Mat_Agnesi
-
Posts
201 -
Joined
-
Last visited
Content Type
Profiles
Forums
Events
Blogs
Posts posted by Mat_Agnesi
-
-
Very Good score !
-
Grande ! non ho mai visto una battaglia cosi accesa come con questi Xeon L3014. 610MHz di FSB, non hanno limiti queste mainboard ?
-
First test; waiting for a competitive cpu & motherboard
-
-
For vCore, I've used only a little overvolt over default, not need for more volt.
For vNB, if I remember well I've used near 1,27v. Elpida mems ?? I've tried 3 different RAMs, (Samsung, elpida) and the better for overclocking proven to be the Elpida !! My Elpida 1GB DDR2 Bank can go up to an incredible 500Mhz 6-7-7 timings. But performance seems better at 420MHz with tight timings. Unfortunately I have only one Elpida RAM, so I cannot use Dual Channel.
-
The PCB has a total of 4 phase, 3 for GPU and 1 for RAM. This came from a 8800GTS, I think 8800GTX has more powerful power section.
however, this one that I have has supplied also my AMD HD 6850 (little overclocked) without problem ! not bad for that old mosfets
-
Thanks everyone for the interest. My custom power supply cutted from the 8800 GTS is modded and has a wide range voltage regulation. GPU voltage can be set from 1,1v to 1,8v. RAM voltage can be regulated from 1,5v to 2,3v.
For that crappy Qimonda RAM I've used 1,9 volt, and for the Core 1,45v
In fact, nachtfalke & poparamiro is right, not so much difference in overclock between the original Asus's power supply and my 8800 GTS's power supply. RAM stop at the same frequency, The core can go only a little higher than before and need a lot of voltages to raise more.
what voltage do you usually give to that G92 chip ??
For sure, this is not a Golden Card...it's crappy, right ?
but I have liked that experiments, it's funny.
-
Yes, the windows I've used is uncleaned, has a lot of software installed. Clean Windows and Dual channel seems to make a good difference, but better than that, the bigger bifference is made if you find a PM45 chipset mainboard. At same clock, It seems to score a lot more if compared with our PM965 board. I have a GM45 board now, don't know if it's good as the PM45 or not, but seems to have far better efficiency if compared to pm965. However I can't mod this board now, because it's not mine
-
I see better efficiency then mine here, very strange my rig score 200 less points with the same hardware
-
So no way to pass 315mhz fsb on that board..
-
FRANKESTEIN CARD, this is an 8800GT with the 8800GTS's Power supply.
Made by me from two broken CARD, the 8800GT has broken VRM, and the 8800 GTS has broken CORE.
Both Core and RAM is powered by the external 8800GTS's power supply.
Very Crappy Qimonda RAM can't Overclock more than 1060MHz even downvolted...
-
Good work. So even the BSEL Mod does not work.
We have to discover if this wall is simply caused by the Overclocking software (Witch cannot set more on that platform) or is an hardware limitation. If you attempt to force the FSB frequency via Hardware mod on logic inputs clockgenerator, we will discover this !
-
Woow, you have taked the 1st place, congratulation bro !
-
In fact I did Socket BSEL mod too, FSB 800 to 1066.
Vccp mod is the key to get high FSB, Vcca mod and Vmem mod seems not so helpful, my Samsung Memory don't like high Voltage.
FSB limited more probably by CPU, I think. I have other CM 925, one can do FSB 383MHz Max on air and 405Mhz by LN2; one only do 336MHz on air. This nice chip can do FSB 392 Max on air
I confirm, The north bridge voltage is the most important thing even in my rig.
So you are limited by the FSB Wall of the CPU ??
-
I'm Just thinking: Who win between a Celeron 925 and Xeon L3014 at the same Frequency ??
Seems that Celeron is better.
A lot of mod here... more than my rig. I haven't performed vmem mod.
I think the limit is still the Chipset.. or not ?
I see 1.550 Volt Voltage, what is this? the CPU voltage ?
-
This is an interesting experiment to do. also because my DDR2 notebook RAM overclock very well, can raise near 500MHz with CAS 5-7-7 or 6-7-6
-
Yes, that's should be tryed. If the system start up directly at 266FSB, it may not lock at 315.
However: with the BSEL mod the CPU force the system to start up at 1066 FSB, but the PM965 only support 800FSB, what does happen ? It may be an interesting thing to try.
-
Yes, I've used Set FSB to change the FSB from Windows! I think this is the only way to overclock a notebook without any option in the bios.
With other CPU, I've got a lower FSB limit. For example with an older 65nm Core 2 DUO model the fsb won't go over 265.
-
Unfortunately I can't find any detailed Datasheet for the RTM875T-606 If anyone can find it, please post here. thanks
-
Hi! yes, the Clock GEN is RTM875T-606. The 3rd pictures is a detail of the VID Vmod for the CPU (The scratched point, is where I have soldered the wire). The 4th pictures, is the Chipset Vmod.
I haven't checked the vDroop.. But I think it doesn't have so much vdroop. That mainboard can use a T7700 CORE duo CPU, witch require so much power if compared to a single core Celeron 925 45nm, even if overclocked.
-
The FSB slider can go up to 380MHz, but the FSB will lock up at 315mhz. This is a strange problem ! may be possible that the clock generator has reached it's limit ??
-
Nope, I change two different version of the software, but never touching this parameter.
I simply raised the clock. I can set all the frequency from 200MHz to 315MHz FSB, but, if I set for example 320MHz, the software seems to accept the command, but the real FSB frequency still remain fixed to 315MHz. Even If I set 330MHz, the FSB freq. still remain 315MHz
Do you think that parameter can unlock the situation ?
-
I have overvolted the PM965, and at this frequency this rig has passed the Wprime 1024, so is it very stable. I think it can go higher, but unfortunately the software " SetFSB " can't set more than 315MHz FSB.
P.S. I notice also better efficiency for PM45/GM45 chipset, more point for same MHz
-
Better tweaked on Threads. Thanks to Gigioracing
Mat_Agnesi - Core i7 4790K @ 4800MHz - 1223 marks XTU
in Result Discussions
Posted
Thanks Bro. This time It seems that I got a chip witch is not bad.