The Official X2 Overclocking Discussion Thread

Printable View

Show 100 post(s) from this thread on one page

06-22-2005, 02:56 PM
ewitte

Quote:

Originally Posted by p0tempkin

CCBWE 0517MPMW

I've got a CCBWE 0519UPMW sitting here. Results later tonight.
06-22-2005, 03:35 PM
s7e9h3n

Quote:

Originally Posted by joe2004

Yup, that is a good tool for stability. I'd still like one SuperPI to run multithreaded so that I can compare the results with singlethreaded scores. That would give me better idea about the performance gain. Sandra's synthetic bench is quite useless (exaggerated).

Courtesy of JNAV89GT:
http://xtremesystems.org/forums/atta...chmentid=32678
06-22-2005, 09:51 PM
PTK

Very promising oc's you guys had.

Waiting for my 4400+ ... if i can hit 2700+ with water im a happy camper.

PTK
06-23-2005, 02:25 AM
Agent-JCDenton

Quote:

Originally Posted by joe2004

Yup, that is a good tool for stability. I'd still like one SuperPI to run multithreaded so that I can compare the results with singlethreaded scores. That would give me better idea about the performance gain. Sandra's synthetic bench is quite useless (exaggerated).

I'll be getting my X2 in a week or two, I want to see if systool SuperPi times when run in 2 instances are significantly different from just 1 instance. I don't think the Systool SuperPi is a true multithreaded application. Cinebench is a good tool if you want a quick benchmark for a multicore setup, but its not exactly an indication of stability. Now, if only SuperPi + Cinebench could be merged....any coders out there? :)
06-23-2005, 02:37 AM
NickK

Just received notification that my X2 4400 is being shipped - should have it tomorrow.

I'll code up something - last time I did that it took 12 hours to complete.. Although it's written in Java it'll use both cores.
06-23-2005, 02:38 AM
[timko]

You lucky bastage NickK :D

I hope it does work "out of the box" for you :)

Don't forget to get the stepping etc info from the core first before putting it in.
06-23-2005, 02:43 AM
Agent-JCDenton

Anandtech have written an article on thier results from overclocking the 4200+ on air. As shown below, thier highest stable OC was 2695MHz

http://images.anandtech.com/reviews/...4200/a64x2.jpg

Performance Test Configuration

Processor AMD Athlon64 x2 4200+ (2.2GHz 512KB cache each) 939
RAM 2 x 512MB OCZ PC3200 Platinum Rev. 2
Power Supply OCZ 520 watt PowerStream
CPU Cooling Thermaltake Silent Boost K8 HeatSink/Fan
Hard Drive Seagate 120GB 7200 RPM SATA (8MB Buffer)
Video AGP & IDE Bus Master Drivers nVidia nForce 6.39
Video Card nVidia 6800 Ultra (PCIe)
Video Driver nVidia nForce 71.89
Operating System Windows XP Professional SP2; Direct X 9.0c
Motherboard DFI LANParty nF4 SLI-DR

Front Side Bus Overclocking Testbed

Processor Athlon 64 x2 4200+ (2.2GHz, 512KB Cache each core)
CPU Voltage 1.55V (default 1.30V)
Cooling Thermaltake Silent Boost K8 Heat sink/Fan
Power Supply OCZ Power Stream 520W
Memory OCZ PC3200 EL Platinum Rev. 2 (Samsung TCCD Memory Chips)
Hard Drive Seagate 120GB 7200RPM SATA 8MB Cache
Maximum OC
(Standard Ratio) 245x11 (Auto HT, 2.5-3-3-7, 1T, 2.8V) 2695MHz (+22.5%)
Maximum FSB
(Lower Ratio) 326x8 (2608MHz) (Auto HT, 2.5-4-3-7, 2.9V) (1:1 Memory, 1T, 2 DIMMs in DC mode) (+63% Bus Overclock)

http://images.anandtech.com/reviews/...4200/326x8.jpg

"In the end the 4200+ appears to be a good-performing dual-core CPU with quite a bit of overclocking headroom. We reached 2.7GHz with a PC Health reported CPU temperature of 61 degrees C at 1.55V. 240 was a breeze at 1.45V, exceptionally stable for days on end, with processor temps generally below 50C with our air cooling. It appears an easy task to reach the highest levels of Dual-Core performance with the cheapest 4200+ if you are willing to overclock a little - and the 4200+ is up to the task." - AnandTech
06-23-2005, 02:44 AM
ewitte

So far systool gives me a max of 2839Mhz at only 1.5v BIOS :banana: I had to pump 1.7v BIOS into my San Diego to get it stable at 2.81Ghz. I'll probably push her up to 1.6v max.

Eric
06-23-2005, 02:46 AM
Agent-JCDenton

Quote:

Originally Posted by NickK

Just received notification that my X2 4400 is being shipped - should have it tomorrow.

Yeah, you lucky bastage. :)

Quote:

Originally Posted by NickK

I'll code up something - last time I did that it took 12 hours to complete.. Although it's written in Java it'll use both cores.

Nice one NickK! Looking forward to running that
06-23-2005, 02:50 AM
Agent-JCDenton

Quote:

Originally Posted by ewitte

So far systool gives me a max of 2839Mhz at only 1.5v BIOS :banana: I had to pump 1.7v BIOS into my San Diego to get it stable at 2.81Ghz. I'll probably push her up to 1.6v max.

Eric

The CPU Overclocking feature in Systool is great, think i'll be using that from now on for OCing. 2839Mhz at only 1.5v uber nice. How many digits were you calculating in SuperPi?
06-23-2005, 02:56 AM
ewitte

Quote:

Originally Posted by Agent-JCDenton

The CPU Overclocking feature in Systool is great, think i'll be using that from now on for OCing. 2839Mhz at only 1.5v uber nice. How many digits were you calculating in SuperPi?

It was a quick test. I ran it with it set to 1million. I've backed down 25Mhz and am trying some benches.
06-23-2005, 03:08 AM
[timko]

Nice result ewitte!

I think I'll count myself lucky if my 4400 can get 2.7Ghz when it arrives :)
06-23-2005, 03:33 AM
ewitte

This is odd. It locked up doing multi CPU Cinebench. I upped it to 1.55v and windows will not even finish loading. Went back to 1.5v and it loads fine! Right now trying 2.75Ghz @ 1.475v. That was good enough for Cinebench to complete in 36 seconds (730 score). If I can get this thing to run stable 2.75Ghz 1.45v or so I'll be happier than 2.8Ghz @ 1.7v :) This thing is going to require a lot of testing. I'm off to work right now though.

Eric
06-23-2005, 04:14 AM
NickK

Quote:

Originally Posted by ewitte

This is odd. It locked up doing multi CPU Cinebench. I upped it to 1.55v and windows will not even finish loading. Went back to 1.5v and it loads fine! Right now trying 2.75Ghz @ 1.475v. That was good enough for Cinebench to complete in 36 seconds (730 score). If I can get this thing to run stable 2.75Ghz 1.45v or so I'll be happier than 2.8Ghz @ 1.7v :) This thing is going to require a lot of testing. I'm off to work right now though.
Eric

What's your system specs and BIOS version?
06-23-2005, 04:26 AM
ewitte

Quote:

Originally Posted by NickK

What's your system specs and BIOS version?

DFI Ultra-D BIOS 5/10-1 (6/18 gave me a LOT of problems). Oddly 5/10 said unknown processor to begin with but after reflashing it now it detects correctly.

2x512MB PQI Turbon PC4000 (TCCD) running 2.7v
ATI X800XL PCIe
Sound Blaster Audigy2 ZS

535W Enermax PSU (new one with two PCIe jacks)

2x200GB Seagate (1 IDE/1 SATA)
8x DVDR burner

Swiftech 6002 Waterblock CPU
MAZE4 GPU Waterblock
Bonneville heater core

Eric
06-23-2005, 04:52 AM
Agent-JCDenton

Quote:

Originally Posted by ewitte

This is odd. It locked up doing multi CPU Cinebench. I upped it to 1.55v and windows will not even finish loading. Went back to 1.5v and it loads fine! Right now trying 2.75Ghz @ 1.475v. That was good enough for Cinebench to complete in 36 seconds (730 score). If I can get this thing to run stable 2.75Ghz 1.45v or so I'll be happier than 2.8Ghz @ 1.7v :) This thing is going to require a lot of testing. I'm off to work right now though.

Eric

Sounds like its time to burn :)

http://www.ocforums.com/showpost.php...7&postcount=11

Use a program like CPU Burn as the guide says, but use 2 instances.
Moved my 3500+ from 1.7v down to 1.58v prime stable at 2.75Ghz
06-23-2005, 05:06 AM
ewitte

Quote:

Originally Posted by Agent-JCDenton

Sounds like its time to burn :)

http://www.ocforums.com/showpost.php...7&postcount=11

Use a program like CPU Burn as the guide says, but use 2 instances.
Moved my 3500+ from 1.7v down to 1.58v prime stable at 2.75Ghz

Getting that high with such little effort and VCORE is getting me excited. At least of getting a few tests ran at 3Ghz.

Eric
06-23-2005, 06:21 AM
Agent-JCDenton

Quote:

Originally Posted by ewitte

Getting that high with such little effort and VCORE is getting me excited. At least of getting a few tests ran at 3Ghz.

Eric

Anandtech noted, in thier article on overclocking the 4200+, that heat increased exponentially with voltages over 1.55v. Try find your max stable between 1.55-1.50 vCore and then burnin at those voltages to get everything you can out of the chip before moving onto 1.55+ vCore.
06-23-2005, 06:44 AM
ben805

looks like these X2 will need atleast WC to keep them cool at >2.6Ghz
06-23-2005, 08:40 AM
NickK

Quote:

Originally Posted by ben805

looks like these X2 will need atleast WC to keep them cool at >2.6Ghz

Which is unsupprising as they'll be drawing nearly twice the current.
06-23-2005, 12:52 PM
euclid

is there a real world advantage with the dual core? for the price of a 4800toledo you could buy an fx57 and a decent watercooling rig.
06-23-2005, 01:02 PM
ewitte

Quote:

Originally Posted by euclid

is there a real world advantage with the dual core? for the price of a 4800toledo you could buy an fx57 and a decent watercooling rig.

In certain apps it will kill an FX57. Although your better off getting a 4400+. They should all clock nearly the same (with the 512KB's having better luck). I don't see a reason why one would want to pay 80% more for nearly the same performance.

Eric
06-23-2005, 01:31 PM
s7e9h3n

Quote:

Originally Posted by NickK

Which is unsupprising as they'll be drawing nearly twice the current.

Not even close ;) :
TDP of 4600+ (2 x 2.4, 512K) = 110W
TDP of 3800+ (1 x 2.4, 512K) = 89W
X2 draws only ~24% more current
06-23-2005, 01:34 PM
bigjohns97

Quote:

Originally Posted by s7e9h3n

Courtesy of JNAV89GT:
http://xtremesystems.org/forums/atta...chmentid=32678

Do you know what he got when just running one instance of spi?
06-23-2005, 01:51 PM
[timko]

Quote:

Originally Posted by s7e9h3n

X2 draws only ~24% more current

Indeed, and here are XBitlabs' measurements...

http://img260.echo.cx/img260/2834/co...umption9rm.png

Dual core AMD cpus consume less power than the latest single core Intel CPUs :)

( http://www.xbitlabs.com/articles/cpu...on64-x2_4.html )
06-23-2005, 01:53 PM
s7e9h3n

Quote:

Originally Posted by bigjohns97

Do you know what he got when just running one instance of spi?

Not at that speed, but here's his spi1m single cpu @ 245:
http://xtremesystems.org/forums/atta...chmentid=32686
06-23-2005, 02:28 PM
joe2004

Quote:

Anandtech noted, in thier article on overclocking the 4200+, that heat increased exponentially with voltages over 1.55v.

Yes, you will not be able to control voltages over 1.6 with air cooling. It is way above 150W of dissipation therefore you would need phase change.
This is now the question how good the chips are and where can they go on default voltage. Or 10 lb heatsinks with 200 mm fans. :D
06-23-2005, 03:20 PM
NickK

Quote:

Originally Posted by s7e9h3n

Not even close ;) :
TDP of 4600+ (2 x 2.4, 512K) = 110W
TDP of 3800+ (1 x 2.4, 512K) = 89W
X2 draws only ~24% more current

Hmm - I looked for the AMD specs a few days ago, no mention of X2s in the power consumption charts.

If that's the case then you an air cool them but it'll be loud!

Had my old Barton at 2.03Vcore on air. IIRC that worked out close to 100W at the overclock it was doing.
06-23-2005, 03:26 PM
s7e9h3n

Quote:

Originally Posted by NickK

Hmm - I looked for the AMD specs a few days ago, no mention of X2s in the power consumption charts.

If that's the case then you an air cool them but it'll be loud!

Had my old Barton at 2.03Vcore on air. IIRC that worked out close to 100W at the overclock it was doing.

*sigh* thinking back to the good ol' days of 2.0Vcore on air *sigh*

http://img253.echo.cx/img253/3613/b031727069qp.jpg
06-23-2005, 04:38 PM
ewitte

Well it looks like with a little work I'd at least be able to get a 1M pi @ 3Ghz :) Just need to burn it in a little bit. I got 4 interations at 3.05Ghz before I got the not exact in round error. Of course that was at 1.7v and just a single instance ;) Temps still under 40C! I think that would change with both processors under heavy use though.
06-23-2005, 05:08 PM
s7e9h3n

Quote:

Originally Posted by ewitte

Well it looks like with a little work I'd at least be able to get a 1M pi @ 3Ghz :) Just need to burn it in a little bit. I got 4 interations at 3.05Ghz before I got the not exact in round error. Of course that was at 1.7v and just a single instance ;) Temps still under 40C! I think that would change with both processors under heavy use though.

Hhehehe, no need to "think" :p: , it most definately will. Trust me, you'll notice ;) Oh, BTW, 1.7Vcore - that's gotta take some serious huevos to try :p: I didn't even go that high with my phasechange and the 4600's.
06-23-2005, 06:04 PM
ewitte

Quote:

Originally Posted by s7e9h3n

Hhehehe, no need to "think" :p: , it most definately will. Trust me, you'll notice ;) Oh, BTW, 1.7Vcore - that's gotta take some serious huevos to try :p: I didn't even go that high with my phasechange and the 4600's.

Just temporary and single core tests. I actually ran S&M at 1.65v (both cores) and in a few minutes the highest I saw was 45C.

Eric
06-23-2005, 06:05 PM
s7e9h3n

Quote:

Originally Posted by ewitte

Just temporary and single core tests. I actually ran S&M at 1.65v (both cores) and in a few minutes the highest I saw was 45C.

Eric

Just because you're using one core doesn't mean that only one core's getting 1.7V :p:
06-23-2005, 06:40 PM
xgman

1 Attachment(s)

Progress so far with the 4800+ retail on water at 1.5v set in bios. I think there will be a bit more headway after I figure out a few settings. 32 deg idle temp and 37 deg during a very brief super pi run.
06-23-2005, 06:47 PM
ewitte

Quote:

Originally Posted by s7e9h3n

Just because you're using one core doesn't mean that only one core's getting 1.7V :p:

I just mean it keeps the heat down a bit. My San Diego was always running 1.7v (1.68 windows) ;)
06-23-2005, 07:03 PM
joe2004

Quote:

Well it looks like with a little work I'd at least be able to get a 1M pi @ 3Ghz Just need to burn it in a little bit. I got 4 interations at 3.05Ghz before I got the not exact in round error. Of course that was at 1.7v and just a single instance Temps still under 40C! I think that would change with both processors under heavy use though.

You need to run two instances of singlethreaded applications, like SuperPi for 32 MB and then check when it is a full load.
Post some screenshots, otherwise it is hard to believe what you are saying. The processor like X2-4600 at full load with 1.7V will develop 160W or more dissipation at 3 GHz and I do not know any air cooler that can handle that. But hey maybe you can surprise us with that magic thing of yours.
06-23-2005, 07:29 PM
ewitte

Quote:

Originally Posted by joe2004

You need to run two instances of singlethreaded applications, like SuperPi for 32 MB and then check when it is a full load.
Post some screenshots, otherwise it is hard to believe what you are saying. The processor like X2-4600 at full load with 1.7V will develop 160W or more dissipation at 3 GHz and I do not know any air cooler that can handle that. But hey maybe you can surprise us with that magic thing of yours.

I'm not running air. Its water. S&M runs on both processors and only bumped me to 45C at times. Although I was at 1.67v (1.64v windows) at the time.

Eric
06-23-2005, 07:45 PM
joe2004

Quote:

I'm not running air. Its water.

That makes sense. I guess that will be required for these beasts. Or phase change. ;)
06-23-2005, 09:16 PM
p0tempkin

Anybody testing for 24 hour Prime Stable clocks? I'm still stuck at 2.7GHz on air.
06-23-2005, 09:31 PM
jinu117

Quote:

Originally Posted by p0tempkin

Anybody testing for 24 hour Prime Stable clocks? I'm still stuck at 2.7GHz on air.

Core 1 stops @ 2.6ghz on air... core 2 stops @ 2.75ghz on air. figures... I am stuck wih 2.6ghz now -_-;

PS) I don't call anything stable unless it pass my testings... prime 95 for reasonable time (could be 8 hours i sleep, could be 24+ hours when I am away from home). Memtest86 and memtest windows stable for hours, 3dmark stable, than last but not least, hours of gaming sessions if i can shove them in.
06-23-2005, 10:09 PM
Ender17

Quote:

Originally Posted by p0tempkin

Anybody testing for 24 hour Prime Stable clocks? I'm still stuck at 2.7GHz on air.

what board?
and what chip?
cooling?
06-24-2005, 01:37 AM
NickK

It's arrived :D

Chip numbers:
ADA4400DAA6CD
CCBWE 0517MPMW

The retail heat sink is a copper heat pipe design with alu vanes.

Just have the rest of the day at work to go.. pfftt.. BF2 should turn up later too :D
06-24-2005, 01:46 AM
[timko]

Looks like the premium you paid by ordering from OCuk has paid off then since no other distributor in the UK has managed to get hold of any 4400s to ship yet, only the 4200s so far :)
06-24-2005, 02:07 AM
ewitte

Some Pics :) Tests ran 2.85Ghz 1.584v.

Sandra 64bit
Cinebench 32bit
Cinebench 64bit
Dual 1MB Pi
Temps with two instances of CPU Burn-in

http://s120570057.onlinehome.us/images/x2/1.jpg
http://s120570057.onlinehome.us/images/x2/2.jpg
http://s120570057.onlinehome.us/images/x2/3.jpg
http://s120570057.onlinehome.us/images/x2/4.jpg
http://s120570057.onlinehome.us/images/x2/5.jpg
http://s120570057.onlinehome.us/images/x2/6.jpg
06-24-2005, 03:14 AM
ewitte

Dual 32MB SuperPi. Running 2.806Ghz and lower vcore.

http://s120570057.onlinehome.us/images/x2/7.jpg
06-24-2005, 06:27 AM
xgman

What is the best way/program to burn in both cores at the same time on the X2? Run two instances of "cpu burn-in"? Does that = one for each core automatically?
06-24-2005, 06:35 AM
xgman

Quote:

Originally Posted by ewitte

Dual 32MB SuperPi. Running 2.806Ghz and lower vcore.

Are you getting higher stable clocks at closer to 1.5v or 1.6v?
06-24-2005, 08:15 AM
p0tempkin

Dual SPI 32MB isn't the most reliable CPU stability test, IMHO. Neither is Dual CPU-Burn. Both don't load your CPU or system as much as Dual Prime95 Small FFTs or S&M.

Dual Prime95 Small FFTs usually loads my CPU about 10C higher than Dual SPI or Dual CPU-Burn. S&M loads my CPU even higher than Prime95.

The best way to test full stability is to use all of them.

Memtest+ Test 5 & 8, 25 Passes
Dual Prime95 Small FFTs 24 Hours
OCCT Torture Test & Stability Test
Dual SPI 32M

I found that I needed a bump of +.02v to get my CPU absolutely stable at 2.7GHz. Still on air cooling. No doubt this CPU would do 2.8GHz stable if I was on water, and at least 3GHz on phase.
06-24-2005, 08:30 AM
ewitte

Quote:

Originally Posted by p0tempkin

Dual SPI 32MB isn't the most reliable CPU stability test, IMHO. Neither is Dual CPU-Burn. Both don't load your CPU or system as much as Dual Prime95 Small FFTs or S&M.

Dual Prime95 Small FFTs usually loads my CPU about 10C higher than Dual SPI or Dual CPU-Burn. S&M loads my CPU even higher than Prime95.

The best way to test full stability is to use all of them.

Memtest+ Test 5 & 8, 25 Passes
Dual Prime95 Small FFTs 24 Hours
OCCT Torture Test & Stability Test
Dual SPI 32M

I found that I needed a bump of +.02v to get my CPU absolutely stable at 2.7GHz. Still on air cooling. No doubt this CPU would do 2.8GHz stable if I was on water, and at least 3GHz on phase.

It wasn't supposed to be a stability test. I posted it for the people requesting screenies. If anything it was just to show how long it took.
06-24-2005, 08:51 AM
xgman

Quote:

Originally Posted by p0tempkin

Dual SPI 32MB isn't the most reliable CPU stability test, IMHO. Neither is Dual CPU-Burn. Both don't load your CPU or system as much as Dual Prime95 Small FFTs or S&M.

Dual Prime95 Small FFTs usually loads my CPU about 10C higher than Dual SPI or Dual CPU-Burn. S&M loads my CPU even higher than Prime95.

The best way to test full stability is to use all of them.

Memtest+ Test 5 & 8, 25 Passes
Dual Prime95 Small FFTs 24 Hours
OCCT Torture Test & Stability Test
Dual SPI 32M

Why not just pour gas on it and use a blow torch? :rolleyes:
06-24-2005, 11:08 AM
jinu117

Quote:

Originally Posted by xgman

What is the best way/program to burn in both cores at the same time on the X2? Run two instances of "cpu burn-in"? Does that = one for each core automatically?

I do prime and set process affinity to test. If you have both core running high and well... lucky you. My first core is sucky while 2nd core is exceptional.... figures....
06-24-2005, 11:25 AM
ewitte

Quote:

Originally Posted by xgman

Why not just pour gas on it and use a blow torch? :rolleyes:

If your going to kill it why not use LN2 and have a little fun :D
06-24-2005, 11:51 AM
xgman

Quote:

Originally Posted by jinu117

I do prime and set process affinity to test. If you have both core running high and well... lucky you. My first core is sucky while 2nd core is exceptional.... figures....

My first core is better than the second. I suppose this is to be expected, but you would think that they would be closer based on the same die, lot etc.
06-24-2005, 12:58 PM
ewitte

I can't pinpoint it exactly but I'm guessing my 2nd core is off by about 20-50Mhz. They will both do over 2.8 fairly easy :p:
06-24-2005, 01:02 PM
xgman

Quote:

Originally Posted by ewitte

I can't pinpoint it exactly but I'm guessing my 2nd core is off by about 20-50Mhz. They will both do over 2.8 fairly easy :p:

How are you testing this? Did you try Systool overclocking utility to see which core errors?
06-24-2005, 01:36 PM
ewitte

Quote:

Originally Posted by xgman

How are you testing this? Did you try Systool overclocking utility to see which core errors?

I got it semi stable with single core and then started throwing the second one into the mix. It just about handles the same workload but either needs a little vcore boost or a drop of 20-50Mhz.

Eric
06-24-2005, 01:58 PM
s7e9h3n

Quote:

Originally Posted by ewitte

I'm not running air. Its water....
Eric

hehehe, must have not caught that earlier...In that case, VOLT AWAY! :explode2:
06-24-2005, 02:48 PM
NickK

Well it's up and running after a slow painful complete reinstall. I haven't even attempted to OC it yet so the poor thing is running with everything at default settings (slack timings etc).
With an X800XL and 4400 at complete stock (not even a whiff of tweaking) I get 5029 3DMark'05 and 23446 3DMark'01.

My X2 install was on a completely brand new DFI nF4 SLI-DR with the 310 BIOS. I used the boot CD BIOS set to flash 510-2. I can see what the BIOS thinks the CPU is as the monitor is too slow!
Also had a little fun with my old OCZ 3500 as it wouldn't post. Using a 512Mb of CAS3 ram for my server it booted first time and allowed me to set up everything.
Put the OCZ back in and it all runs fine.

I'm not expecting a massive OC as it's on an XP90C.

I'll get some proper benching over the weekend. Too tired now..
06-24-2005, 03:09 PM
ewitte

Quote:

Originally Posted by s7e9h3n

hehehe, must have not caught that earlier...In that case, VOLT AWAY! :explode2:

1.7 BIOS is my max and only for benchmarking. This thing should be fairly easy to get stable running 2.8 @ 1.5 or so volts :)
06-24-2005, 03:43 PM
ewitte

After a few more tests I think I'm off on the 20-50Mhz estimate. I think 3.0Ghz would have been completely doable at 1.65v on the first CPU. The second one seems to need about 1.625v to get 2.8Ghz stable. Both should be able to handle 24 hour prime @ 1.5-1.575v WITH some burning in.

Eric
06-24-2005, 05:02 PM
ewitte

Seeing the grammar check passed for some odd reason you can guess why I did not publish it ;)

http://s120570057.onlinehome.us//images/x2/8.jpg
06-24-2005, 05:08 PM
s7e9h3n

Quote:

Originally Posted by ewitte

Seeing the grammar check passed for some odd reason you can guess why I did not publish it ;)

http://s120570057.onlinehome.us//images/x2/8.jpg

Why? :confused:
06-24-2005, 05:13 PM
ewitte

Quote:

Originally Posted by s7e9h3n

Why? :confused:

I'm running the free version. The only way to get a score is to run grammar by itself before running the suite. An option thats not normally available ;)
06-24-2005, 05:18 PM
s7e9h3n

Quote:

Originally Posted by ewitte

I'm running the free version. The only way to get a score is to run grammar by itself before running the suite. An option thats not normally available ;)

Yes I know how to get that score...If you read sampsa's post, I was the one who had asked him originally ;) But I'm not quite sure why you wouldn't post your score. I guess it could be one of two reasons:
A) You're running a hacked pcmark04 :woot: or
B) You were able to complete the PCMark Benchmark from start to finish without running grammar check separate and you think the orb won't accept it? :p:
My money's on A :D
06-24-2005, 05:19 PM
xgman

Quote:

Originally Posted by ewitte

1.7 BIOS is my max and only for benchmarking. This thing should be fairly easy to get stable running 2.8 @ 1.5 or so volts :)

I'm finding 1.55v in smartguardian will run stable at over 2900 mhz. at 34 deg idle and 42 deg under dual cpu-burn running. I may try to go up to 1.6 later on and see if I can get closer to 3ghz. maybe when my G5 block arrives that will help.
06-24-2005, 05:20 PM
xgman

Quote:

Originally Posted by ewitte

After a few more tests I think I'm off on the 20-50Mhz estimate. I think 3.0Ghz would have been completely doable at 1.65v on the first CPU. The second one seems to need about 1.625v to get 2.8Ghz stable. Both should be able to handle 24 hour prime @ 1.5-1.575v WITH some burning in.

Eric

yeah, I'm finding burn in seems to help matters.
06-24-2005, 05:42 PM
ewitte

Quote:

Originally Posted by s7e9h3n

Yes I know how to get that score...If you read sampsa's post, I was the one who had asked him originally ;) But I'm not quite sure why you wouldn't post your score. I guess it could be one of two reasons:
A) You're running a hacked pcmark04 :woot: or
B) You were able to complete the PCMark Benchmark from start to finish without running grammar check separate and you think the orb won't accept it? :p:
My money's on A :D

Its A.
06-24-2005, 09:43 PM
h4nc0

Just got my 4400+. Won't be able to install it until Monday tho. :slobber:
It's CCBWE 0522TPMW.
06-25-2005, 02:09 AM
Mr.President

hey guys which bios are ypu using for your x2?
06-25-2005, 02:11 AM
jinu117

Quote:

Originally Posted by Mr.President

hey guys which bios are ypu using for your x2?

Was using 510-2. Trying 623-2 and seems to be just as good.
06-25-2005, 05:41 AM
ewitte

There is one really nice thing about the X2. While I'm testing stability on CPU2 I don't have to wait for the system to complete. I have prime95 running on it while I'm playing KOTOR2 :D
06-25-2005, 11:24 AM
shimmishim

so it appears that the 4400 X2 is the most popular chip by far...

how come no one wants to test out the 4200+?

i know it only has a total of 1 meg L2 compared to the 2 meg L2 of the 4400+ but doesn't having less cache help in overclocking due to less heat produced?

i dunno... what do you guys think?
06-25-2005, 12:20 PM
Agent-JCDenton

Quote:

Originally Posted by shimmishim

how come no one wants to test out the 4200+?

i know it only has a total of 1 meg L2 compared to the 2 meg L2 of the 4400+ ....

A 4200+, same clock speed as the 4400+ but with half the cache and only approximately £35 cheaper. A 4200+, Would you?

And I haven't seen any 4200s clocking a whole lot higher than 4400s

As for performance, a picture says a thousand words.
06-25-2005, 12:30 PM
shimmishim

it shows that at least when it's 1024 vs 512 that the extra 512 cache makes like no difference at all but that's single core...

i assume it doesn't make a huge difference in dual core either...

seems like the price of the 4400+ keeps going up and up at most vendors and the price dfiference between the 4200+ and 4400+ is going up... from about $40 to over $70 at some vendors...

is the extra cache really worth it?
06-25-2005, 12:39 PM
Agent-JCDenton

Quote:

Originally Posted by shimmishim

it shows that at least when it's 1024 vs 512 that the extra 512 cache makes like no difference at all

Revisit the link
06-26-2005, 10:46 AM
jinu117

Got replacement X2 from outpost.com and was able to get 50mhz more (one thing is though, it seems like both core do like voltage vs previous chip. As for stepping... its exactly same stepping still CDBHE 0515RPAW) 2.65ghz @ 1.57v. I think it might be heat related problem on air for limitation I have as well as very hot PWM temp... (up to 64c I saw a t one point -_-) That is my prime 95 stable 3d stable, game stable, memtest stable setting. Memory is running 2t 4x 512mb mushkin blue 2-2-2-5 @ 240mhz 3.25v. Overall, system response time is great, gaming isn't bad at all vs 2.8ghz venice setup I had. Maybe I will get back into some cheap water solution with some left over parts I STILL had and shove it in to see how it fares.
PS) This time, core 1 is better than core 2.... figures. Core 1 actually can seem to go 2.75ghz with about 1.5v while core 2 likes to ramp up with more volt... I need core 1 from this cpu and core 2 from previous one.... that would make this 2.75ghz prime 95 stable machine on air... chuckle...
06-26-2005, 02:06 PM
NickK

I think the extremely hot PWM is a common trait for X2 overclocks due to the increased current draw compare to single core overclocks.
06-26-2005, 02:26 PM
joe2004

In other words it is extremely unlikely there will be overclocks comparable to FX-55. Unless Vapochill comes out with something that can handle more than 200W. Watercooling is the best way to go here, but certainly no low end stuff like my old crappy Exos.
06-26-2005, 05:50 PM
Agent-JCDenton

Hence i'm looking to install a few TEC mods to my Koolance setup
06-26-2005, 09:55 PM
Usurper

i dont think they produce anywhere near 200w tbh.
06-27-2005, 05:27 AM
ewitte

Quote:

Originally Posted by NickK

I think the extremely hot PWM is a common trait for X2 overclocks due to the increased current draw compare to single core overclocks.

My PWM temps are about the same as my 3700+ SD. Although that was running 2.8Ghz @ 1.7v. I've got a 80mm fan right above them but it still hits 53-54C or so during the day when its hot in the house. If I'm not home I'll turn the fan up to max and it stays around 50C. I don't want to hear that thing if I'm home ;)

Eric
06-27-2005, 05:31 AM
ewitte

Core 2 needs extreme voltage to get anything over 2.8Ghz stable while Core 1 probably would have scaled to about 3Ghz. For instance running 2.816Ghz 1.65 or 1.67v core 1 9 hours prime before I cancelled it, core 2 lasted 11 minutes. I'm running 2.805Ghz @ 1.625v and it looks a little better. Did not try core 1 but core2 made it 40 minutes before it had an error. I'm burning it in, checking again and then trying 1.6v.

Eric
06-27-2005, 05:42 AM
xgman

Quote:

Originally Posted by jinu117

Got replacement X2 from outpost.com and was able to get 50mhz more (one thing is though, it seems like both core do like voltage vs previous chip. As for stepping... its exactly same stepping still CDBHE 0515RPAW) 2.65ghz @ 1.57v. I think it might be heat related problem on air for limitation I have as well as very hot PWM temp... (up to 64c I saw a t one point -_-) That is my prime 95 stable 3d stable, game stable, memtest stable setting. Memory is running 2t 4x 512mb mushkin blue 2-2-2-5 @ 240mhz 3.25v. Overall, system response time is great, gaming isn't bad at all vs 2.8ghz venice setup I had. Maybe I will get back into some cheap water solution with some left over parts I STILL had and shove it in to see how it fares.
PS) This time, core 1 is better than core 2.... figures. Core 1 actually can seem to go 2.75ghz with about 1.5v while core 2 likes to ramp up with more volt... I need core 1 from this cpu and core 2 from previous one.... that would make this 2.75ghz prime 95 stable machine on air... chuckle...

I guess it is better to have core 1 be stronger than core 2 if they are going to be different since core 1 will get most of the work right?
06-27-2005, 05:48 AM
xgman

Yeah, I'm finding too much voltage even at reasonable temps starts adversely effecting stability. I wonder why most reposrts are that the two cores are so far off from each other in stability at the same overclock? Really limits the overclocking potential.
06-27-2005, 06:13 AM
Agent-JCDenton

Quote:

Originally Posted by ewitte

My PWM temps are about the same as my 3700+ SD. Although that was running 2.8Ghz @ 1.7v. I've got a 80mm fan right above them but it still hits 53-54C or so during the day when its hot in the house. If I'm not home I'll turn the fan up to max and it stays around 50C. I don't want to hear that thing if I'm home ;)

Eric

are you using TCCDs or UTT/BH5?
06-27-2005, 06:15 AM
ewitte

Quote:

Originally Posted by Agent-JCDenton

are you using TCCDs or UTT/BH5?

TCCD
06-27-2005, 06:53 AM
xgman

Quote:

Originally Posted by Mr.President

hey guys which bios are ypu using for your x2?

was on 623-3 but tried 510-3fixed and got better performance and bandwith for some reason. Not sure what the deal is with the Erata 123 etc. referred to in bios 623 where there is a diable switch that syas better performance but may cause problems. It seems that AMD has released a defective processor. Not sure how to tell how this exhibits itself.
06-27-2005, 09:03 AM
Agent-JCDenton

Quote:

Originally Posted by ewitte

TCCD

50C is quite high for the PWM using TCCDs, tho i managed 48C the other day without cooling and vDimm set to 3.6v. whats your ambient temp?
06-27-2005, 09:25 AM
ewitte

Quote:

Originally Posted by Agent-JCDenton

50C is quite high for the PWM using TCCDs, tho i managed 48C the other day without cooling and vDimm set to 3.6v. whats your ambient temp?

During the day the AC is set to 80F. That specific room can be 85-90F.

Eric
06-27-2005, 09:41 AM
jinu117

Quote:

Originally Posted by Agent-JCDenton

50C is quite high for the PWM using TCCDs, tho i managed 48C the other day without cooling and vDimm set to 3.6v. whats your ambient temp?

1) PWM temp is for the CPU PWM... not VDimm.
2) Setting vdimm high on DFI actually lowers the temp on mosfet for the top right corner as you are dissipating less thrown out power through that circuit vs say 3.4v....
3) Peltier on Koolance.... :stick:
06-27-2005, 09:47 AM
mongoled

Quote:

Originally Posted by ewitte

There is one really nice thing about the X2. While I'm testing stability on CPU2 I don't have to wait for the system to complete. I have prime95 running on it while I'm playing KOTOR2 :D

LOL, damn right, is so annoying tht while stressing a single CPU u cant do too much with the computer, dual chip means tht problem goes bye byes

:D
06-27-2005, 12:14 PM
Agent-JCDenton

Quote:

Originally Posted by jinu117

1) PWM temp is for the CPU PWM... not VDimm.
2) Setting vdimm high on DFI actually lowers the temp on mosfet for the top right corner as you are dissipating less thrown out power through that circuit vs say 3.4v....
3) Peltier on Koolance.... :stick:

When i was burning in my Redlines today i noticed that the PWM temp would increase with vDimm,

vCore -> 1.61v

vDimm:
3.3v -> 43C
3.4v -> varied between 43-44C
3.5v -> a steady 44C
=>3.6 -> 45C

I've gone back to 3.3v for the time being while i sort out cooling in that area (I need to tidy my cables)

And yes, thats TEC on a Koolance.
06-27-2005, 12:19 PM
ewitte

Quote:

Originally Posted by Agent-JCDenton

When i was burning in my Redlines today i noticed that the PWM temp would increase with vDimm,

vCore -> 1.61v

vDimm:
3.3v -> 43C
3.4v -> varied between 43-44C
3.5v -> a steady 44C
=>3.6 -> 45C

I've gone back to 3.3v for the time being while i sort out cooling in that area (I need to tidy my cables)

VDimm @ 2.7
06-27-2005, 12:21 PM
Agent-JCDenton

Quote:

Originally Posted by mongoled

LOL, damn right, is so annoying tht while stressing a single CPU u cant do too much with the computer, dual chip means tht problem goes bye byes

:D

Possibbly run 2 games at the same time if you've got a dual monitor setup? I'm thinking Championship Manager (or Football Manager 2005) & say Operation Flashpoint (Annoying for long waits while searching and connecting to online servers). C Manager doesn't use much of the GPU being mostly text based...i'll out when i try running them 2gether...whenever www.Komplett.com see it fit to ship my X2.
06-27-2005, 12:33 PM
Agent-JCDenton

Quote:

Originally Posted by ewitte

VDimm @ 2.7

Is your chipset >50C? What PSU are you using?
06-27-2005, 01:49 PM
Bandit1

Quote:

Originally Posted by p0tempkin

CCBWE 0517MPMW

same as mine,hehe
06-27-2005, 04:36 PM
ewitte

Quote:

Originally Posted by Agent-JCDenton

Is your chipset >50C? What PSU are you using?

Chipset only gets up to about 46C.
06-27-2005, 04:43 PM
ewitte

After about 24hours burnin on second CPU.

http://s120570057.onlinehome.us/images/x2/9.jpg
06-27-2005, 05:00 PM
johnrr6

Probably stupid questions

How do you engage each core separately so you know what is running what??

Run something like Prime to occupy first core----then run the bench??

When does the 2nd core kick in???

Is it a percentage of CPU use on the first core???

what happens when you multi taks with say 5 apps running at the same time, i.e.: Word, PowerPoint, Phtoshop, Dreamweaver and Outlook???

Is there rhyme and reason as to what core handles what???

Please excuse what may be stupid questions....

I have a 4800+ X2 arriving tomorrow for my DFI rig.

I love the DFI and my 4000+ diego overclocked to 2.8-----and in single apps or a single game it is astonishingly quick....

But multitasking.....

It is a sad 2nd place to a stupid 2.4 cheap Intel Compaq I have at work.

I multi task a lot-----and lag in multitasking drives me bananas....

Am hoping the x2 will solve this as I'd like to stay with AMD....

Here's a pic of my yellow and black themed DFI rig it will go into:

http://www.frontpage2000.family-net....is/Inside1.jpg

John
06-27-2005, 05:08 PM
joe2004

Quote:

Originally Posted by Usurper

i dont think they produce anywhere near 200w tbh.

Oh, I think they do, and go over that. I know that dual Opteron setups spend more than 100W on full load more than single Opteron setups. If this is only 50W increase now then you are looking at the 400-500 MHz overclocked processor that goes to 200W and anything higher more than that.

This will not be tamed by Vapochill, only by solid watercooling, or perhaps modded Vapochill.

Multicore processor overclocking is not for the faint of heart. AMD has done good job of keeping it under the roof at the default frequencies but do not expect that to be the case once overclocked.

Show 100 post(s) from this thread on one page

All times are GMT -8. The time now is 04:01 AM.

XtremeSystems