[Titan7171]

1.45V high for 3.7GHz? C'mon man, my q6600 is taking 1.475V to run 3.2GHz so I don't think it is high at all

look at the first pic above 1.52v for 3.7ghz 1.456 volts was for 3.65ghz........that makes all the difference if it takes 1.52 for 3.7.....at that rate it will take 1.7v to hit 4ghz

Also I dont have any problem with high voltages Im running 3.3ghz at 1.48v but the thing Im saying is if we want to hit 4ghz what voltage will it take....the lower the better......my goal is 3.8ghz for 24/7 and a high benching speed of 4ghz.

[superpyton]

and 32M..almost sub 20 minutes...
(picture)

NB clock @ 2610MHz... wow .

[Sparky]

what it's highest temp on air at those voltages do you know ? ( I know your water now I bett it was hot as hell on air)

I think lap mine phenom II down to copper when I get one.

My q6600? Never been on air while I've had it. It is 50's full load sharing a single loop with a radeon 4870 on an MCR320.

Q6600 is 65nm .
P.S. @zif33rs, that are good clocks now, good job

True. But IIRC AMD's 45nm isn't near so sensitive as intel's for whatever reason.

[marten_larsson]

http://www.xtremesystems.org/forums/...d.php?t=213200

Hadn't seen that. But it's clear that a 3,6GHz Phenom II will beat a 3,2GHz Kentsfield if you look at all of those numbers too. Considering that the Q6600 beats the Yorkfield and Nehalem in several benchmarks too makes it quite obvious that not even Intel has progressed all that much. I knew Kentsfield was great but wasn't aware of that it still beats Intels own top processors at same speed (in some benchmarks). This rather shows that Kentsfield is "one of a kind" rather than Deneb being bad.

(BTW, have a Q6600 myself and no AMD systems for now.)

[villa1n]

and 32M..almost sub 20 minutes...

Very nice Work. How hot is your board getting with the HTT at 260+, do you have only the cpu under your water loop? and your doing this on the Board in your sig.. the gigabyte ds4?

[duploxxx]

at 3.65ghz, 2600mhz NB 1.45vcore, 1.4v NB

looks like BE is not needed at all for normal air cooling configs.... tempting to buy one next week and not waiting for am3 procs.

[Caveman787]

I think he just hasn't dialed in the voltage yet since he's been doing steps so far slowly increasing the clocks.

Isn't 1.4v nb really high? specially when I thought it was like 1.12 or something stock.

Makes me fear for the mosfets and northbridge.

[zif33rs]

what it's highest temp on air at those voltages do you know ? ( I know your water now I bett it was hot as hell on air)

Temps from what I can tell havent peaked above 27c on the cpu die..for some reason im not getting any core temp readings.

zif33rs: That looks GREAT man! Gratz...

Does the HT Ref clock seem to go higher using the PhII? I pushed mine to the limit on the Asus 79-T and could only get 250 with stability (was just testing for limits). If anything is keeping you from going higher that may be it.

Personally I only up the HTRef to tweak the memory and use multis for everything else, but those are some pretty impressive results for a locked chip!!

Can't wait for the 940's...

Seems The HT is bound to the MB maybe getting alittle higher with the Phenom II but as of now I'm having a hard time getting over 270

Q6600 is 65nm .
P.S. @zif33rs, that are good clocks now, good job

Thanks Stuen4y

look at the first pic above 1.52v for 3.7ghz 1.456 volts was for 3.65ghz........that makes all the difference if it takes 1.52 for 3.7.....at that rate it will take 1.7v to hit 4ghz

Also I dont have any problem with high voltages Im running 3.3ghz at 1.48v but the thing Im saying is if we want to hit 4ghz what voltage will it take....the lower the better......my goal is 3.8ghz for 24/7 and a high benching speed of 4ghz.

Yea..its taking me abit to dial this in..right now I'm at 3.7ghz at 1.475v. Seems to be pretty stable..havent primed yet...running some vantage and cinebench now.

NB clock @ 2610MHz... wow .

Havent found max NB speed yet...tuned NB speed down to 2400mhz to match HT link speeds for now.

Very nice Work. How hot is your board getting with the HTT at 260+, do you have only the cpu under your water loop? and your doing this on the Board in your sig.. the gigabyte ds4?

Thanks villa1n, the board is staying VERY cool compared to my 9950. Max system temps are 34c

I think he just hasn't dialed in the voltage yet since he's been doing steps so far slowly increasing the clocks.

Isn't 1.4v nb really high? specially when I thought it was like 1.12 or something stock.

Makes me fear for the mosfets and northbridge.

My mother board has two setting for NB voltages. One seems to be direct to the NB(set at 1.4v...up from 1.3v default) and the other is NB VID. This is default at 1.175v. I have added up to .05v here but it is VERY sensitive. I think this is the voltage causing people to blow MB's

Thanks every one for your interest..Post any Benches you would like me to try out and I'll see what I can do.

p.s. First Vantage run at 3.64ghz

[villa1n]

I think he just hasn't dialed in the voltage yet since he's been doing steps so far slowly increasing the clocks.

Isn't 1.4v nb really high? specially when I thought it was like 1.12 or something stock.

Makes me fear for the mosfets and northbridge.

Yes, this was my concern as well, as I am considering grabbing and building one of these in the next couple weeks.. but not sure on what mobo's will be able to support high htt clocking.

Edit: zif33rs, you just posted as i posted this... 34C sounds great, doesn't sound like much chance of burned mosfets or northbridges with those temps

[Brother Esau]

voltage seems a little high for 3.7 but great oc none the less.

Seems a little high because 64bit OS requires more voltage then 32bit

[informal]

voltage seems a little high for 3.7 but great oc none the less.

The default vcore for Phenom IIs is 1.35V ,so 1.47V is just 8(eight) percent above stock .

[AbelJemka]

I heard AMD says no more than 1.55V for aircooling so not to be worried at 1.47V.

[informal]

I heard AMD says no more than 1.55V for aircooling so not to be worried at 1.47V.

Yep that's roughly the upper Vcore border for the air cooling.Phenom IIs like cold much more than voltage,so if user can provide a good cooling device,4Ghz could be done with much lower volts than 1.55V.

[WarEagleAU]

Wow, the temps are very nice and impressive to reach those speeds at stock voltage. Seems the new deneb chips are doing pretty well.

[Titan7171]

The default vcore for Phenom IIs is 1.35V ,so 1.47V is just 8(eight) percent above stock .

I know I was refering to the 1.52v in the first post on this page

Seems a little high because 64bit OS requires more voltage then 32bit

Ahh didnt know it was 64bit......the cpu is doing really well then, this looks very promising and I cant wait

[Titan7171]

would you run the X3 terran conflict rolling demo bench?? http://downloads.guru3d.com/X3-Terra...load-2115.html

[zif33rs]

Cinebench R10 64bit

There are 2 scores...ran it with the HT link at 7x(about 1830mhz) the first pass and then (9x the second), NB speed was the same...the scores are interesting. The OpenGL and multi cpu scores were higher with the higher HT speed. The single cpu pass was lower with increased HT speed. Not sure if the difference is within the margin of error but maybe.




Sure Titan..give me a few to download it and install...

[HondaGuy]

Nice Cinebench....Phenom II 940 here :58 seconds here on M3A32-MVp Deluxe