Test of the Real Nvidia Gtx Titan, single-sli-tri Sli, 4 WAY SLI
-
I don't exactly have Titan's, but I do have Kepler, a 680, I've never overclocked them, I have one at 82.3% and the other at 78.8%.
So what do you think, could they reach good clocks?
-
I don't exactly have Titan's, but I do have Kepler, a 680, I've never overclocked them, I have one at 82.3% and the other at 78.8%.
So what do you think, could they reach good clocks?
On GK104 I don't know which ASICs are the best, I'm also not sure if the rule applies as in Titan, the best way to see it would be to look at what boost frequency each card reaches and compare with others that are the same.
If the one at 82% reaches a better frequency with boots as it comes from the factory it's very easy that it can be overclocked a little better, although there's not much difference in ASICs, you can monitor the frequency and see which one reaches more in games, that way you'll see if the one with the best ASIC has a little more boost.
regards
-
Do you reach the 1200 of Stock?, well if so... first congratulations and second ¡¡ vaya tela!!, since I am reading and reading and I think I have not vsito any that does that.
I suppose, or I want to suppose, that these 1200 are enough for you but giving it the +38mv (that is to 1.2 or so, with stock BIOS).
Is that so?.
Greetings.
-
In GK104 I don't know which ASICs are the best, I'm also not sure if the rule is met as in Titan, the best way to see it would be to look at what boost frequency each card reaches and compare with others that are the same.
If the one with 82% reaches a better frequency with boots as it comes from the factory, it's very easy that it could overclock a little better, although there isn't much difference in ASICs, you can monitor the frequency and see which one reaches more in games, that way you'll see if the one with the best ASIC has a little more boost.
regards
Well if the one with 82.3% really has about 26 MHz more (1176) than the one with 78.8% (1150).
Well they are the TwinFrozr versions with factory OC, that is 1058 clocks, but as I say, I haven't really needed to OC them manually. I only asked because of pure curiosity.
But seeing my case, then I suppose that the theory of ASICs can be real, and also for Kepler.
We'll have to try overclocking them manually, see what they can do.
Well, thanks a greeting!!
-
Stradivarius, I haven't touched the voltage, I haven't raised 38mV or anything, because I don't like or want to touch voltages. My Asics are 80.1 and 80.5. The 80.5 one runs at 1215 Mhz long with 1.18V, although I haven't tested it alone yet, so I suppose that SLI will slow it down, but this particular card is a real powerhouse according to my colleague. With the first drivers they were at 1.16V. with RL and with the 320 they went up to 1.20 and 1.18 respectively without me touching anything, just reinstalling the OS and doing a clean installation of everything, and with the latest ones they are still there. I haven't tried to raise 38mV, maybe I will someday, but it's impossible because no matter how much I raise +38mV, they feed with 1.23V or 1.21 respectively. I use the PC very little, I'm not at home right now, but when I spend some time at home I will do more tests. Some of the units tested also did that, I don't know why some do and others don't and I don't think it's just a matter of the ASIC, not at all. It's impossible to access all the core data because Nvidia protects a part of them by fusion and another by encryption.
Jotole, mine don't have more ASICs than one of yours, so this demonstrates that it's not all ASIC in these cards, it has its influence but not all of it -
Ok, thanks "majo" …jeje (in both senses).
But there are things I don't understand about your cards, for example that it goes to 1.18v; since all the default maximum voltage they give is 1.16V; another thing is that they hold up better to OC with this voltage and/or do throttling. So I don't understand that... have you changed the BIOS or something?
I already read that with the 320 they went up to 1.20 and 1.18 but you must have touched some parameter because this is not how it is. I suppose the other users will be able to corroborate this...
Many thanks, and let's see if we can keep putting tests and those things.
Regards.
-
Stradivarius, I have tested more cards than I have and in some the voltages were variable, so I don't think this is strange. From what I read in the forums, between some Titans and others, they do very, very variable things with the boost. At idle, they are all almost the same, but in terms of boost, they are very variable. I don't remember in which forum, I don't know if overclock or which one, where there are many more users than here with Titans, I have seen cases like mine. The only thing I have is the EVGA precision and I used it to show here the graphs and visualize the speeds and voltages of the boost and I assure you that the voltage is not touched and the BIOS less. What I have seen very, very little are graphs that with only 1.18V give 1.215Mhz
-
Of course 1215@1,18V is amazing... but what I really haven't read or seen is the variable voltage "upwards", downwards obviously since it relaxes frequencies and so on... but anyway, I thought the maximum stock voltage for all of them was 1.16v.
Good luck then, but I'm very interested in your tests.
What assembler and ASIC do you have?
Thanks, regards.
-
Stradivarius the voltage I have not touched, I have not raised 38mV or anything since I do not like or want to touch voltages. My Asics are 80,1 and 80,5. The one of 80,5 is set to 1215 Mhz long with 1,18V, although I have not tested it alone so I suppose that SLI will slow it down, but this particular card according to my colleague is a real cannon. With the first drivers they were at 1,16V. with RL and with the 320 they went up to 1,20 and 1,18 respectively without me touching anything, just reinstalling OS and doing a clean installation of everything and with the latest they are still there. I have not tried to raise 38mV, perhaps someday I will do it but it is impossible because no matter how much I raise +38mV they feed with 1,23V or 1,21 respectively. I use the PC very little, now I am not at home, but when I spend a little time at home I will do more tests. Some of the units tested also did that, I do not know why some yes and others no and I do not think it is a matter of the ASIC alone, not at all. It is impossible to access all the core data since Nvidia protects a part of them by fusion and another by encryption.
Jotole mine do not have more ASIC than one of yours so here it is demonstrated that not everything is ASIC in these cards, if it has its influence but not allWell we have already managed to find out something more. I have the SC, I have not managed to test it alone. Let's see if I finish the new mess I have got myself into, and I will do some tests to see where she can go alone...
A greeting... -
Stradivarius the voltage I have not touched, I have not raised 38mV or anything since I do not like or want to touch voltages. My Asics are 80,1 and 80,5. The one of 80,5 is set to 1215 Mhz long with 1,18V, although I have not tested it alone so I suppose that SLI will slow it down, but this particular card according to my colleague is a real cannon. With the first drivers they were at 1,16V. with RL and with the 320 they went up to 1,20 and 1,18 respectively without me touching anything, only reinstalling the OS and doing a clean installation of everything and with the latest they are still there. I have not tried to raise 38mV, perhaps someday I will do it but it is impossible because no matter how much I raise +38mV they feed with 1,23V or 1,21 respectively. I use the PC very little, now I am not at home, but when I spend a little time at home I will do more tests. Some of the units tested also did that, I do not know why some do and others do not and I do not think it is a matter of the ASIC alone, not at all. It is impossible to access all the data of the core since Nvidia protects a part of it by fusion and another by encryption.
Jotole my ones do not have more ASIC than one of yours so here it is demonstrated that not everything is ASIC in these cards., it has its influence but not allWhat beasts you have got majo
Yesterday I put the water blocks, and for the little I could test I have gained something in terms of gpu and memories.
The best part is that before you had to give it wood to the fan so that it did not exceed 70- 80 degrees under load, and now with the blocks it does not exceed 32…..On the other hand I do not know why, when I raise the voltage +38 one gets to 1.2mv and the other to 1.16mv. I have to test the bios to see if I can squeeze everything out of them.
Testing them independently, the one that reached me to 1.2 I managed to submit it to 1200mhz which was impossible for me before, I remind you that mine have a 69% ASIC.
I will continue to do more tests............
Regards.
-
What a great beast you've got there, man
Yesterday I put the water blocks on it, and from the little I could test I've gained something in terms of GPU and memory.
The best part is that before you had to give it full throttle to the fan to keep it from going over 70-80 degrees under load, and now with the blocks it doesn't go over 32…..On the other hand, I don't know why, when I increase the voltage by +38 one goes to 1.2mv and the other to 1.16mv. I need to test the BIOS to see if I can squeeze everything out of them.
Testing them independently, the one that used to go up to 1.2 I was able to push it to 1200mhz which was impossible for me before, remember that mine have a 69% ASIC.
I'll keep doing more tests............
Regards.
Great temperature change from 80 to 32, equally this card has surprised me with its idle temperature with the same tower, my cards had these temperatures
GTX 285 - idle 32º
GTX 580 - idle 28º
GTX Titan - idle 17ºRegards.-
-
Great temperature change from 80 to 32 also to me this graph has me surprised its temperature at idle with the same tower my graphics had these temperatures
GTX 285 - idle 32º
GTX 580 - idle 28º
GTX Titan - idle 17ºRegards.-
Patagonico, those values are with R.L.??? With the Titan at idle you have it lower than me, taking into account that right now in Almeria it makes a good ambient temperature, around 20 Degrees….
I can't wait to get home tonight and start tinkering XD.
Regards.
-
Great temperature change from 80 to 32 also to me this graph has surprised me its temperature at idle with the same tower my graphics had these temperatures
GTX 285 - idle 32º
GTX 580 - idle 28º
GTX Titan - idle 17ºRegards.-
you live in Patagonia don't you?
It's that I'm not surprised by the cold that must be there, let's say that there you appreciate 4 way of 480, because at least they give heating.regards
-
What a great beasts you've got, man
Yesterday I put the water blocks on it, and from the little I could test I've gained something in terms of gpu and memories.
The best part is that before you had to give it hell to the fan so that it didn't go over 70-80 degrees under load, and now with the blocks it doesn't go over 32…..On the other hand, I don't know why, when I increase the voltage by +38 one goes to 1.2mv and the other to 1.16mv. I have to test the bios to see if I can squeeze everything out of them.
Testing them independently, the one that reached 1.2 I was able to submit it to 1200mhz which was impossible for me before, remember that mine have an asic of 69%.
I will continue to do more tests............
Regards.
Ruruloko, also mine changed a lot when I put RL on them, well really RL and drivers
Str@divarius mine have Asics 80,1 and 80,5 and both are eVGA SC
-
Good ASICs, at least they look the same… which is cool.
Do you have stock BIOS? How long do they stay stable without throttling?... and, I suppose the voltage doesn't go over 1.162v (if you have them stock, just adjusting TDP and Offset...). For example, majo's voltage goes up "on its own"... I don't know what this depends on....
Greetings!!!.
-
Patagonico, I'm linking you to a message I posted a while ago:
HardLimit - View post - P4 2.40 Ghz to 3.36, 280fsb, should I be satisfied?be careful with those 17º that it's showing you.
-
Patagonico, te enlazo a un mensaje que puse hace ya un tiempo:
HardLimit - Ver post - P4 2,40 Ghz a 3,36, 280fsb, deberia estar satisfecho?ten cuidado con esos 17º que te está marcando.
Son reales temperatura ambiente 8.9º C, son mediciones que se pueden hacer 1 o 2 veces al año
Saludos.-
tu vives en la Patagonia no?
Es que no me extraña el frio que debe hacer allí, vamos que allí se agradece 4 way de 480, por que al menos dan calefacción.saludos
Vivo en Buenos Aires pero amo la Patagonia y la visito casi todos los años, el tema es que ingreso una ola fría polar y como no sufro el frio me doy el lujo de hacer algunas pruebas.
17ºC minimos pero 73º en carga.-
Saludos.-
-
It's real room temperature 8.9º C, these are measurements that can be taken 1 or 2 times a year
Best regards.
I live in Buenos Aires but I love Patagonia and I visit it almost every year, the thing is that a polar cold wave comes in and as I don't suffer from the cold I can afford to do some tests.
17ºC minimums but 73º in load.
Best regards.
8 or 9º is nothing, some areas in Spain are much colder, here in winter it goes below 0º, we've already caught some colds trying to ocean with the windows open.
Patagonia is amazing, I hope they keep it, civilization ends up spoiling all those natural places.
Best regards
-
8 or 9º is nothing, some areas in Spain do much more, here in winter it goes below 0º, we have already caught some colds trying to sleep with the windows open.
Patagonia is amazing, I hope they keep it, civilization ends up spoiling all those natural places.
Regards
He he that's the temperature of my room, if here in Buenos Aires we also reach 0 and in Patagonia they are usually at -10 or -15ºC.
If it is wonderful I meet quite a few people from Spain who are fascinated by it.
Regards.
-
Now the data fits well with me. That you are entering winter.
Offtopic: I have a friend who came from Ushuaia to Lanzarote. He went from being below zero to twenty-five above zero. And now he is going to Australia, to forty above zero. He is going to work as a sign painter-graphic arts. Here at 10º he was already freezing, if he turns completely around he will die of hypothermia.
Greetings
