• 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.

AMD 3800 X2

Massive Attack

Massive Attack

Massive Attack

Permabanned
Joined
9 Aug 2006
Posts
2,273
hehe, i have just managed to get hold of a LDBHE 0601 UPMW 3800x2 N , i was wondering what core these processors are as i have heard that they are toledo cores with some of the cache disabled , but on the other hand i have heard they are manchester cores.

Also what sort of speed have people had these processors on air as i have a scythe ninja and didnt want to change m,y cooler as i have only just bought it.

Thanks for your help
 
As far as I know they're Manchester core. Regards speed, mines sitting happy at 2.4GHz on the stock cooler, at 1.4V, doesnt seem to like going higher, but that could be the cooler.
 
ive heard rummours that they go to 2.8 on air, i have got two coolers a big typhoon and a scythe ninja, i am unsure what cooler to use when i get my processor.

what model is your 3800, i have the LDBHE 0601 UPMW ( Waiting for delivery )
which is a N chip and is supposed to be the best 3800 at ocing
 
CPUZ will tell you if it's a manchester or toledo core. The toledo core chips clock better supposedly, they are revision E6 as opposed to E4 and I think the memory controller on the E6 is better generally. I have an E4 version and it's overlocking quite nicely stepping is LDBFE 0623XPMW.
 
when i get it mine will be LDBHE 0601 UPMW stepping with an N in the corner, so i take it the n stands for newcastle and not toledo. i was told the above stepping was better than toledo, i will have to wiat and see and post some pics up next week when i have her set up and running.

What cooler are you using and what are your temps for this processor,
 
Massive Attack said:
hehe, i have just managed to get hold of a LDBHE 0601 UPMW 3800x2 N , i was wondering what core these processors are as i have heard that they are toledo cores with some of the cache disabled , but on the other hand i have heard they are manchester cores.

Also what sort of speed have people had these processors on air as i have a scythe ninja and didnt want to change m,y cooler as i have only just bought it.

Thanks for your help
Click to expand...


Check my sig. My chip will do 2700 Mhz but it locks up in 3DMark2006 (passes Prime 95 and OCCT though). I'm sure it could go even higher.
 
My chips a toledo but it can run at 2750MHz 24/7 on less than stock volts, even 2800 is stable but my RAM doesn't allow me to run it at whilst folding is running.

The chip itself can do 3006MHz stable on 1.55v but my RAM, again forces me to run it on the 100MHz divider so I leave it at 2750MHz 24/7.

The code of it is: lcbie 0619RPMW

Regards,
ghgh
 
i have 2 gig of corsair value ram and i was wondering if i would be able to get 2.8ghz as my curent 3700 sits at 2.8 all day long, an d to be quite honest with you i would like to see some sort of performance increase for the money i have just paid.
 
The X2 3800 will have a multiplier one time lower than the 3700+...10 compared to 11 so when you set the ram divider the RAM is going to be running faster than it was before on the 3700+ CPU.

I reckon if your RAM is up for it, if not lower the divider to 133MHz....then that chip could goto 3GHz on roughly 1.55v or maybe 1.5v from what I've googled and found on another forum - looks like a very decent stepping.
 
Just to add to the above, my chip which doesnt like doing above 2.4Ghz on the retail coolers is a

ADA3800DAA5BV
LDBHE0601UPAW
2877279A60604

Which is very close you yours, so I'd wager the performance may be similar.
 
El Jimben said:
CPUZ will tell you if it's a manchester or toledo core. The toledo core chips clock better supposedly, they are revision E6 as opposed to E4 and I think the memory controller on the E6 is better generally. I have an E4 version and it's overlocking quite nicely stepping is LDBFE 0623XPMW.
Click to expand...

Hmm similar here, I have a LDBFE 0622WPMW with N in the corner and it's sitting @ 2.6Ghz 1.4V overnight dual prime stable, well last time I ran it anyway.

2.0Ghz - 1.325V
2.2Ghz - 1.35V
2.4Ghz - 1.35V
2.5Ghz - 1.375V
2.6Ghz - 1.4V
2.65/2.7Ghz - 1.475V not stable, lock up in 3dmark2005 canyon demo although 3dmark2006 passed!?

I would like to try 1.5V+ on mine but my mobo won't allow anything more than 1.475V (1.4V + Overvolt)...
 
Mine's similar to yours Steve, does 2.5GHz @ 1.35V, stepping LDBFE 0623XPMW. I think it could do 2.8GHz easy but I've not had a tweak in a while.
 
Sounds like I got a really bad clocker, or my PSU isnt providing stable enough power unless I set it higher where it always has minimum power+ required for the OC then :(
 
Well, you've got your RAM overclocked by 40MHz so I guess your running at 1:1. I doubt you'll be able to get it higher unless you slacken the timings and give it more volts - http://www.pureoverclock.com/article25-4.html.

I would use a 166 divider, you'll lose abit of the overclock on the RAM but you'll be able to get the CPU higher. You should be able to run at CAS 2 also, I've got the same RAM as you and I'm running at 2-3-3-6.
 
I have been, my ram wont boot at 6-3-3 @2 over about 220Mhz.
I've been running on the divider up til now, and managed to get it with 2.8V, @5-2-3 @2.5
I've always been trying to get the CPU stable before the RAM, as you can see from the following paragraph, my CPU bottles out, before the ram on the divider does.

As it was my b'day yesterday, I've been trying for a higher OC as I have more time on my hands. I've managed to get 2.6Ghz out of my chip (after about 4hours of fiddling with the voltages) (running the divider on ram 333Mhz [email protected], [On my board 400 is running 1:1 with fsb, so 333Mhz is what this board calls 166 divider]), but it requires 1.59V Vcore on my board, my RAM doesnt want to clock stably 1:1 at 8-4-4 even with higher V without locking up/reset looping :(, and my Nforce4 Ultra Chipset is at 1.65V (instead of 1.5) to stabilise the buffed fsb, with 3xHT and an extra 4% voltage to the HT to try and make sure everything runs stable. Unlike the RAM recieved by the above dude, my sticks wont go higher with timings that low, and my CPU seems to be a thirsty one too :(

My PSU is the 450W one that comes with the Antec SonataII if that makes any difference. Am currently trying the volts, and timings listed by your article to see if I can get memtest 1.65+ stable with those, then bump my CPU up to what I got it to if they can, although I have to say I love RAM timings and volts, it can be such an imprecise science, because high volts can make it clock worse than low volts sometimes!
 
Last edited:
Nope, my ram will not do 260Mhz, no matter the timings or voltage.
With my CPU already at 1.59V I dont feel like pushing that more than 2.6Ghz anyway, its reaching close to 90 degrees c like that when 100% stress tested, and I dont want it any hotter!
RAM is currently running with the 333Mhz Divider (approximately 218Mhz) at 2.8V 5-2-3 CAS 2.
Am memtesting now. Shame my CPU wont clock further, I know that RAM is stable at 240Mhz with CAS2.5 at those timings!

Obvious question though would be what gives better performance:
  • CPU @ 2600Mhz, RAM at ~220Mhz 5-2-3 CAS2
  • CPU @ 2500Mhz, RAM at 250Mhz 5-2-3 @ CAS 2.5 (1:1 with FSB)

Wanted to ask which is worth more, that extra 100Mhz CPU and tighter timings, or 100Mhz off CPU, slacker timings, but 30MHz extra on the RAM...or is there no decent answer, aside from it's machine specific and I'd have to test this myself (IE 3DMark03/06 Scores)?
 
Last edited:
Alexrose1uk said:
its reaching close to 90 degrees c like that when 100% stress tested, and I dont want it any hotter!



Obvious question though would be what gives better performance:
  • CPU @ 2600Mhz, RAM at 220Mhz 5-2-3 CAS2
  • CPU @ 2500Mhz, RAM at 250Mhz 5-2-3 @ CAS 2.5
.
Click to expand...


cpu speed over all else, without a doubt, a64s dont need the extra bandwith, they do however need low latencies.

And those temps are wrong, no way could it survive 90c
 
Possibly. It survived stably through a S&M 100% load CPU test, and shy of 90 degrees was the max temperature registered (was mid 80s). Will try and take a screeny if I can reproduce it, although its possible that the thermal sensor/sensor being read was incorrect. We are talking the stock cooler here though, not an uber Ninja cooler etc though, so I dont discount it :D
 
Last edited:
Alexrose1uk said:
Possibly. It survived stably through a S&M 100% load CPU test, and shy of 90 degrees was the max temperature registered (was mid 80s). Will try and take a screeny if I can reproduce it, although its possible that the thermal sensor/sensor being read was incorrect.
Click to expand...

Do you have a thermal probe? only way to be sure

What progs are you recording temps with?
 
You must log in or register to reply here.
Back
Top Bottom