I know I saw a registry tweak somewhere that let you OC the memory on these cards with coolbits past the 505Mhz limit.. But I can't remeber where i saw it or how to do it.. I made a killer HSF for the card and got it to 290/505 with no trouble.. I want to see if I can get the memory past the 505 barrier.. Anyone know how?
-
Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!
Geforce 2Ti Memory over 505Mhz?
- Thread starter ptcg
- Start date
- Joined
- Dec 14, 2001
- Location
- Littleton, Co
I have the same card as you. (need to update below, took my GF4 4200 back) I've maxed mine out to 505 as well you might want to try rivatuner or something like that. I'm pretty sure that coolbits will only go to 505 on the memory. I got mine up to about 520 with riva though. here is a link http://guru3d.com/rivatuner/
- Joined
- Apr 28, 2002
- Location
- Chicago suburbs (Bartlett)
i saw that post 2. I think u go into the coolbits derectory and find a document thgat says the highestv speed it overclocks to. Then change that number 2 a higher 1 and coolbits should let u overclock higher.
- Joined
- Dec 14, 2001
- Location
- Littleton, Co
i guess i need to look for that....sound like it might be nice.. because i am sure i can go higher as well as ptcg.
- Joined
- Dec 14, 2001
- Location
- Littleton, Co
i am using the stock cooling on my card as well. mine is at 270/505 and i'm sure that there is more in the card. i'd use 3dmark or a favorite game and just play it for a few hours to test the temps. when they get to high it will begin to artifact(you'll know when you see them). this is from the ram gettting to hot. just back it down until you notice them no more. I have noticed that artifacts start to appear in medal of Honor after about 3 to 4 hours of staight game play with as high a setting as was playable. that game pushes the card pretty hard. (i had a GF4 for about a week and the game didnt even phase it with everything maxed out plus 4x antialising and vert-sync, god i looked good, going to get one) it is just luck on how high you can get the card to go.. start with small increments until you notice artifacting. if it locks up or crashes the program the GPU is too high
- Joined
- Dec 28, 2001
Re: Hmm
I've a Chaintech GF-2 Ti. I'm running it at 294/444. I lapped the stock heatsink, put an old socket 7 fan on top of the stock heatsink fan, and I made a bracket out of coathanger wire, and put a 60 MM case fan directly above the top of the GPU (opposite side of the heatsink). I also had to bump the I/O voltage to 3.5 to get the core stable at 294. I'm planning on getting some heatsinks for the memory, I can't go any higher as is. Just trying to get all I can out of it until I can afford to upgrade.Bundifrog said:
![]-[itman](/forums/data/avatars/m/9/9309.jpg?1650325391){kind=avatar}
Nimps said:
Hey, could you run on the tests on 3dmark? I personally(and I know a lot of others)won't take a score seriously unless all tests were performed. You don't have to of course, but if you did, you could also join the overclockers.com 3dMark team (link in the sig)!
- Joined
- Nov 8, 2001
GF2 ti's
I can run my leadtek card at 527. More if I'm I'm just benching for the sake of numbers..
Do join up with the 3dmark team, as I'm in the top spot for GF2's and could really use a good thrashing to knock my ego back into spec!
HAHAHAHAHAHAHAHHA!!!! I dare ya to beat me!
Check the link in my sig for our site and the sign up form.
We have some rules that you'll find on the 1st page of the 3dmark team sticky at the top of the Video/soundcards forum. Basically just labeling information and the standards for our 3 score catergories- default, stable, and suicide.
I can run my leadtek card at 527. More if I'm I'm just benching for the sake of numbers..
Do join up with the 3dmark team, as I'm in the top spot for GF2's and could really use a good thrashing to knock my ego back into spec!
HAHAHAHAHAHAHAHHA!!!! I dare ya to beat me!
Check the link in my sig for our site and the sign up form.
We have some rules that you'll find on the 1st page of the 3dmark team sticky at the top of the Video/soundcards forum. Basically just labeling information and the standards for our 3 score catergories- default, stable, and suicide.
If I ever pop my GeForce2 back into my system I'll run the full tests. But here's something you guys may want to read. Basically only the game tests count, there's no point to run the rest of the tests. As you can see in the following image, it clearly states that in the Read Me file that accompanies 3dMark. Although, you're still entitled to your opinion, this is just mine.
Yes, I understand that and I can see why you would only run the game tests, but for me and many others that I mentioned, it's a matter of stability. If you can oc a card to the moon, what's the point if you can't play a single game with it? I'm not saying that the oc was worthless or it shouldn't count, I'm just saying I don't consider it a succesful overclock unless it's stable. Yours may be stable, I don't know, but that's the thing. Unless all the tests are run, I don't know if it could finish them or not. On the 3dMark site, we sort users' scores by their stable run. Then, we throw in the suicide or unstable run to show we can compete, but we don't really count the suicide as competition. For us, there's just no point in it.
- Joined
- Nov 8, 2001
3dmark
The O/C 3dmark team believes in stability over pure benchmarking numbers. 3dmark is intensive, but for our default and stable scores, we like the extra tests for comparison purposes. The theoretical tests are useful for something, you know. And just to fully certify stability, we require that anyone submitting deafult or stable category scores run all the tests, which is sometimes not possible to do when a card or cpu is clocked too high. We have a separate category for competing for pure numbers in which stability is not a factor... any score submitted with only the game tests is automatically consider to be in this, the suicide category.
Just some standards the team adopted when it was formed... we like our cake and eat it too.
The O/C 3dmark team believes in stability over pure benchmarking numbers. 3dmark is intensive, but for our default and stable scores, we like the extra tests for comparison purposes. The theoretical tests are useful for something, you know. And just to fully certify stability, we require that anyone submitting deafult or stable category scores run all the tests, which is sometimes not possible to do when a card or cpu is clocked too high. We have a separate category for competing for pure numbers in which stability is not a factor... any score submitted with only the game tests is automatically consider to be in this, the suicide category.
Just some standards the team adopted when it was formed... we like our cake and eat it too.
Well, i've found that if my card gets passed Lobby in High Detail, it's good to go. It will always lock before or on Lobby High Detail.
I run all test most of the time, but if it makes it into Nature, I know it's home free. That's just how mine works.
I run all test most of the time, but if it makes it into Nature, I know it's home free. That's just how mine works.
I understand the whole stability thing, but if I ran through the game tests one by one and got my score, and then downclocked my video card and waltzed through the theoreticals my score would stay exactly the same, this defeats the whole stability testing you're going for and there's no real way to tell if someone downclocked or not, unless it was a signifcant underclocking. You can say your system was stable through all the tests, but that would the only proof of the is your word and since you're all pretty much strangers, that doesn't mean much to me. I just run suicide runs, because all that matters to me is my score, I know my system isn't stable at those speeds, that's why I don't keep it at that speed 24/7. Ah well, it just comes down to your view on things. 
