Printable View
This mod permit to go more high with FSB with thoro B 133MHz FSB Def.Quote:
Originally posted by sysfailur
B4E.. what should we be looking at? What does that do?
EDIT: Nevermind.... *Smacks himself in the head*
Before this mod my max FSB was 230MHz ... after 250MHz!
CYA
Ok, I reconnected the L12 bridge as per the picture on my 1700+ JUIHB, and the cpu was indeed recognized as a 2600+ at 166fsb. But it made the overclockability worse, not better. It would not run above 222fsb stably regardless of the memory settings. This mod sounds bogus to me. This is on a Abit NF7-S v2.0 with the 1.1 bios. Disconnecting the mod let me run at 230fsb once again. Disconnecting the L12 bridge on a 2600+ should give you the same advantage, but reconnecting it on a 1700+ is a mistake.
rms
rsquires@flash.net
Have you tried to use older version of bios? IMHO 1.0 is better than 1.1.Quote:
Originally posted by rms
Ok, I reconnected the L12 bridge as per the picture on my 1700+ JUIHB, and the cpu was indeed recognized as a 2600+ at 166fsb. But it made the overclockability worse, not better. It would not run above 222fsb stably regardless of the memory settings. This mod sounds bogus to me. This is on a Abit NF7-S v2.0 with the 1.1 bios. Disconnecting the mod let me run at 230fsb once again. Disconnecting the L12 bridge on a 2600+ should give you the same advantage, but reconnecting it on a 1700+ is a mistake.
rms
rsquires@flash.net
CYA
EDIT: you can see discussions about L12 mod here
http://www.amdforums.com/search.php?...der=descending
No I didn't try it with 1.0 bios. There was another very odd thing about the reconnected L12 mod behavior: my cpu temp skyrocketed to 65C idle (watercooled) from 33C with the L12 bridge disconnected. This was not an artifact or a bad job mounting the cpu, as the water radiator was also quite warm. After disconnecting the bridge and reinstalling the temp went right back to 33C. So there's something strange going on, possibly bios-related.
But there needs to be more than just the one "reconnected 1700+ L12" positive result before I'd recommend people do this assuming it will work.
rms
Remember this; http://www.sudhian.com/showdocs.cfm?aid=307Quote:
Originally posted by rms
Ok, I reconnected the L12 bridge as per the picture on my 1700+ JUIHB, and the cpu was indeed recognized as a 2600+ at 166fsb. But it made the overclockability worse, not better. It would not run above 222fsb stably regardless of the memory settings...
I just tried another time
Still using pc2700 ram! rated 7.5ns
..:
http://www.oc-central.dk/pics/1.jpg
and
http://www.oc-central.dk/pics/2.jpg
Timings: 7-3-3-2.5
timings is a little on the slow side but hey its fair enough..
new bios for The a7n8x
a screen
It's not me but kazama (creator of setfsb)
http://www.page.sannet.ne.jp/k-hazam.../277-a7n8x.gif
simply crasy!! wow...Quote:
Originally posted by cpulloverclock
new bios for The a7n8x
a screen
It's not me but kazama (creator of setfsb)
http://www.page.sannet.ne.jp/k-hazam.../277-a7n8x.gif
btw. I hope they would make a bios like that for the rda+´s
can you give me a link??Quote:
Originally posted by cpulloverclock
new bios for The a7n8x
a screen
It's not me but kazama (creator of setfsb)
http://www.page.sannet.ne.jp/k-hazam.../277-a7n8x.gif
That`s some crazy bandwith with your pc2700 :slobber:Quote:
Originally posted by bias_hjorth
I just tried another time
Still using pc2700 ram! rated 7.5ns
..:
http://www.oc-central.dk/pics/1.jpg
and
http://www.oc-central.dk/pics/2.jpg
Timings: 7-3-3-2.5
timings is a little on the slow side but hey its fair enough..
I almost beat you on 228 FSB man i gotta get higher
http://www.mounewetter.nl/memory3.JPG
On 228 FSB with Twinmos pc 2700 no winbond chips 2.9 volt, overvolt on NF7-S to 3.04 volt.
I think the chipset is limiting me i cant go higher with loose timings or not but i`m trying, this is a single stick of 512 mb:D
no problemQuote:
Originally posted by ztop
can you give me a link??
http://www.page.sannet.ne.jp/k-hazama/ ---> BENCH RESULT
http://www.page.sannet.ne.jp/k-hazam...gogo/max55.htm
http://www.page.sannet.ne.jp/k-hazam...amd-wcpuid.htm
thanks:p
:slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber::slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber::slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber: :slobber:
Hehe... okay this score is shamed by you psychotic Prometia owners, but I finally got my 10,10.5 multis working so I'll post my work (in progress) :)
This is all on air. I'm hoping I can get FSB up a little higher... right now just ensuring stability at this level. Ran Prime95 all night with no errors, so I think it's doing fine so far.
EPoX 8rda+ A1
HyperX PC3500
http://www.fury-tech.com/images/user...a_10.5x220.gif
http://www.fury-tech.com/images/user...d_10.5x220.gif
caspergt:
thx man! btw mine is also a single stick
Here's my pathetic OCZ EL-PC3700 (10x221FSB), 2-3-3-6 and 2.6VDIMM. I'm having problemo's getting above 221FSB. I can acheive it, but I have to lower the mutli, to the point where it defeats the bandwidth. Foolish me, for some reason I thought the bandwidth would be an independent measurement from internal clock speed. Apparantly the overall speed up's the bandwidth as much as the FSB, and Latency?
If you run the test unbuffered the cpu speed has nothing to do with the benchmark of the ram.Quote:
Originally posted by Liquid3D
Here's my pathetic OCZ EL-PC3700 (10x221FSB), 2-3-3-6 and 2.6VDIMM. I'm having problemo's getting above 221FSB. I can acheive it, but I have to lower the mutli, to the point where it defeats the bandwidth. Foolish me, for some reason I thought the bandwidth would be an independent measurement from internal clock speed. Apparantly the overall speed up's the bandwidth as much as the FSB, and Latency?
Steve
Quote:
Originally posted by Liquid3D
Here's my pathetic OCZ EL-PC3700 (10x221FSB), 2-3-3-6 and 2.6VDIMM. I'm having problemo's getting above 221FSB. I can acheive it, but I have to lower the mutli, to the point where it defeats the bandwidth. Foolish me, for some reason I thought the bandwidth would be an independent measurement from internal clock speed. Apparantly the overall speed up's the bandwidth as much as the FSB, and Latency?
yeah I can hit 228FSB with an 8x multi, but I'm not running at 1.8GHz just for the added FSB!
I think part of it can be attributed to the stress factor. As you increase the clock speed of your CPU, it actually takes full advantage of the FSB and puts a higher strain on it. Could be wrong though, this is entirely speculation. I'm sure there are a number of factors, latency being one of them as well.
those 6gb/s scores of the 300fsb P4 rigs seem so much faster, but i just remembered that its pure bandwidth. they are not really much faster in gaming then an xp or barton with 250fsb.
with all this quadpumped fsb stories i just see the history repeating itself, intel is now taking the mghz legend to the fsb... and many people are falling for the 1.2ghz fsb. but paired with the low latency i must admit that its a different story then P4s mhz thing.
in Sandra I can disable "Buffering/Block-Prefetch Benchmarks" is this what your referring to? And if I did, then wouldn't my test be different from everyone elses?Quote:
Originally posted by Spec3
If you run the test unbuffered the cpu speed has nothing to do with the benchmark of the ram. Steve
Also is prudent to disable, "Use 50% Memory Only for Test"? Could this adversly affect the purpose of the twinbank memory architecture?
BTW Tekime I think your onto something there, or it may be as simple as the vcore needs to raised to compensate for the higher clock-speed. I think your correct as far as latency is concerned at higher clock-speeds, however; raising the latency would again defeat the benchmark.
My score seems pretty low, running 230MHz 4-2-2-2 @ NF7-S rev2
Dual Channel :confused:
http://www.biologigrand.ac/lgh11/temp/san-230-4222.jpg
Have I missed a setting in bios?
BTW, Hi everyone (first post).
Most recently (what's holding me back from 2.5, 2.6GHz is the dam Vcore 1.85V max)
Doh! Forgot the bandwidth.
Quote:
Originally posted by emp
My score seems pretty low, running 230MHz 4-2-2-2 @ NF7-S rev2
Dual Channel :confused:
http://www.biologigrand.ac/lgh11/temp/san-230-4222.jpg
Have I missed a setting in bios?
BTW, Hi everyone (first post).
to low multip?