Battle of the titans "780 ti Vs 290x" first results.
-
Do you remember when the TITAN was about to be released, the fake leaks from chiphell making a supposed TITAN with more than 6500 points in the graphics score of 3D Mark 11 in extreme?
Well, look what happens, that in the end, it's going to be true and all..;)
1421MHZ…from there, I don't go..I don't feel like frying my graphics card..but the increase is absolutely proportional...40% frequency, a result of more than 40% over stock (you have to add the memories)...it's already giving much more than 690 or SLIs of GTX 680 oceados at maximum..incredible..
Best regards.
How do you master the unlock voltage now, ELP3, hehe :p:p:p
The Titan are great cards, but once they unlock the Ti, it's going to be a beast, because they do the same clocks or even better than the Titan when these couldn't be unlocked...
Pass me a firestrike to the sli with maximum oc, let's see how much you surpass me...
-
How do you master the unlock voltage now, ELP3, hehe :p:p:p
The Titans are great cards, but once they unlock the Ti, it's going to be a beast, because they do the same clocks or even better than the Titans when these couldn't be unlocked...
Pass me a firestrike to the sli with maximum oc, let's see how much you surpass me...
Right now I'm with the 4 stuck in the Ivy and it doesn't allow me to oc freely the two, and firestrike in general and perfomace in particular, is totally bandwidth.La the difference in core is minimal, and if very important what was said above. It would be better to see it in extreme..
But anyway, here it is:
They touched the ceiling of spending more than 150 in the power limit, even with that, so I imagine the frequencies have dropped.. in the first test it's clearly seen, what I don't know very well is to what extent..
However, this result in Extreme, I think is more appropriate:
And man.. I was disconnected from this unlock thing, but one has never been bad at oc-ing..:ugly:
Best regards and thanks.
-
¡Esta publicación está eliminada! -
Right now I'm with the 4 punctures on the Ivy and it doesn't allow me to oc freely the two, and firestrike in general and perfomace in particular, it's totally a bandwidth.La difference of core is minimal, and if very important what was said above. It would be yours to see it in extreme..
But well, here it is:
They have touched anyway and all ceiling of spending of more than 150 in the power limit, so I imagine it will have lowered frequencies.. in the first test it is clearly seen, what I don't know very well is to what..
However, this result in Extreme, if I see it more appropriate:
And man.. I was disconnected from this of the unlock, but one has never been bad at oc..:ugly:
Best regards and thanks.
Very good result, pushing them to the limit… The 780ti do it without added voltage, which means they are surely throttling. That's what I tell you, as soon as these beasts can go a little further, it will be seen that they are some brown beasts...
Regarding overclocking, I was joking with you, man. Sometimes, due to life circumstances you can't be into everything, it's normal. And for that, precisely, there are colleagues/companions of the moment who can help one.
Regards and thanks to you for your collaboration...
-
Lo importante es que tienes una buena grafica, fresca oceadora y silenciosa, además ahora ya no es tan cara, lo de la Ti pues si algunas han salido rana lo que me extraña es que no les revienten a los de overclocked. net o a ELP3 con la caña que les están dando, si fueran como la 590 ya tendría que haber muchos casos en todo el mundo.
La red va muy deprisa para decir el problema de la Ti pero muy lenta para decir las pantallas negras y bloqueos de las R9-290x, si es que fueron rapidos hasta para criticar el vsync adaptativo cuando lo acababan de meter al driver, pero de decir que no funciona el vsync normal desde el catalyst nada.
Vamos que se debería informar rápido para todo sea el que sea el fabricante, yo por eso termino no creyendo lo que se dice hasta que vea un caso de gente que conozco, de todas formas a ver si con la mala publicidad bajan esas Ti de precio.
saludos
No te entusiasmes tanto que ya quedan pocas Titan, por que al final la Titan va a seguir siendo la mejor tarjeta de Nvidia.
Aunque algo raro ocurre con las Ti para que no se note tanto el rendimiento con OC, pues deberían rendir un poco mejor a igual clock, parece que algo han capado por que es raro.
saludos
También es raro que un sitio con un nivel técnico y de seriedad como es CHW, siempre neutral y conocedor al detalle de cada problema, que nunca deja pasar una sea cual sea la marca, sea la que se haga eco de estos casos y no otros como los que mencionas.
[mode ironic OFF]
A mí lo que me hace coña es ver cómo se explica que es un problema que ha tenido UN FABRICANTE al cambiar algunos componentes del diseño de referencia por su CUENTA Y RIESGO, y se pasa a decir que todas las GTX 780 Ti tienen "problemas" en las fases, siendo como es un diseño netamente, no superior, sino lo siguiente a las 290/X con sus más que flojas fases de regulación (6 señores, 6 en total).
Después de meses de negacionismo seguido de las unidades fallidas de 7870, sí, las de sapphire, insistiendo sites como ése mencionado y foros varios (sus foreros, más bien) que ese problema no era real, que ya estaba parcheado, y con datos objetivos como éstos:
http://www.hardware.fr/articles/911-5/cartes-graphiques.html
Donde podemos leer partes tan interesantes como:
- 12,67% Sapphire Radeon HD 7850
- 7,44% Sapphire Radeon HD 7870 OC V2
- 7,41% Sapphire Radeon HD 7870 OC V1
- 7,02% Sapphire HD 7950 With Boost (11196-16)
- 6,09% ASUS HD7750-DCSL-1GD5
- 5,82% Sapphire Radeon HD 7870 V1
- 5,65% Sapphire Radeon HD 7870 V2
- 5,30% Gainward GeForce GTX 670
…
Si on s'intéresse aux chiffres par GPU nous obtenons :
-
Radeon HD 7850 : 3,74%
-
Radeon HD 7870 : 5,48%
-
Radeon HD 7870 XT : 4,25%
-
Radeon HD 7950 : 5,75%
-
Radeon HD 7970 : 5,31%
-
GeForce GTX 660 : 1,01%
-
GeForce GTX 660 Ti : 2,81%
-
GeForce GTX 670 : 2,87%
-
GeForce GTX 680 : 1,99%
Donde podemos ver que los problemas puntuales y ya arreglados no son tales (versiones V2), que además estos problemas masivos se presentan también en las 7850, y en la serie 7900 tampoco está libre de ratios de fallos muy elevados (un 5,75% de RMAs en un año de control de componentes es MUCHO).
En fín, en todos sitios cuecen habas, pero me gustaría ver menos escandalera montada por un fallo puntual de un fabricante y que ha afectado a unas pocas unidades, y que sí se critique más los fallos masivos vistos en modelos como los aquí listados, que todo el mundo que las tiene dice que son una maravilla y que no presentan problemas, pero los hilos de pantallazos y las estadísticas de RMAs dicen lo contrario.
¿Alguien recuerda cuando desde otro sitio tan serio como CHW se empezó a decir (y por supuesto CWH "rebotó" la noticia como site serio que es) que las GTX 680/670 estaban con un fallo de diseño masivo y que se iban a llamar a todas a fábrica por esto?
Eso fue a principios del verano del 2012, y corrieron ríos de tinta de críticas contra nvidia basadas en… un site desconocido con fuentes "fiabilísimas" (según ellos), pero secretas. Ya vemos cómo va el cotarro.
-
Very good result, pushing them to the limit… The 780ti does it without added voltage, which means they are surely throttling. That's what I'm saying, as soon as these beasts can go a little further, it will be seen that they are some brown beasts...
Regarding overclocking, I was joking with you, man. Sometimes, due to life circumstances, one cannot be everywhere, it's normal. And for that, precisely, there are colleagues/companions of the moment who can help one.
Regards and thanks to you for your collaboration...
The thing is that the 780 Ti do not currently allow either voltage unlocking or maximum TDP, so they suffer throttling and cannot go beyond stock voltages plus the minimum plus that nvidia allows (which is equally impressive, running at 1300 stock).
So it's clear that the potential is there, now we will have to wait for modified BIOS as has happened before, and with waits of many months, so I think that with the Ti this wait will not be so long (something has been learned along the way unlocking limits in the BIOS of their sisters), but it will not appear overnight (unless something has appeared these days and I have not noticed, I am a bit disconnected from the topic, jajaja).
-
The issue is that the 780 Ti does not currently allow voltage or maximum TDP unlocking, so they suffer from throttling and cannot go beyond stock voltages plus the minimum plus that NVIDIA allows (which is equally impressive, running at 1300 stock).
So it's clear that the potential is there, now we'll have to wait for modified BIOS as has happened before, and with waits of many months, so I think that with the Ti this wait won't be so long (something has been learned along the way unlocking limits in the BIOS of their sisters), but it won't appear overnight (unless something has appeared these days and I haven't noticed, I'm a bit disconnected from the topic, jajaja).
There are already modified BIOS.
In fact, both ConanR, Vmanuel and even I myself are using them. They nullify the boost, give more TDP and allow voltage-
All the captures you've seen in these last posts of the 780Ti are from modified BIOS that allow you to raise voltage up to 1,21V, give more TDP, nullify boost and not enter throttling.
What is being waited for is the voltage unlock of the MSI as I have in the TITAN to give more voltage and at least reach 1,3V which should make them go up without problems up to 1400MHZ or more.. but I do see a problem and also, if you notice for the first time in a long time there is no professional overclocker who has beaten any record with them officially. And it is perhaps, due to their very high cost... the reference phases despite having been modified, maybe are not prepared to endure what can come their way.
I certainly if I had Ti and the unlock came out, I wouldn't think of overclocking them with extra voltage without a good RL...
I don't see it as normal, as I comment, that no overclocker or brand, has not already sponsored some record-breaking event and especially when the graphics card has just come out and it's interesting to attract attention. Something must be going on that is more complicated than we thought at first..
Regards.
-
It is also strange that a site with a technical level and seriousness like CHW, always neutral and knowledgeable about every problem in detail, which never lets any pass by regardless of the brand, is the one that covers these cases and not others like the ones you mention.
[mode ironic OFF]
What amuses me is how they explain that it is a problem that one manufacturer had when changing some components of the reference design at their OWN RISK, and they move on to say that all GTX 780 Ti have "problems" in the phases, when it is a design that is clearly not superior, but the next to the 290/X with their more than weak regulation phases (6 gentlemen, 6 in total).
After months of denialism followed by the failed units of the 7870, yes, the ones from Sapphire, sites like the one mentioned and various forums (their forum members, rather) insisted that this problem was not real, that it was already patched, and with objective data like these:
http://www.hardware.fr/articles/911-5/cartes-graphiques.html
Where we can read interesting parts like:
Where we can see that the punctual and already fixed problems are not such (V2 versions), that these massive problems also appear in the 7850, and the 7900 series is not free of very high failure rates (a 5.75% of RMAs in a year of component control is A LOT).
In short, everyone is cooking beans, but I would like to see less scandal made about a punctual failure of a manufacturer that has affected a few units, and that the massive failures seen in models like the ones listed here are criticized more, which everyone who has them says are wonderful and do not present problems, but the threads of crashes and RMA statistics say otherwise.
Does anyone remember when from another site as serious as CHW they started saying (and of course CWH "reposted" the news as a serious site as it is) that the GTX 680/670 had a massive design flaw and that they were going to call them all back to the factory for this?
That was in the beginning of the summer of 2012, and rivers of ink of criticism against NVIDIA were written based on... an unknown site with "very reliable" sources (according to them), but secret. We can see how the situation is going.
I remember when some red screens of death appeared, the situation always gets blown out of proportion and I think it's good that they report any problem, but they should report everything, the driver that burns the graphics is like a snowball, curiously I have several cards with all the drivers including beta and they have not killed a single one, in the end there comes a point where I don't even pay attention, I learn much more by reading forums and users especially if they are known, than by reading news.
It's like the press or television, they only count what their news agency tells them, they all say the same thing, and that agency counts only what interests them, that's why in the end they make you not believe anything.
regards
-
If I remember when some red screens of death, things always get blown out of proportion and I think it's good that they report any problems, but they report everything, the graphics driver is like a snowball, curiously I have a few cards and I put all the drivers including beta and they haven't killed me a single time, in the end there comes a point where I don't even pay attention, I find out much more by reading forums and users especially if they are known, than by reading news.
It's like the press or television, they only count what their news agency tells them, they all count the same thing, and that agency counts only what interests them, that's why in the end they make you not believe anything.
saludos
That week I was "arguing" with a writer from CHW about the reason for the non-existence of any material about the BS and other problems presented in Hawaii. His explanation is that he waits for an official note from AMD, I deduce that he waits for AMD's permission or instructions ON which to publish about the topic.
A friend installed the latest CAT BETA and went through the BS trying to play BF4… He tried it after each forced reset of the machine... until the VGA wouldn't start up anymore.. His GA - HD 7970 GHZ is dead waiting for transport for RMA...
-
There are already modified bios.
In fact, both ConanR, Vmanuel and even I myself are using them. They cancel the boost, give more TPD and support voltage-
All the captures you have seen in these last posts of the 780Ti are from modified bios that allow you to increase voltage up to 1.21V, give more TPD, cancel boost and not enter throttling.
What is being waited for is the voltage unlock of the MSI as I have in the TITAN to give more voltage and at least reach 1.3V which should make them go up without problems up to 1400MHZ or more.. but I do see a problem and also, if you notice for the first time in a long time there is no professional overcloker who has beaten any record with them officially. And it is perhaps, due to their very high expense… the reference phases despite having been modified, maybe are not prepared to withstand what can come their way.
I certainly if I had TI and the unlock came out, I wouldn't think of overclocking them with extra voltage without a good RL...
I don't see it normal as I comment, that no overcloker or brand, has not already sponsored some record-breaking event and especially when the graphics card has just come out to the market and it is interesting to attract attention. Something must be going on that is more complicated than we thought at first..
Regards.
Hello guys.
It must also be taken into account how many months the Titan was on the market until they unlocked the voltage… Before going above 1225 was quite difficult...
With the 780ti it will happen a bit the same..
Greetings to all.
-
Hello guys.
You also have to take into account how many months the Titan was on the market before they unlocked the voltage... Before, going over 1225 was pretty hard...
The 780ti will be a bit the same..
Greetings to all.
Of course..it's clear that the B2 steeping is different and they raise much more stock.
But I don't see that real gain of +300MHZ in the core as effective in % as in the case of the TITAN. Maybe it's because of lack of voltage...
And then, what really bugs me is that no record has been beaten yet with them through volt-mod..it makes me think that maybe raising the voltage is more complicated than in their sisters. Maybe it has some kind of OCP implemented in hardware or whatever...
By the way, my last blunders with a single card at 1398MHZ...
Unigine valley (it's pure bandwidth too, but it's not bad..)
Unigine heaven:
I hope to get a response from the TIs...;)
Regards.
-
If I remember when some screens of death were red, things always get blown out of proportion and I think it's good that they report any problem, but they report everything, the graphics driver is like a snowball, curiously I have several cards and I put all the drivers including beta and they haven't killed a single one, in the end there comes a point where I don't even pay attention, I learn much more by reading forums and users especially if they are known, than by reading news.
It's like the press or television, they only tell what their news agency tells them, they all tell the same thing, and that agency tells only what interests them, that's why in the end they make you not believe anything.
saludos
Well, you've given a great example, and possibly without realizing it.
The issue of "red screens" wasn't what many were saying by making a scandal. With red screens nothing died and, beyond having to restart or at most turn off and on, nothing else happened.
This problem was due to... guess what. Versions of GPU-z pre-release to Kepler, used a way of accessing BIOS registers "at a fixed pitch", accessing the range of memory addresses reserved for interacting with the BIOS (mainly reading, but not only), and therefore when changing the position of some of these addresses or the registers and data they pointed to within the BIOS, it resulted in an "illegal access" to the BIOS of the Kepler graphics and as a result said red screen.
But nothing more than an off and on problem, at most. Of course every time you used an outdated version of GPU-z, it happened again. But... not a single problem with the hardware or a problem with the drivers, a pure and hard problem of how a program happily accessed the BIOS assuming that everything was the same as in Fermi and previous chips. But it wasn't.
You see how serious and how evil NVIDIA is with their drivers and poorly designed graphics, how provoked, all the dust you would have seen then about this issue, was pure fiction.
About graphics and drivers that kill cards, it's another story, I've seen many cases with that "accusation" that border on the ridiculous, to the point of users with brand new graphics cards that "die" when trying a 3D game for the first time (meaning they came damaged from the factory or the store, nothing more), and seeing how the user raves about the NVIDIA driver that supposedly killed the card (supposedly) since it had it installed for a while BEFORE trying the graphics in 3D, but quickly uninstalled it for a "safe" version of the driver and still... surely it had been the driver that killed the card. While in 2D and for a little while. XD
If that story were true, I would have gone bad having intensive sessions with these drivers and games, with the card overclocked, TDP unleashed, same with voltage, etc. That is, no, there are typical failures of defective units or simply bad luck, but many people invent stories for who knows what, to justify some of their own mistakes?
I haven't seen this said so cheerfully about any AMD driver, and with the 12.11 there were a thousand stability and graphic artifact problems, and you see the RMA ratios in the distributors I mentioned before, the data says that GCN fail significantly more, which doesn't necessarily mean they are bad, but if there were drivers that killed cards and the red screens were as important as some say, then RMA ratios would be higher.
That week I was "arguing" with a writer from CHW about the reason for the lack of material on the BS and other problems presented in the Hawaii. His explanation is that he waits for an official note from AMD, I deduce that he waits for permission from AMD or instructions ON what to publish about the topic...
A friend installed the latest CAT BETA and went through the BS trying to play BF4... He tried it after each forced reset of the machine... until the VGA wouldn't start anymore... His GA - HD 7970 GHZ is dead waiting for transport for RMA...
Well, these people at CHW had no qualms about publishing rumors of "failures" with NVIDIA graphics at the time, it seems they pay homage to some manufacturers and for those they don't, they do the dirty work for those who do.
I don't believe the BS story, you see everything in God's vineyard, but I do want those who report to do so truthfully, not inventing police novels but also not denying well-analyzed data. I see too many examples of these two extremes in the "technical press".
-
Obviously we all know the editorial line of CHW.. and it also makes sense, because let's not forget that it is aimed at South American countries, where access to Nvidia graphics and Intel processors is absolutely prohibitive for 90% of the usarios.No I honestly don't know what this is due to.. I suppose it's tariffs, taxes, fluctuations or who knows.. AMD, however, is much more affordable, not to mention the only reasonably payable in those parts...
And they logically, besides being a simple pamphlet in a network format of rumors and press news, put what they know will report more visits and more money... neither more nor less...
It happens everywhere.
Right here, we have thousands of AMD followers in Spain. However, sales in our country of the new R9 are very low compared to the GTX 780 and I would even dare to say the 780Ti.. because perhaps those who buy the latter are not spending all day giving it their all on the internet...
Normally, the Nvidia customer and the AMD customer in the high range, are usually not the same in age, experience, or even in purchasing power.. it's sad to say this, but it's true. Another thing is the official gadget enthusiasts who don't care about 8 or 80...
Best regards.
-
¡Esta publicación está eliminada! -
Of course.. it's clear that the B2 steeping is different and they raise much more stock.
But I don't see that real gain of +300MHZ in the core as effective in % as in the case of the TITAN. Maybe it's because of lack of voltage..
And then, what really bugs me is that no record has been beaten yet with them through volt-mod.. it makes me think that maybe raising voltage is more complicated than in their sisters. It might also have some kind of OCP implemented in hardware or whatever..
By the way, my last blunders with a single card at 1398MHZ…
Unigine valley (it's pure bandwidth too, but it's not bad..)
Unigine heaven:
I hope to get a response from the TIs..;)
Regards.
The 780ti lacks voltage, a bios mod revision and another of the nvidia drivers to give it their all. The sli performance at 1265 would be in graphics for sure in 21000 upwards in the firestrike…
I was testing the bios from skyn3t and it doesn't convince me. I'll go back to the one from svl7, which doesn't make artifacts. Then I'll run a bench of the unigine to see if I can make a decent note and get close to your oc, hehe...
Greetings..
-
The 780ti lacks voltage, a bios mod revision and another of the nvidia drivers to give it its all. The sli performance at 1265 would be in graphics sure at 21000 or above in the firestrike…
I was testing the bios from skyn3t and it doesn't convince me. I'll go back to the one from svl7, which doesn't make artifacts. Then I'll run a benchmark from unigine to see if I'm able to make a decent note and get close to your oc, hehe...
Greetings..
I didn't notice anything strange in the one from skynet. I attribute the artifacts to the memories being very tight or even the core.
At the same frequency doesn't the one from svl7 do it? this one has less TPD than the other..
Greetings.
-
I didn't notice anything strange in the skynet one. I attribute the artifacts to the memories being very tight or even the core.
At the same frequency, doesn't the svl7 do it? This one has less TPD than the other..
Best regards.
Both the svl7 and the skyn3t are the first versions of the 780ti.
In the coming months we will see new versions of the assemblers to work on, and they will do their own revisions..
Right now, and it's being corroborated by other users, the svl7 bios, although it doesn't allow as much power target, is more stable and doesn't produce artifacts at the maximum overclock of the card at 1.21v.
We'll see how things evolve, and whether it's possible to unlock the voltage soon...
Best regards.
-
Both the svl7 and skyn3t are the first versions of the 780ti.
In the coming months we will see new versions of the assemblers to work on, and they will do their own reviews.
Right now, and it is being corroborated by other users, the svl7 bios, although it does not allow as much power target, is more stable and does not produce artifacts at the maximum overclock of the card at 1.21v.
We will see how things evolve, and whether it is possible to unlock the voltage soon.
Best regards.
Honestly, I like the slv7 bios better because they are more neutral than the Skynet ones. But of course, both in the 780Ti and in the TITAN this is true, with overvoltage, I am limited by the TPD. And the slv7s cap that sooner.. but they really go finer if.
Best regards to you too..;)
P.D. And already to play eh? you can't live on firestrike.. or at least, we can't..jeje
-
wwwendigo, I know it was either GPU-Z or Everest or AIDA, that we had to use a newer one, but the news was red screen of death, all very sensationalist.
Now there are many R9s affected by black screens, they don't die but it's annoying, the thing with the flickering in videos or similar things, that's not given as news and you see people suffering from it in the forums, although mine happened in the 7000 series.
None are free of faults, I know someone whose 780 died after two days, some 770s have come out with problems for having such fast memory and they give graphic faults in games, there's always something defective whether it's graphics or motherboards, but if there are few cases then it's not worth giving importance, if you already see many cases that's when you should be careful.
I leave here the Topic that they will tell us something.
Greetings
-
Well, after the EVGA SC turned out to be a dud, today I received a GB to pair with the other one I already had, and after messing around with them, SLI at 1310-1980:
