Ad

News

US science teachers pass on climate DVDs

Psst! Want 50,000 free DVDs? The US National Science Teachers Association didn't, even though the DVDs in question were of An Inconvenient Truth, the climate change movie that is required viewing for all children in Norway and Sweden. Read more

Intel claims edge over AMD on dual core chips

Chip giant Intel is moving heaven and earth to deliver dual core technology, but according to a web cast its president Paul Otellini made to staff, there are limitations on AMD technology which could give the chip giant the edge. Read more

Relevance Of Blu-ray Up In The Air, Says ISuppli

Following the recent revelation by Toshiba that it will no longer make any HD DVD products, the death knell of its next-generation, high-definition DVD standard has been sounded - and rival Blu-ray's victory assured. Read more

ATI confirms R580 for January or February - sort of

R580, ATI's new high-end graphics processor based on the X1000 architecture has not been quite what one would call a secret for quite some time. Read more

Latest Reviews & Articles

GeForce GTX 295 Performance: Previewed

GeForce GTX 295 Performance: Previewed

After a frustrating second half of 2008, Nvidia is looking to start the new year off by reclaiming its single-card performance crown. We got our hands on an engineering sample GeForce GTX 295 to give you a taste of what you can expect in two weeks. Read more

Phenom Recycled: Athlon X2 7000-Series

Phenom Recycled: Athlon X2 7000-Series

Just a couple of weeks before the introduction of its 45 nm Phenom II, AMD introduces a new dual-core chip. The Athlon X2 7000-series is basically a 65 nm Phenom with two active cores, but with the full L2 and L3 cache memory. Read more

Does Saving Power Mean Hurting Performance?

Does Saving Power Mean Hurting Performance?

Modern processors are capable of switching into power-efficient modes to save power when they’re idle, and an increasing number of motherboards offer dynamic features for the same purpose. Yet, the benefits come at a price. Read more

Four Full Tower Cases From $150 To $600

Four Full Tower Cases From $150 To $600

With Intel's Core i7 920 looking like a solid overclocking play, we wanted to find the right full-tower case for our test benches. ABS, Antec, Cooler Master, and Thermaltake battle it out for chassis supremacy. Read more

All the Reviews & Articles
  Tom's Hardware Forums » Overclocking » AMD » FX-55 Question
 

FX-55 Question




Word :   Username :  
 
Bottom
Author
 Thread : FX-55 Question
 
Profile: member
More Information

I have an FX-55 (S939), running with some decent memory on an ASUS A8N-E. However, I can't go faster than about 2.75GHz unless raising my voltage way up, which disappoints me when compared to the successes of other people with the same chip. I have to run 1.425v to even get it to do 2.75GHz when others do 3.0GHz with that much vcore... applying as much as 1.475v might get me up to 2.8GHz, but it's still slightly unstable. I have reapplied paste, tried BIOS updates, etc. I have a fairly large Zalman (7700 Al/Cu) on it, and I'm hovering around 50c full load.

Thus, I've decided to stay at my safe 2.75GHz unless anyone has ideas...
When overclocking, I have either of the two possible choices:

229 x 12 = 2748MHz (2-5-3-2, 1T - HTT 4x @ 916)
275 x 10 = 2750MHz (3-8-4-4, 2T - HTT 3x @ 825)

I can't decide which is better and can't figure out a decent way to bench in 'nix (don't have a legitimate copy of M$ at the moment, nor do I intend to pirate my way out of it).

Anyone think they know the answer to which is faster?

Related Product

Register or log in to remove.

Profile: Eternal Poster
More Information

is amd cool&quite enabled?

the fx-57 is 2.8 stock.

Profile: member
More Information

I meant FX-55 :oops:
C&Q is disabled.

Profile: Eternal Poster
More Information

Quote :

I meant FX-55 :oops:
C&Q is disabled.




well i was tottaly cofused on that.

thats not too bad for a 55 i guess, since it is clocked so high already.

but i have had my 3700sandy core up to 2730mgz before.
from stock of 2.2 :P

anyway there are more limiting factors such as
ram, mobo, cooling, and power.

Profile: member
More Information

Other have claimed to get 3.0GHz on stock... I can't come close no matter how high my voltage is :(. My guess is that I just got unlucky and got a bad chip...

System Specs:
500W PSU, Dual 12v @ 20A/ea.
ASUS A8N-E, nForce4 Ultra Based
AMD FX-55 (San Diego) @ 2.75 (275 * 10)
2x1GB DDR550 @ 3-8-4-4, 2T

Audigy2 Value Edition
7800GTX, 256MB @ 500/1300
80GBx2WD in RAID-0
DVD/CD-R, 16x


Have tried playing around with direction of airflow, to no avail:

120mm, 80mm intakes
90mm exhaust (high RPM)

Profile: Eternal Poster
More Information

maybe not all cpu,s will oc the same.

Profile: Forum Fixture
More Information

I'd say the one with tighter timings is better, amd's always perform better with tighter timings, and I think you could keep the ht mult at 5 and stay fine, maybe try the 13 mult, tighter timings again will do you more good

Profile: addict
More Information

stick with this option... 229 x 12 = 2748MHz (2-5-3-2, 1T - HTT 4x @ 916). lower memory timings at 1t and a higher htt.

it is true, not all chips overclock the same. that being said, have you tried dropping your memory divider down to 183, or even 166?

Profile: member
More Information

Yes, but then my memory frequency dips under DDR400, because I only have the option of 400 or 333 on my board, which takes away so much of my bandwidth. At my current setup, I'm almost hitting 3gb/s in memtest 86+.

Something interesting... when I set the memory interface to DDR466 @ 2-5-3-2-1t, the storage controller seems to give out, even though I locked my PCI bus at 33.33MHz...

13x (~3.00GHz) multiplier failed, tried upto 1.45v and no luck. I figure it's not such a good idea to go past that point, especially when I can do 1.40v at my current setting, 2.75GHz.

Profile: addict
More Information

all i can really say right now is that if you're hellbent on getting to 3ghz on this chip, upgrade your motherboard. kinda retarded to do considering s939 is basically dead.

Profile: Forum Fixture
More Information

Quote :

all i can really say right now is that if you're hellbent on getting to 3ghz on this chip, upgrade your motherboard. kinda retarded to do considering s939 is basically dead.



In that case, I recommend the same mobo you have, the a8n32-sli deluxe, pcworld did a review on it and it turned out to be the best ocer out of the 939 boards of 2005 and most of 2006 (don't know if anything is better that's out yet)

Profile: old hand
More Information

The higher the fsb 1:1 with the cpu, the slightly better games and heavily memory intensive apps are going to run (@ the same overall mhz)...

That said, you may be running in to a limitation with your mobo... The a8n-x series has known issues with running anywhere from 230-270fsb 1:1. I personally have run in to this with the a8n-del, and not run in to it with the a8n-premium (but i ran in to voltage issues with the premium...).

I haven't used the -e, so i couldn't say for sure... It could be just as you say...The chip... You could drop the divider to 400:333 to take the memory controller out of the equation (a little bit...), or 400:200 or whatever, and see how far you can push it (instead of running 1:1).

My 3500+ edited, and 3800+ x2 were both able to hit 2.75 between 1.45v-1.52v on different boards. I haven't even played with my 55 yet and i've had it for months now...

Profile: Forum Fixture
More Information

no I disagree, I will say the 1:1 thing is correct, but since he has a fx-55, the lower mhz with 1:1 with better timings will by far outperform anything but a 1:1 ratio, even the higher mhz with the 10 mult. All that pushing it will do is risk the chance of frying his memory

Profile: addict
More Information

Quote :

all i can really say right now is that if you're hellbent on getting to 3ghz on this chip, upgrade your motherboard. kinda retarded to do considering s939 is basically dead.



In that case, I recommend the same mobo you have, the a8n32-sli deluxe, pcworld did a review on it and it turned out to be the best ocer out of the 939 boards of 2005 and most of 2006 (don't know if anything is better that's out yet)

eh...i wouldnt recommend this motherboard. overall, it's allright, but in retrospect, i should have gone with a dfi.

Profile: old hand
More Information

You will only fry your memory if you overvolt it...

If it's within it's normal working voltages and can go to higher fsb (which he's already said it can work at *faster* [unspecified] fsb's...)... Then the faster the fsb is (with a given same overall mhz/ghz of the cpu)...

The arguement of memory latency vs bandwidth i think can be answered fairly easily by a quick google... ddr vs ddr2 or ddr vs ddr3...or sdram vs ddr or sdram vs ddr3... So far in sdram/ddr tech, increased bandwidth/fsb is more important then lower latencies, or we'd still be using sdram or some derivitive... I'm telling you, higher fsb = faster, at the same overall cpu rating/mhz. Anandtech did some easy to read testing about a year ago on this...

I've done some fairly extensive testing with this as well...a link to some of my testing...

edit: one more thing...uh the 32-del...it also isn't an extreme oc'er board...if you're interested in extreme oc'ing dfi is the answer (the only answer). That said...i'd only recommend dfi if you're interested in putting a lot of time in to it...like weeks to get a stable* overclock...

Profile: Forum Fixture
More Information

that's true for intel, but look at it this way. When you have an ht of already 1000mhz and it is a theoretical 2ghz, then having your memory run at speeds of say 500ddr with loose 3-5-5-10 timings (were talking cheap ram here), then the 500ddr will still be slower than the ht link, meaning that have much the latency is will actually effect the gb/s. So if you have your ram running at something like 2-2-2-5 1t with only 400ddr instead, then you will get far better performance, because the speed is already a bottleneck, but the latency is far lower

Ali

Profile: member
More Information