NiBiTor v2.9 Feedback... - 2006/05/12 11:55Here we go, a new version is again available. Thanks to a a decent amount of BIOS's and xtremetheme for the new update with some big updates and several other fixes. As well as some extra support and new features... Some nice updates are available that have been requested! And with the memory timings update for the GeForce 7900 and GeForce 7600 series.
Updated support for Editing Memory Timings - GeForce 7900 Series - GeForce 7600 Series
And please provide your feedback... That way we will be able to support you even better! Also provide your Original BIOS as that help a lot to get thing better supported in the next versions!
| | The administrator has disabled public write access.
czuk
User Junior Boarder
Posts: 19
Karma: 2
Re:NiBiTor v2.9 Feedback... - 2006/05/18 08:00Is it possible to readout the CAS latency CL timing of the video memories? Well some background on this... I am asking because there seems to be some big issue with the Gainward BLISS 7800 GS+ 512MB AGP cards. As you probably may know, these cards use fast 1.2ns memories (1666MHz specified) however the actual reachable memory clocks are limited at about 1500MHz. I am in touch with a couple of guys from XS, VR-Zone and 3DCenter, however this issue is still unsolved. It is a fact, that memory voltages shouldn't be the problem (2.1V).
So at the moment we think, it might be the memory timings. NiBiTor says unfortunately that timings are okay. However we didn't get the CL timing. So at the moment our hope is, that CL timing is to fast respectively to short. Thanks for that great tool by the way. I believe the whole overclocking community is really happy with it!
| | The administrator has disabled public write access.
Mavke
Admin Admin
Posts: 13453
Karma: 340
Re:NiBiTor v2.9 Feedback... - 2006/05/18 15:23I would like to know a bit more and what you mean exactly, cause normally NiBiTor provides all memory timings parameters that can be changed... Now you are a bit lucky I think cause I receive a Gainward BLISS 7800 GS+ 512MB AGP card for review, which I will have for about a month. So with enough details I might be able to figure out and see how high I can overclock the card and relay that back to you.
| | The administrator has disabled public write access.
czuk
User Junior Boarder
Posts: 19
Karma: 2
Re:NiBiTor v2.9 Feedback... - 2006/05/18 15:53That would be really really really great!
More facts on the mem issue: As I said, all the enthusiast 7800GS+ users which are going for highest possible clocks are stucked with mem speeds around 1500Mhz, instead of much higher ones. The problem is, at the moment nobody knows why mem-clocks are so bad (keeping the used 1.2ns mems in mind) So the question is why.
Some possibilities were: - no 1.2ns RAMs but 1.4ns ones (false, as highres pictures of some users clearly shows 1.2ns markings on memories)
- undervolted memory (false, as they got their 2.1V)
- bad pcb layout which does not allow high mem clocks (not confirmed and hopefully will not be confirmed )
- too fast memory timings (partly false, as Nibitor readout confirms that the timings are between 1.2ns and 1.4ns default memory timings according to samsung datasheets. However as far as we see, nibitor doesn't show any CL timings. Thus we can not confirm that. So our idea is, perhaps the CL timings in BIOS is to fast for high mem clock speeds).
GPU speeds are okay. There is already a simple GPU voltage mod available, which gives 1.4 instead 1.2v. More GPU voltage is also possible.
| | The administrator has disabled public write access.
ChristTheGreat
User Platinum Boarder
Posts: 1066
Karma: 18
Re:NiBiTor v2.9 Feedback... - 2006/05/18 16:22Well actually, there is some problem with memory clcok of some Card. If I remember, that Gainward is a G71 right?
A few people have this problem, with evga, Asus, etc.. 7900GT as example. The card will be able to clock memory very high, then 2 days after, not even 1mhz more. It's like a protection or something..... Very strange. I'll try to find the problem, with the card of my friend. Also, it can be rubber on memory(ramsink) that may cause a few problem (already saw a problem like that).
| | The administrator has disabled public write access.
czuk
User Junior Boarder
Posts: 19
Karma: 2
Re:NiBiTor v2.9 Feedback... - 2006/05/18 17:32Chris, this Gainward card is a G71 on AGP, just to mention it.
It has yellow rubber-like stuff on the thermal pads. Thus, the pads are only pressed against the memories. However Gainward says, this stuff is supposed to be there. I took a look on it just a couple of minutes ago. It doesn't feel like normal plastic cover, it is really rubber-like and feels very soft. I guess it is kind of "mechanical fixing" for the pads, as the used pads are quite thick and really really soft.
| | The administrator has disabled public write access.
Capt_MonkeyBoy
User Fresh Boarder
Posts: 8
Karma: 0
Re:NiBiTor v2.9 Feedback... - 2006/08/11 21:20Hi,
about the RAM i have a 7800GS+, and noted the text from the ram fitted,.
and trcdwd is worked out from trcdrd and WL and tck,. and they both are meant to run CL11.
i tryed setting up the timings to the 1.2ns set but it still tops out at 1520mhzDDR'ish. i guess there could be other timings that are holding it back, as changing these timings only changes values in timing set 0 and 2, so is the ram the same unit just told to set up differently, like tccd running 200mhz 5222 or 300mhz at 8443., if so we need to unlock these puppys...
| | The administrator has disabled public write access.
EVGA 9600GT (SC) PCIe 512MB - Rev 01
