Could B3 Phenom be the rebirth of AMD....+1000MHZ OC done!

[justapost]

Even a 12x nb multi would be nice. Also 3.2GHz at just 1.3V is cool, id still be ok if one whould need 1.4V for that. I wonder if watercooling is making a difference now.
Any success with HTT ? I smell 300.

[i found nemo]

hey tony, you never mentioned the cpu stepping.... maybe next time you swtich cooling, or try a new block.... write it down? lol. but hey, just stoppin' in to say good job and thanx for contributing to the community; and as usual GREAT results.

[Lightman]

LOL...i wish

its fun to tweak AMD's CPU's, I get bored of tRD etc on intel.

I wonder how many I sold today ?

One for me please!

Damn holiday! Why, oh why they weren't available yesterday!


I will get one as soon as I'm back from Poland! (or tomorrow if I will find £20 to finance Saturday shipping! )

BTW Tony congrats on new AMD Phenom OC WR!

[keithlm]

I'd just be happier than a pig in e if I can get 3.2 on high end air with a good NB speed!

I don't personally overclock... I might play around a bit and I do enjoy doing that... but I'd never leave it at the higher speeds. I always end up back at stock. (Or perhaps a little bit higher.)

But I can see how others might get excited.

Actually I'd be excited to see 3.2Ghz - 3.4Ghz verified. That will take away the current "party" line... the one where people bemoan the Phenom and the fact that it can't even reach 3.0Ghz. (Of course that "party line" won't disappear... it'll just be higher... just above what the highest verified overclock is.)

I'm already starting to see the next "iteration" of fanboy arguments: people adamantly claiming that the Phenom at 3.2Ghz can't keep up with an Intel Q6600 at 3.2Ghz. And some others starting to say: "Who cares... it's less than 3.6Ghz and EVERYONE gets that with an Intel."

I so want to order a 9850... and DFI 790FX... RIGHT THIS MINUTE... <insert blood curdling scream><then whimpering turning to pathetic sobbing>

[Tony]

you smell 300 and thats all you will do. I stopped for the day and will resume tomorrow.

Overall I think HTT clocks have not increased much over what i had out of a B2 which was around 240 to 250. I don't think Phenom really has the ability to push HTT like K8 did, and with BE CPU's you don't need it being honest.
The fact is I now have 1066 working well on 2 boards even with 2x2GB dimms, the new B3 processor allowed me to use a lower vdimm on the 8000Reaper X over the old B2 and im confident 1066 improves performance enough to warrant releasing a special product for this processor.

I will have a play with HTT and see what the limits are with the GBT board, my DFI seriously has an issue and I will look to swap it and retest as i do feel its an individual board issue. It clocks with multi's very well though

[justapost]

Nevermind, I ordered one.
In 1066 mode we are limited by the imc to 4-5-5-15 timings, for better timings 800 mode at 300MHz would be nice I think. Lower Vdimm with B3's is actualy good news btw.

[MonsterMash]

Great result! I'm curious to see how it scales in performance with the increase of the NB clock. Can you make a 3dmark06 cpu test?


Any of you know if it's possible to remove the HIS from a phenom? Some peaple told me that it will easily pick up if heated at around 100 °C.

Thank you, and sorry for my bad english

[NaeKuh]

Whats wrong with a sub 13second SuperPI 1m Run?

Tony is it not stable enough for one? It shouldnt take more then 30 sec tops.

Just want to see some numbers on how they scale up there. :T

[SeLecT]

Very nice work Tony!

I hope most of the chips will run like yours

[Tony]

Whats wrong with a sub 13second SuperPI 1m Run?

Tony is it not stable enough for one? It shouldnt take more then 30 sec tops.

Just want to see some numbers on how they scale up there. :T

you know everyone wants me to do this and that, the exercise was to get 1066 running well...not run pi etc

I don't use pi, see no need for it, so won't be posting it...sorry m8

[Charles Wirth]

It takes what 30 seconds for a 1M? Sure it looks horrid compared to Intel but its not about that. People just want to see where its at in terms of single thread performance not hear about how you hate pi all of a sudden...

BTW, good showing for AMD finally, moving in the right direction.

[Tony]

http://valid.x86-secret.com/show_oc.php?id=339861

Validation for 3.5GHZ, this time on GBT DQ6

Charles I have 0 interest in superpi, people think it matters..I do not.

If other's want to show it they can, i will not fuel the debate about why Intel is better when using pi.

Sorry...its just how I feel and I hope you respect my feelings.

[Charles Wirth]

NP, thanks for the initial results.

Hopefully you reach 4Ghz + on phase.

noticing the core temps are all exact, if you load up two cores do they change?

[Jethro]

Grats on phenomenal chip!! Wowsor's! AMD fan's rejoice! Theres hope

[JumpingJack]

NP, thanks for the initial results.

Hopefully you reach 4Ghz + on phase.

noticing the core temps are all exact, if you load up two cores do they change?

Unfortunately, no.... core temps are not reproted from the DTS on phenom, only from the analog package sensor, and it is not calibrated:

http://www.alcpu.com/CoreTemp/howitworks.html

At least according to the core temp author:

AMD chips report the temperature by a special register in the CPU's NB. Core Temp reads that register and uses a formula provided by AMD to calculate the current temperature.
The formula for the K8 is: 'Core Temp = Value - 49'.
The formula for the K10* is: 'CPU Temp** = Value / 8'.

The sensor in AMD CPUs can report temperatures between -49C and 206C.

*K10 = Phenom (Agena), Opteron (Barcelona). The K10 reports a temperature value that is relative to a certain predefined value, it doesn't report the actual processor temperature! So take that into consideration.
**CPU Temp is because the Phenom\Opteron (K10) have only one sensor per package, meaning there is only one reading per processor.

To Tony ... kudos on the highest observed Phenom clock yet.

[JumpingJack]

hey tony, you never mentioned the cpu stepping.... maybe next time you swtich cooling, or try a new block.... write it down? lol. but hey, just stoppin' in to say good job and thanx for contributing to the community; and as usual GREAT results.

It's a 9850, thus B3 stepping.

[Tony]

CPU score 3D06, this is in vista64bit with 2GB ram.

[aGeoM]

It's a 9850, thus B3 stepping.

Revision=B3; Stepping= eg. CAAZB AA 0743APCW

[Loser777]

Hmm, does the DS5 clock as well as the DRI M2R or M2RS then?

[Tony]

3d06 full run, driver set to high performance...only tweak. Only gained me like 20 points

Ultra at stock clocks...no pcie clocking.

[Boschwanza]

CPU score 3D06, this is in vista64bit with 2GB ram.

Pretty nice so far, try to optimize memory timings+subtiming this will boost up performance and if u have time the cinbench10 score would be nice to see.

[Loser777]

You using the M2RS or M2R for 3.5GHz?

[i found nemo]

It's a 9850, thus B3 stepping.

uhhh ..... i mean more like what is in my sig ( cabze 0631 cpmw ) why would i need him to take off the block to see what cpu-z tells me? lol

[NickS]

superpi? we needz it. share your pies.

edit: 3.5ghz "almost stable," 3.2ghz 3d?

[Tony]

You using the M2RS or M2R for 3.5GHz?

you asking about the board?..MA790FX-DQ6