Weird FX-55 (2.8 @ 1.36V Yes) (2.9GHz @ Anyvolt NO!)

[Aleman]

Hi guys, long time no see.

I'm having hard time overclocking my FX-55.

Specs:
DFI LP Ultra-D (3/29)
2GB Gskill HZ (UCCC)
FX-55 Sandy

Current Stable OC:
255*11=2.8GHz @ 1.36V
24hrs PRIME 95 TORTURE STABLE

My problem is that I can't OC the CPU any higher using any volt.

Any freaking idea Why???

I already tried to lower HTT (using high multi's)
or lower HTT multi (because of the bus 1GHz "limit"
or increase all kind of volts (LDT, vCore, vChipset, u name it)

.. nooooo result at all!!!

Any idea?

ps. Temps are low 40's anyway, that shouldnt be the problem. Ive tried with stock 200MHz for RAM and still no success with the CPU.

[v0dka]

I assume youre only overclocking using an upward multiplier?

The problemmust be in your memory though, take a good look at your secondary mem timings.

[Serge84]

Thats because your board is limited. We all know FX55's can do 3.5ghz 4ghz easy. They use to be a common thing here specially with the FX57 before conroe took over the OC forums. Unless you have a crappy chip or your a beginner OCer. I have never seen a FX OC that bad before in a bad way. Maybe you need to lower your memory speed timmings and Hyper transport width. It needs a CPU vcore of 1.5v or 1.6v to go higher btw to a resonable level. And the chip can take it np. Max might be around 1.7v for resonable limits. You'd want better cooling like liquid or phase 24/7 for it to use more volts then that all the time or its life would be only about a year.

[KingDingeling]

serge84, DFI Ultra-D's are NOT limited at 2.8GHz or anything somewhere near that! I've seen people boot these guys at 3.3GHz for once, and your information that FX55's and FX57's hit 3-5-4GHz easily is just plain wrong! On air, you might be lucky to hit something around 3.5 because of temp limitations, but why do you think it took so long for people to hit 4GHz on AMD?

[p0tter]

umm..FX55 will not do 3.5-4 easy....


Lower your mult (x8) and see what the FSB max's out at.

OC Guide = DFI NF4 LP's (http://www.dfi-street.com/forum/showthread.php?t=20823) dfi-street would also be helpful, make sure you read the forum rules.

You are also running an older bios, ive had the best luck with the 704BTA

[Aleman]

Lower your mult (x8) and see what the FSB max's out at.

OC Guide = DFI NF4 LP's (http://www.dfi-street.com/forum/showthread.php?t=20823) dfi-street would also be helpful, make sure you read the forum rules.

You are also running an older bios, ive had the best luck with the 704BTA
Thx 4 the link,

im no OC begginner lol. used 2 OC while u still were in kindergarten (jk).

The 3/29 is the Newest NF4 ultra-D bios i could find, its way newer than 704BTA i suppose.

I'll give it a try with high FSB.

[RPGWiZaRD]

Who knows could be deadspots in an unusually low FSB. I've seen it happen for people before where higher FSB would work but just not that certain value. So checking for the board's max FSB value using a lower multi would be a great start.

Sux my SLI-DR board is limited at already 310/311 with 1.6v chipset and 1.3v ldt voltage, dun wanna use any higher chipset voltage as it only goes a little further and it's already quite hot due to the oven to 7900GTO card having those very hot components just a few centimetres in front of the chipset fan which is an Evercool VC-RE with 120mm fan blowing from beside at that area! The Opty 165 is Orthos stable (only let it ran for 6hrs which is enough for me at least) at 2.8GHz @ 1.41v tho so it would have a lot more potential... :(

[uOpt]

Is that one of those that newegg dumped out for $139 lately? What is the production code and week?

[Aleman]

Yeah its from the 250GB HDD Combo

ADAFX55DAA5BN

CABCE 0520 RPMW

[Serge84]

I been OCing for many years and gone through about 16 platforms myself. I'm pretty sure I'd know what I'm talking about with my experionce as a moder and selling computers, and what I'm capable of OCing to XXXX speed. Who said it was a walk in the park, for a beginner it would be a challange but not to somebody who has maxed out some chips before in there time. I could max out the chip way beyond 3ghz if I had the same parts to work with.

[Serge84]

serge84, DFI Ultra-D's are NOT limited at 2.8GHz or anything somewhere near that! I've seen people boot these guys at 3.3GHz for once, and your information that FX55's and FX57's hit 3-5-4GHz easily is just plain wrong! On air, you might be lucky to hit something around 3.5 because of temp limitations, but why do you think it took so long for people to hit 4GHz on AMD?

Did I ever say all DFI's are like that? Doesn't look that way. It could be his or has a bad production chip. I seen FX55 57 60 62 hit way beyond 3.5ghz even 4.2ghz speaking of a FX57. It only depends on the production limits of the chip. How well it was fabed, the quality of the silicon, the SOI manufacturing, DSL phase, finally how good the COMS Cu structure is in the whole transistor makeup that will ultimately result in its max performance.

So there are only 4 conclusions. Its a bad chip, the mobo is limited mainly the chipsets HTT for the CPU, powersupply may not be giving enough clean power to the cpu (Fluds it with dirty power and = crappy OC), or the OCer (Human error, miss set timings, miss configured settings.) I highly don't think the memory could be the factor here.

[GMX]

Did I ever say all DFI's are like that? Doesn't look that way. It could be his or has a bad production chip. I seen FX55 57 60 62 hit way beyond 3.5ghz even 4.2ghz speaking of a FX57. It only depends on the production limits of the chip. How well it was fabed, the quality of the silicon, the SOI manufacturing, DSL phase, finally how good the COMS Cu structure is in the whole transistor makeup that will ultimately result in its max performance.

So there are only 4 conclusions. Its a bad chip, the mobo is limited mainly the chipsets HTT for the CPU, powersupply may not be giving enough clean power to the cpu (Fluds it with dirty power and = crappy OC), or the OCer (Human error, miss set timings, miss configured settings.) I highly don't think the memory could be the factor here.

EDITed out by IFMU
Another one to add to the ignore list.
I'd like to Serge's hands on a sweet setup(well considered sweet a while ago) consisting of things along the lines of an FX55 and OC it past 3.5ghz assuming he gets positive 40C at 2.8ghz and uses the same cooling.

Aleman try 11/14 bios and looser MAL/RP

EDITed by IFMU
Wow.... straight up flaming and you have been here for just over 2 years and you dont know better?
tsk tsk tsk.
Flaming edited out.
IFMU

[DuceGT]

Serge, you've been voted off the island.. please leave... thanks

if your hitting any "brick walls" its usually due to the mainboard/memory system...when your starting to reach the peak of your CPU's OC you'll notice it'll go unstable and fail stress testing.. you gotta kinda creep up on the brick wall of the CPU... gl bud

[uOpt]

Yeah its from the 250GB HDD Combo

ADAFX55DAA5BN

CABCE 0520 RPMW

I have a similar one from the same sale, don't have the week code handy right now.

As your, mine goes very well on low volts but doesn't improve much with higher volts. Mine does go to 3.0 GHz 1000 minutes mprime stable at 1.50 Volts, though.

I did notice that mine doesn't drive my TCCD as high as my CABNE Opteron. You might want to lower the RAM devider regardless of whether you use a high CPU multiplier.

[Sparky]

Check your secondary RAM timings, I found I was stuck at 2.5GHz on my opty 165 until I fiddled with those and then I could get up to 2.81GHz where I am now. It didn't take much and I forget exactly what I changed but it was only a clock cycle or two and I was up and running.

Even with the RAM set to stock speeds sometimes it isn't the RAM but maybe the memory controller inside the CPU that makes you need to adjust those timings.

[Aleman]

Aleman try 11/14 bios and looser MAL/RP
Tried, no success whatsoever :( The CPU did not get any higher.



Check your secondary RAM timings, I found I was stuck at 2.5GHz on my opty 165 until I fiddled with those and then I could get up to 2.81GHz where I am now. It didn't take much and I forget exactly what I changed but it was only a clock cycle or two and I was up and running.

Even with the RAM set to stock speeds sometimes it isn't the RAM but maybe the memory controller inside the CPU that makes you need to adjust those timings.
I fidled with them, no success either :(

[Rickster_64]

I also had an FX from the same sale which even with 1.65V would not run stable @ 3ghz. That was using 15X 200 fsb, and even a 166 divider to rule it all out. My system is stable to about 280fsb 1:1. So what did I do? Put my real cpu back in, LOL.

[uOpt]

You need to try it in a different board, I'm afraid.

What multiplier and FSB are you trying, exactly?

[tvdang7]

alot of ppl that got in on the fx-55 deal dont get a good overclock such as your self. they are just not good clockers. check around. most ppl top out around 2.8- 2.9 with the newegg 139 dollar ones. (even tho urs was the free hardrive one its prob the same cpu's just with a free HD).
its a shame really since they used to be 1000 dollar cpus and a 70 dollar opteron 144 can overclock high than that batch.

[syne_24]

same here, my previous fx-55 would not hit 2.9 with max volt. Although mine was a clawhammer and I was running on air. But I suspect it might have been my board also but never really got to the bottom of it.

[.sentinel]

EDITed out Flaming ~ IFMU No wonder everyone ignores you. Another one to add to the ignore list.
I'd like to Serge's hands on a sweet setup(well considered sweet a while ago) consisting of things along the lines of an FX55 and OC it past 3.5ghz assuming he gets positive 40C at 2.8ghz and uses the same cooling.

Aleman try 11/14 bios and looser MAL/RP

Serge, you've been voted off the island.. please leave... thanks

if your hitting any "brick walls" its usually due to the mainboard/memory system...when your starting to reach the peak of your CPU's OC you'll notice it'll go unstable and fail stress testing.. you gotta kinda creep up on the brick wall of the CPU... gl bud


OOOOOO more people to add to my ignore list for EDITed out Flame by IFMU Even, I have seen a FX-57 at 4.2..... Ever hear of K|ngP|n....

*EDITed by IFMU*
.sentinel.... I've seen you around a bit and I know you know better too... dont make me send out an official warning.

[GMX]

OOOOOO more people to add to my ignore list for EDITed out by IFMU Even, I have seen a FX-57 at 4.2..... Ever hear of K|ngP|n....

EDITed by IFMU Now this guy (serge) is saying u can EASILY get 4.2ghz on an FX-55, where at 2.8ghz you get 40C temps(ie same cooling for this instance). Please say to me how it is possible
EDITed by IFMU

EDITed by IFMU
Flaming edited out.
IFMU

[learners permit]

A friend picked up one of those FX's from Newegg and it don't clock for crap.

[Lastviking]

What psu are you using?

[Aleman]

Im using an Enermax EG485AX-VHB(W)

3.3V 32A
5.0V 32A
12V1 18A
12V2 18A

its ATX 2.01 and all.

[KingDingeling]

Thats because your board is limited. We all know FX55's can do 3.5ghz 4ghz easy. They use to be a common thing here specially with the FX57 before conroe took over the OC forums. Unless you have a crappy chip or your a beginner OCer. I have never seen a FX OC that bad before in a bad way. Maybe you need to lower your memory speed timmings and Hyper transport width. It needs a CPU vcore of 1.5v or 1.6v to go higher btw to a resonable level. And the chip can take it np. Max might be around 1.7v for resonable limits. You'd want better cooling like liquid or phase 24/7 for it to use more volts then that all the time or its life would be only about a year.

yes, I think you said that Utra-D's are limited....

anyways, apart from that, 485Watts might be a little low for that kind of setup. The rails might be ok, but you hav 2 gigs of RAM, probably not a low end vid card, and you have a high end OVERCLOCKED CPU. Look into a higher end PSU and see what that comes out to

[uOpt]

You guys with the badly going FXes probably want to try them in a different board that is known to give better results. Then you know whether it actually is the CPU.

[Aleman]

i tried a different FX-55 and it goes 250x12 (3.0) with 1.41v no problem.

[Rickster_64]

Its not the enermax 485W. I saw one of the first day reviews for Quad FX that one of the guys hooked up a 550W power supply and it ran just fine with a single 8800 GTX. That was a dual FX74, 4GB ram, fans, and 8800GTX. I think that is a little more draw than a single core with 2gb and single vid card.

[osckhar]

Same problem with my FX-57. The Vcore does not help to reach 3GHz.

Oscar

[Serge84]

EDITed by IFMU Yeah you've seen a FX @ 4.2ghz on what cooling u little crap? No wonder everyone ignores you. Another one to add to the ignore list.
I'd like to Serge's hands on a sweet setup(well considered sweet a while ago) consisting of things along the lines of an FX55 and OC it past 3.5ghz assuming he gets positive 40C at 2.8ghz and uses the same cooling.

Aleman try 11/14 bios and looser MAL/RP

Phase mother fletcher! :D

http://www.xtremesystems.org/forums/showthread.php?t=100443

AMD

FX
FX-62: OPPAINTER - 3635.9 MHz http://www.xtremesystems.org/forums/...d.php?t=100443
FX-60: gloatlizard - 3875.44 MHz http://valid.x86-secret.com/show_oc.php?id=97927
FX-57: before - 4208.1 MHz http://img199.imageshack.us/img199/9243/cpu42089kx.jpg
FX-55 (ClawHammer): |RickY|&Pedro Rocha - 4005.23 MHz http://valid.x86-secret.com/show_oc.php?id=10680
FX-55 (San Diego): -De$troyeR- - 4056.86 MHz http://valid.x86-secret.com/show_oc?id=108727
FX-53: |RickY| & Pedro Rocha - 3602.8 MHz http://resources.vr-zone.com/ocdb/da...tel_or_amd=amd

Read the fourms more pall before the words come out of your mouth before somebody pops your fat head. Conroe isn't the only fraling chip that can do 4.2ghz YPOS!!! Have I nice bloody day mate before ISOWAOY. :D

[Serge84]

I bet he got confused with Intels core 2 due chips. My personal tip to serge is stop smoking, lol . J/K.

Does the above post look anything like conroe? ;) Joke taken as. :D lol

[Serge84]

EDITed by IFMU
Now this guy (serge) is saying u can EASILY get 4.2ghz on an FX-55, where at 2.8ghz you get 40C temps(ie same cooling for this instance).
EDITed by IFMU

From my own personal experionce I have hit 3ghz on air and thats a 1000mhz OC. Not a 200mhz or 400mhz OC with a 2.6ghz cpu or fansy FX. Its a 50% OC with a 2ghz cpu. It was as easy because i have OCed other peoples products before for them. Too bad it was my chips limit.

http://valid.x86-secret.com/show_oc?id=132953
http://valid.x86-secret.com/show_oc?id=132951

Since I'm capable of these things I could OC a FX57 to about 4ghz with the right chip and cooling. Its not a hard thing to do if you been OCing for as long as I have. One customer wanted me to OC there rig on phase and did so up to 3.5ghz with a FX-57 to run all the time on a phase set up about a year ago. So don't tell me I can't OC or get to a sertent speed because I done it before. :rolleyes: You have no room to talk hotshot because you don't even know me in the outback mate!

Why don't you put your foot where your mouth is and show us some 50%+ OCes your so great at doing then with a AMD? :rolleyes:

[GMX]

EDITed by IFMU
I have been a member longer than you here, i think i know very well that 4ghz has been reached on CASCADE you EDITed by IFMU." But you say "it cant be chips limit because another did 4ghz" Wrong. Simple as that, that is not the way to argue "because another chip did" If you been OCing for years you wouldn't say a dumb:banana::banana::banana::banana: ideology like that.
EDITed by IFMU

50% OC? I have reached 2.88ghz on a Sempron 2600(rev E) with stock cooling
ALso a sempron 2800(rev E) to 2.92ghz. SO dont talk to me about how good you are, because you are not. It's straight forward to OC 50%. Th e Chip is capable, then yes easy. If its not, then hard

For Eg, tell me about your Sempron D0 OC? Any over 50%? another 2600 sempron rev d0 @ 2.5ghz 1.8v

Or tell me about Chips with 1mb L2 or dual core that you can get to 3ghz+, those are alot harder to so obviously, oh but you too stupid to see that and oblivious to any fact. Your posts are filled with opinion(which are twisted in words)

There you go now, you better shut your mouth now.

EDITed by IFMU
Nah, you will be, if you have been here longer than he has then you SHOULD KNOW that the flaming youve done in this thread is WAY out of line, PERIOD.
You sir, are taking a 3 day.
Accept it.
Acknowledge it.
Move on.

EDITed by IFMU[/b]
How about this?

[el rolio]

im echoing it here but:

whenever i hit OC walls after near perfect mhz-voltage scaling in this past year of s939 clocking, i was always directed to loosen up MAL and RP as well as tweaking a few of hte other settings in the secondary timings. it worked everytime. looks like that might be what is happening with you

edit: i'll try to get the settings for you tonight specifically that i used to get my 4x512mb happy with my FX @ 3ghz (1.42v)

[IFMU]

I am going to say this ONE (1) Time. If I see ANYONE in this thread break any forum rule or guideline that has broken them in here anytime soon, I dont give rats a** if it is your first offense or not, expect to take some vacation time.
Period.

IFMU

[el rolio]

i tried a different FX-55 and it goes 250x12 (3.0) with 1.41v no problem.

basically the same exact thing as me.

ps: thanks for the cleanup IFMU. one bit of hyperbole sent this whole thread spiraling outta control, and all this guy was trying to figure out is the same thing OZZIMARK had to help mee with 3 times this year already.

[Rickster_64]

They just must be dumping inventory of ALL their inventory of anything except Opterons and AM2 chips.

[AceGoober]

I bought two FX-55 CPUs: one from BOLt11 on [H]ard|forum and one from Newegg. Both are CABCE 0516 WPMW.

After messing around with them for several days, the CPU from NewEgg was found to be the better OCer. I could hit 282 FSB without OCCT or SuperPi failures whereas the other wouldn't get past 266 FSB no matter what voltage or multiplier I used. This, of course, was all done with the RAM

I am at work right now and this is the only pics I have access to (sorry for the large images, the workstation I am at does not have Photoshop or GIMP on it):

http://img.photobucket.com/albums/v210/AceGoober/FX-55/DSCN0521.jpg

http://img.photobucket.com/albums/v210/AceGoober/FX-55/FX-55-3102-11x282-CPU-14-CP.jpg

I have had it running at 3.1Ghz for the past 3 days and it seems stable though I haven't completed the rigorous testing I put each and every piece of hardware I purchase through before calling it good.

And, pertaining to the OPs predicament, I would

[[XC] leviathan18]

Phase mother fletcher! :D

http://www.xtremesystems.org/forums/showthread.php?t=100443

AMD

FX
FX-62: OPPAINTER - 3635.9 MHz http://www.xtremesystems.org/forums/...d.php?t=100443
FX-60: gloatlizard - 3875.44 MHz http://valid.x86-secret.com/show_oc.php?id=97927
FX-57: before - 4208.1 MHz http://img199.imageshack.us/img199/9243/cpu42089kx.jpg
FX-55 (ClawHammer): |RickY|&Pedro Rocha - 4005.23 MHz http://valid.x86-secret.com/show_oc.php?id=10680
FX-55 (San Diego): -De$troyeR- - 4056.86 MHz http://valid.x86-secret.com/show_oc?id=108727
FX-53: |RickY| & Pedro Rocha - 3602.8 MHz http://resources.vr-zone.com/ocdb/da...tel_or_amd=amd

Read the fourms more pall before the words come out of your mouth before somebody pops your fat head. Conroe isn't the only fraling chip that can do 4.2ghz YPOS!!! Have I nice bloody day mate before ISOWAOY. :D


there are almost no amd cpu near stable at 4ghz a few counted with my right hand bench spi1m near that speed, you say its easy to OC a fx55 at 3.5ghz or 4.0ghz i can tell you its almost impossible to get a stable oc above 3.5ghz for everyday use specially with dual cores you need subzero cooling that doesnt hit your cold bug and besides that you need to get it low enough to oc it high....

you know where its easy to oc to 3.5 4.0 4.2 with conroe thats an easy cpu to oc if you dont have a FSB wall... get your facts straigt ricky and pedro broke 4ghz after a lot of hard work with their AMD setup dont come here and say thats an easy task