980 Vs 970 Vs 780Ti a.k.a And you more

[Fassou]

You can follow if you want, but be careful with the way you do it.

Bye!

[Javisoft]

@Fassou:

You can follow if you want, but be careful with the way you do it.

Bye!

I won't follow, I'm going to leave the topic, I don't want any more controversy, it's useless to comment anything else, data has been provided, reviews have been posted, all that's left is for me to get another 970 G1 to fully get into their game... Because this is already bordering on ridiculous with the amount of user data and reviews that are on forums.

Goodbye.

[wwwendigo]

It is logical and normal that if someone uploads tests and performance of their units, the one who intends to refute them does not search the internet for the best possible results, but rather, having in possession the exact model in question, they should pass tests and performance on the units of that user.

It is as basic as the principle of equal opportunities. Not to mention that there are units on Mars that, taking advantage of the low temperatures of the Martian winter, reach unheard-of speeds.

There is no ounce of grace in uploading one's own results so that they can be "refuted" with a selection of the best results in each test among the thousands and thousands of results uploaded on the internet. It is just basic.

Apart from this, I see that beyond the acute reviewitis of some, there is genuine love in the air. I want a photo of the passionate kiss, the one at the end of the movie.

On the other hand, the debate 970 vs 780 Ti is sterile in itself, when one is going to crap for having low performance, surely the other will suffer the same. That is, you do not win any fundamental battle by demonstrating that one is 0.000001% faster than the other.

:love:

[Handrox]

@wwwendigo:

The logical and normal thing is that if someone uploads tests and performance of their units, the one who intends to refute them does not dive into the internet to look for the best possible results, but rather, having in possession the exact model in question, they should pass tests and performance on the units of that user.

It is as basic as the principle of equal opportunities. Not to mention that there are units on Mars that, taking advantage of the low temperatures of the Martian winter, reach unheard-of speeds.

There is no ounce of grace in uploading one's own results so that they can be "refuted" with a selection of the best results in each test among the thousands and thousands of results uploaded on the internet. It is just basic.

Apart from that, I see that beyond the acute reviewitis of some, there is genuine love in the air, I want a photo of the passionate kiss, the one at the end of the movie.

On the other hand, the debate 970 vs 780 Ti is sterile in itself, when one is going to crap for having low performance, surely the other will suffer the same way. That is, you do not win any fundamental battle by demonstrating that one is 0.000001% faster than the other.

:love:

Amen

[Javisoft]

@wwwendigo:

The logical and normal thing is that if someone uploads tests and performance of their units, the one who intends to refute them does not scour the internet to find the best possible results, but rather, having in possession the exact model in question, they should have the tests and performance of that user's units.

It's as basic as the principle of equal opportunities. Not to mention that there are units on Mars that, taking advantage of the low Martian winter temperatures, reach unheard-of speeds.

There is no ounce of grace in uploading one's own results so that they can be "refuted" with a selection of the best results in each test among the thousands and thousands of results uploaded on the internet. It's just basic.

Apart from that, I see that beyond the acute reviewitis of some, there is genuine love in the air. I want a photo of the passionate kiss, the one at the end of the movie.

On the other hand, the 970 vs 780 Ti debate is sterile in itself, when one is going to crap for having low performance, the other will probably suffer the same way. That is, you don't win any fundamental battle by proving that one is 0.000001% faster than the other.

:love:

Then you get banned from the forums for comments as implausible and useless as this, only seeking to create controversy...

My data is there, with throttling, while theirs are in full OC without throttling, I don't think I need to post photos with the thermal gun to show the core on the verge of burning with high consumption and the cooler at full capacity and with the temps over 60 degrees.

You talk about the "principle of equal opportunities" ¬¬, reference against the supposedly better custom version of the 970 according to its owner, do you want more in its favor besides what it already has? Don't be hypocritical, man, and stop sowing discord, you don't know which way to go to justify the unjustifiable...

[Ciclito]

I repeat that I passed the walk meters…(you seem not to understand) where I gave it my all is in the synthetics and also without going to the extreme since I had to sell them and I wasn't going to burn them. Wwwendigo knows well that it's true because by mp in a thread that I created in another forum, I said that I still had an ace up my sleeve and could improve my results by raising it to 1.312 and putting the bios that I made for it. All of that was at 1.27 and raising the tdp a little...

And on top of the one that came out worse of the two, because the other you can't even smell. At 1570mhz it ate your 970 at 1650mhz taking it two fps in heaven... something a bit hard to believe but there are the results. (If you did it many months ago I already tell you that it's because of drivers because I gained 2 fps at the same clocks with the penultimate ones vs others from months ago)

Leaving that aside and seeing that one of my blacks does the hell of time towards almost 61 fps all ultra with physx enabled in the metro LL when you barely reach 59 fps without physx (which are 6-7 fps less on average approx), you come to tell me that your lousy 780ti pulls more than my 970. When I have compared in a thousand tests my sli of blacks vs the 970 and I see the differences.

I'll tell you something when the 970 and the 980 came out during the first months that the 780ti was superior even the 780ti was at the level of the 980, but then months passed and the situation changed with the drivers, the maxwells started to perform more and the keplers were left more behind, to the point that to this day I can say with rotundity that the 970 is superior to the 780 ti without a doubt.

Now you can water them and put them in a freezer and demonstrate to me with synthetic tests (the same ones that I have contributed) how you clearly surpass me as you assure. Leave aside the webs and stories, if you want to prove your superiority upload that. For the moment you uploaded a fire strike, in which you beat me both in gpu score and in the general. So 1-0 for me.:fumeta:

I will continue waiting for your results... maybe with luck nvidia makes some magic driver for kepler and you surpass me. Until then I look forward to that supposed thrashing of your 780 ti good vs my normal 970.

[Ciclito]

¡Esta publicación está eliminada!

[Ciclito]

¡Esta publicación está eliminada!

[Javisoft]

@Ciclito:

Te repito que pase los metros de paseo…(que parece que no te enteras ) donde le di caña es en los sintéticos y tampoco sin llegar al extremo ya que tenía que venderlas y no las iba a quemar. Wwwendigo sabe bien que es cierto porque por mp en un hilo que creé en otro foro, le dije que aún guardaba un as en la manga y podía mejorar mis resultados subiéndola a 1,312 y poniendo la bios que le hice para ello. Todo eso era a 1,27 y subiendo algo el TDP...

Y encima de la que me salió peor de las dos, porque la otra ni la hueles. A 1570mhz me comía tu 970 a 1650mhz sacándole dos FPS en el Heaven... cosa algo difícil de creer pero ahí están los resultados. (Si lo hiciste hace muchos meses ya te digo que es por drivers porque yo gané 2 FPS a los mismos clocks con los penúltimos vs otros de meses atrás)

Dejando eso a un lado y viendo que una de mis Black hace la ostia de tiempo hacia casi 61 FPS todo ultra con PhysX habilitado en el metro LL cuando tú a duras penas llegas a 59 FPS sin PhysX (que son 6-7 FPS menos de media aprox), me vienes a decir que tu roñosa 780Ti tira más que mi 970. Cuando he comparado en mil pruebas mi sli de Blacks vs el de 970 y se las diferencias.

Te diré algo: cuando salieron las 970 y la 980 durante los primeros meses sí que la 780Ti era superior, incluso la 780Ti estaba a la altura de la 980, pero luego pasaron los meses y cambió el asunto con los drivers, las Maxwell empezaron a rendir más y las Kepler se quedaron más rezagadas, hasta el punto que a día de hoy aseguro con rotundidad que la 970 es superior a la 780Ti sin dudas.

Ahora ya puedes pasarlas por agua y meterlas en un congelador y demostrarme con tests sintéticos (los mismos que yo he aportado) cómo me superas claramente como tú aseguras. Déjate de webs ni historias, si quieres probar tu superioridad sube eso. De momento subiste un Fire Strike, en el cual te gané tanto en GPU Score como en el general. Así que 1-0 para mí.:fumeta:

Seguiré esperando tus resultados... igual con suerte NVIDIA hace algún driver mágico para Kepler y me superas. Hasta entonces espero con ansias ese rapapolvo supuesto de tu 780Ti buena vs mi 970 normalita.

De verdad, tu prepotencia y chulería solo es igualable a tu ego … baja los humos si quieres que sigamos hablando :verguen:...

Vamos por partes, los 1.31V... va a ser que no:

http://www.onsemi.com/pub_link/Collateral/NCP81174-D.PDF

Datasheet de tu regulador de voltaje, si entiendes algo de lo que por ahí pone y debido a tu número de fases tu voltaje máximo por PCB es 1.27V que con el Vdroop se quedará en 1.265V, que tú en BIOS puedas deslimitar el TDP y poner 1.6V como tienen las EVGA FTW pues fantástico (que no lo van a coger ni de broma lógicamente), que el AB te lo pone y tú te lo crees fantástico, coge un polímetro y mira lo que sale en tu tarjeta, eso no lo haces ¿verdad;D?

O sea que tu as en la manga … es más bien humo, por no decirte que la diferencia entre 1.265V y 1.31V en Maxwell son apenas 15 MHz, probado en una G1... como dije antes, a Maxwell lo puedes exprimir si necesidad de sufrir throttling o necesitar BIOS custom (a no ser que quieras batir récords mundiales y con voltmod).

Estoy hasta el gorro de modificar BIOS y del prueba-error, teniendo que re-flashear con segundas gráficas cuando me paso en algún parámetro y la primera VGA se queda off como para que vengas a decirme lo contrario, para más info sobre lo anterior:

Extract and Flash GTX 970 and 980 Firmware - Zoson's H2O and AIR Custom BIOS

Luego te contradices cuando afirmas que la 970 rinde más que el GK110B, tú mismo afirmas que tus Titan dan más puntuación con PhysX activo que las 970 sin él … en Heaven por debajo, en 3DMark por debajo... Tú solo te das las respuestas :facepalm:

En cuanto a drivers, NVIDIA ya ha reconocido bugs para Kepler que según dicen, están subsanando en los sucesivos drivers, ¿que lo han hecho a posta? Puede ser, el rendimiento era superior antes, aunque a mí me da mejor rendimiento por poco estos últimos...

Por lo demás, ayer noche con algo menos de temperatura, 1267 por aire:

Con PhysX 4 FPS menos … Y como digo, esta gráfica no está saliendo buena...

En 3DMark no creo que pueda superar a Maxwell, ni siquiera a la 970, por lo menos con la reference, ya veremos con la Classi, de todas formas, como siempre he dicho y mantenido, es un test que no representa el rendimiento real, ni de un sistema, ni de una tarjeta, solo vale para competir con lo que mejor puntuación de en ese momento, no refleja la realidad como tal.

De todas formas estamos hablando de juegos, ¿no? Porque en Heaven te meto bastante, en 3DMark estoy a la par y en juegos... despegado, como digo, con una reference y por aire con throttling.

No sé por qué intentas tergiversar los números y la realidad de lo que se está viendo <:(.

La Classi ya veremos cuánto te saca .

EDITO:

Como curiosidad, y salvando las distancias en FPS por drivers y plataforma lógicamente, aquí se puede ver más o menos lo que se está viendo en esta rama en cuanto a diferencias se refiere:

http://www.overclockersclub.com/reviews/msi_gtx_970_gaming_4g/4.htm

Todas con máximo OC por aire, se ve perfectamente que aun a 1600/2100 no la puede superar, y eso que la 780Ti va a 1291 y fijo que tiene throttling aun con el venti al 100%.

Un saludo :risitas:

[Ciclito]

@Javisoft:

De verdad, tu prepotencia y chuleria solo es igualable a tu ego … bajate los humos si quieres que sigamos hablando :verguen:...

Vamos por partes, los 1.31v... va a ser que no:

http://www.onsemi.com/pub_link/Collateral/NCP81174-D.PDF

Datasheet de tu regulador de voltaje, si entiendes algo de lo que por ahi pone y debido a tu numero de fases tu voltaje maximo por pcb es 1.27v que con el vdroop se quedará en 1.265v, que tu en bios puedas deslimitar el tdp y poner 1.6v como tienen las evga FTW pues fantastico ( que no lo van a coger ni de broma logicamente ), que el AB te lo pone y tu te lo crees fantastico, coge un polimetro y mira lo que sale en tu tarjeta, eso no lo haces verdad ;D?

Osease que tu as en la manga … es mas bien humo, por no decirte que la diferencia entre 1.265v y 1.31v en maxwell son apenas 15 mhz, probado en una g1... como dije antes, a maxwell lo puedes estrujar si necesidad de sufrir throttling o necesitar bios custom ( a no ser que quieras batir records mundiales y con voltmod ).

Estoy hasta el gorro de modificar bios y del prueba-error, teniendo que re-flashear con segundas graficas cuando me paso en algun parametro y la primera vga se queda off como para que vengas a decirme lo contrario, para mas info sobre lo anterior:

Extract and Flash GTX 970 and 980 Firmware - Zoson's H2O and AIR Custom BIOS

Luego te contradices cuando afirmas que la 970 rinde mas que el gk110b, tu mismo afirmas que tus titan dan mas puntuacion con physix activo que las 970 sin el … en heaven por debajo, en 3dmark por debajo... Tu solo te das las respuestas :facepalm:.

En cuanto a drivers, nvidia ya ha reconocido bugs para kepler que segun dicen, estan subsanando en los sucesivos drivers, que lo han hecho a posta? puede ser, el rendimiento era superior antes, aunque a mi me da mejor rendimiento por poco estos ultimos...

Por lo demás ayer noche con algo menos de temperatura, 1267 por aire:

Con physix 4 fps menos … Y como digo, esta grafica no está saliendo buena...

En 3Dmark no creo que pueda superar a maxwell, ni siquiera a la 970, por lo menos con la reference, ya veremos con la classi, de todas formas, como siempre he dicho y mantenido, es un test que no representa el rendimiento real, ni de un sistema, ni de una tarjeta, solo vale para competir con lo que mejor puntuacion de en ese momento, no refleja la realidad como tal.

De todas formas estamos hablando de juegos no? Por que en heaven te meto bastante, en 3dmark estoy a la par y en juegos... despegado, como digo, con una reference y por aire con throttling.

No se por que intentas tergiversar los numeros y la realidad de lo que se está viendo <:(.

La classi ya veremos cuanto te saca .

EDITO:

Como curiosidad, y salvando las distancias en fps por drivers y plataforma logicamente, aqui se puede ver mas o menos lo que se está viendo en esta rama en cuanto a diferencias se refiere:

MSI GTX 970 Gaming 4G Review » Page 4 - MSI GTX 970 Gaming 4G Testing: Setup & Overclocking - Overclockers Club

Todas con maximo oc por aire, se ve perfectamente que aun a 1600/2100 no la puede superar, y eso que la 780Ti va a 1291 y fijo que tiene throttling aun con el venti al 100%.

Un saludo :risitas:

Este era mi as en la manga:

Unlocking the Nvidia Power Limit (all cards!) | Overclocking.Guide

Dejo la discusión me aburre soberanamente, si tu 780ti rinde mas, supera mis 7 bencheos donde iba a tope (metro destartalo ya que iba a medio gas) cuando me superes en los siete con tus bloques y rl te dare la razón hasta entonces no cambio de idea. Ya te dije que las kepler tiraban algo mas al principio. Al cabo de los meses eso fue cambiando. Y a dia de hoy eso es una realidad. Si compartes esa opinion demuéstralo tu mismo con tus propias pruebas y las adjuntas al hilo como he hecho yo.

Si me superas en todos con holgura (pues dices que se la come ) me quitare el sombrero y te daré la razón sin problema. Pero deberías superaros todos… sino te habré owneado de nuevo.

Salu2 y a la espera de tus pruebas (propias, no de webs)

[Javisoft]

@Ciclito:

This was my trump card:

Unlocking the Nvidia Power Limit (all cards!) | Overclocking.Guide

I leave the discussion as it bores me immensely, if your 780ti performs better, surpass my 7 benchmarks where I went full throttle (I stopped measuring as it was running at half throttle) when you surpass me in all seven with your blocks and RL I will give you credit until then I won't change my mind. I already told you that Kepler cards initially performed a bit better. Over the months that changed. And today that's a reality. If you share that opinion, demonstrate it yourself with your own tests and attach them to the thread as I have done.

If you surpass me in all of them by a wide margin (since you say you can eat it up) I will take my hat off and give you credit without any problem. But you should surpass all of them... otherwise I will have owned you again.

Best regards and waiting for your tests (your own, not from websites)

But let's see... I'm flabbergasted, your trump card? :ugly:

You can spot a liar faster than a cripple, I bet you've been looking all day for how to raise the voltage to 1.31v and you're trying to sell me the power limit unlocking mod :dead: (or what's the same, the power target mod) as if I were a total noob and you're completely mistaken...

I'm going to explain what the mod is for... It's for nothing, absolutely nothing, and I'll explain why:

The only thing you're doing is removing the SMDs that act as shunt resistors on the PCB, or in other words, increasing the card's TDP by removing a protection against watt/ampere overvoltage, replacing it with a conductive material like silver paint (yes, that which many of us used to connect the pins on Athlons that had their multipliers unlocked).

In no case will it provide more voltage and instead it's very possible that you'll end up frying either the voltage regulator, the card, or, through excessive overclocking, damaging the power supply if it doesn't have OCP.

The funny thing is that you release this as if you did the same thing via BIOS in the power table section, which means you don't know how to modify BIOS, not a clue about the subject (another catch), moreover, the author of the guide tells you that if you already have a modded BIOS (like Skynet) this won't bring you any benefit...

What he doesn't say is that those shunt SMDs are the ones that lower the frequency when you go too far with the voltage and prevent the chip or the controller from frying...

Moreover... if you force the fixed frequency via BIOS by disabling boost and also by unlocking the voltage and power limit, what will happen to you with a lot of luck is that the OCP of the power supply will trip and either the PC will shut down or restart.

That is, if you haven't fried any of the previously mentioned parts, lucky you...

Now that in Metro LL I get more FPS, you come with synthetic tests, before with 1.31v, then you weren't going full throttle and so on... in other forums with peaks of 1800 MHz on the GPU (without voltmod and with ref PCB, come on... :wall: )

If you've been banned and kicked from so many forums it's because of things like this...

Before it was in games, now we only limit ourselves to asking in 3Dmark, where you know I'll have to push to the limit to tell you again that I've surpassed you and by a lot, when I do it will be that I'm going with water and you by air (although the test is favorable to Maxwell).

The funny thing is that in the middle of a heatwave and with a GPU with a 73% ASIC and throttling I'm taking the credit (the one that others ask for so much and fewer reviews... ) another thing is that it's not convenient for you to see it...

I still have to wait for my classi to arrive and for another friend's to arrive, I don't want to tell you where the comparison between custom vs custom will end up.

Well, goodbye...

[whoololon]

Damn, how nice it is to have a popcorn thread like this every now and then. :sisi:

[Javisoft]

@whoololon:

Damn, how nice it is to have a popcorn thread like this every now and then. :sisi:

Make room, the curves are coming, I'm already ahead in Firestrike and by AIRE:

NVIDIA GeForce GTX 780 Ti video card benchmark result - Intel Core i7-3930K Processor,ASUSTeK COMPUTER INC. RAMPAGE IV EXTREME

I had to disable boost in bios because of throttling and increase the maximum fan speed from 85 to 100% so that it doesn't fry with the reference cooler and can tolerate more mhz.

The rest is as you can read in the gpu-z bios at home. I have more data but for now I'm not going to upload them XD.

Best regards.

[wwwendigo]

@wwwendigo:

The logical and normal thing is that if someone uploads tests and performance of their units, the one who intends to refute them does not dive into the internet to look for the best possible results, but rather, having in possession the exact model in question, they should pass tests and performance on the units of that user.

It is as basic as the principle of equal opportunities. Not to mention that there are units on Mars that, taking advantage of the low temperatures of the Martian winter, reach unheard-of speeds.

There is no ounce of grace in uploading one's own results so that they can be "refuted" with a selection of the best results in each test among the thousands and thousands of results uploaded on the internet. It is just basic.

Apart from this, I see that beyond the acute reviewitis of some, there is genuine love in the air. I want a photo of the passionate kiss, the one at the end of the movie.

On the other hand, the debate 970 vs 780 Ti is sterile in itself, when one goes to hell for having low performance, surely the other will suffer the same. That is, you do not win any fundamental battle by demonstrating that one is 0.000001% faster than the other.

:love:

I quote myself because I love seeing debates where the differences in a test after several attempts and "yours is longer" "mine is longer" are left at:

+0.598% in favor of the 780 Ti with OC.

Yessss…. indeed. 14197 points in the graphics tests on the GTX 970 side, 14282 of the 780 Ti. OC to death in the second and possibly in the first (I am clear that almost 1300 in a GK110 is approaching its absolute limits, unfortunately in the GTX 970 1449 in a BOOST ACTIVATED does not tell us anything, it is a frequency in principle low for the GM204 although that depends on how much the boost adds to the chip, if it is closer to 1500 or 1600 MHz, etc).

Great score, without a doubt this totally transforms the panorama in favor of the 780 Ti.

I think you do not realize how ridiculous it sounds to those of us who see the competition from the outside when you talk about "crushing" one graphics card or the other by differences of almost half a percentage point, or shaving less than a frame difference in games... :ugly:

The minimum sensible thing would be to recognize the equality of results when the differences are so negligible. But no, popcorn and let this war of attrition and Pyrrhic victories continue.

[whoololon]

Summary of the thread:

@Ciclito:

[…]Javisoft takes the best of your 780ti and tries to beat some of my results when I had the 970 (and note that it was the bad one, since I sold the good one before those benchmarks) many of those results without a modified bios.
Let's see when you beat them…ji ji ji... in games a 970 with oc max I already tell you that surpasses a 780ti also at max oc in many cases. (remember that I had two blacks top notch... and also sli 970 pretty decent) Admit it already man...[…]
Ale you already have entertainment to prove the superiority of one of your 780ti watercooled vs a 970 with reference pcb air cooled (that is if you had, good dissipation):sisi:

@Javisoft:

Get ready for some curves, I'm already ahead in Firestrike and by AIR:

NVIDIA GeForce GTX 780 Ti video card benchmark result - Intel Core i7-3930K Processor,ASUSTeK COMPUTER INC. RAMPAGE IV EXTREME

I had to disable boost in bios because of throttling and increase the maximum fan speed from 85 to 100% so that it doesn't fry with the reference cooler and tolerate more mhz.

The rest as you can read in the gpu-z bios at home. I have more data but for now I'm not going to upload them XD.

Best regards.

@wwwendigo:

[…]+0,598% in favor of the 780 Ti with OC.[…]

[…]Great score, this definitely changes the landscape in favor of the 780 Ti.[…]

Point made.

[Javisoft]

@wwwendigo:

I quote myself because I love seeing debates where the differences in a test after several attempts and "you more" "mine is longer" are reduced to:

+0.598% in favor of the 780 Ti with OC.

Siiii…. indeed. 14197 points in the graphics tests on the GTX 970 side, 14282 for the 780 Ti. OC to death in the second and possibly in the first (I'm sure that almost 1300 in a GK110 is approaching its absolute limits, unfortunately in the GTX 970 1449 in a BOOST ACTIVATED doesn't tell us anything, it's a frequency in principle low for the GM204 although that depends on how much the boost adds to the chip, if it's closer to 1500 or 1600 MHz, etc).

Great score, without a doubt this completely transforms the landscape in favor of the 780 Ti.

I think you don't realize how ridiculous it sounds to those of us watching the competition from the outside when you talk about "crushing" one card or the other by differences of almost half a percentage point, or shaving less than a frame difference in games... :ugly:

The minimum sensible thing would be to recognize the equality of results when the differences are so negligible. But no, popcorn and let this war of attrition and Pyrrhic victories continue.

Do we better not talk about the 9.22% I put in heaven? Or the almost 6% in metro? :ugly: What a coincidence, thank goodness I'm on air and water the reference will make more foam. What a shame that you have to rely on what interests you, 3Dmark, it's well known that kepler scores less... in tomb raider it would have been even more of a percentage difference, but no, I'm stubborn on air, when I see that it's not possible to get more performance, I'll switch to water and we'll talk again ;D.

You take the data as you like and the comeback is yours and you'll have even more, I guarantee it.

For now, above all the data I'm showing, when I have them all we'll see in how many I'm below, but let me anticipate it, it's sure to have to do with 3Dmark, because of course, it passes tests everywhere, but only puts the ones it's interested in, lest it get caught with X fps in some game and be exposed...

I'm following his script, the one he's interested in, when it ends, I'll present mine and we'll see the excuses flow down the river...

For now you're not affirming as in other places that the 970 is faster, now it's a little slower, the next thing is that you see what many websites have been showing for a long time, a 7% in favor of kepler on average...

Then you can cry about the boost of the 970 as much as you want, the result will be the same.

[Javisoft]

@whoololon:

Thread summary:

Point ball.

Re-read the thread if you are so kind :facepalm:...

[wwwendigo]

@Javisoft:

Of the 9.22% that I put into heaven, let's not even talk about it? What about the almost 6% in metro? :ugly: What a coincidence, it's a good thing I'm in air and water, the reference will make more foam. Too bad you have to rely on what interests you, 3Dmark, it's well known that kepler scores less... in tomb raider it would have been even more of a percentage difference, but no, I'm stubborn in air, when I see that it's not possible to get more performance, I'll switch to water and we'll talk again;D.

You take the data as you like and the comeback is yours and it will be even more yours, I guarantee it.

For now, above all in all the data that I'm showing, when I have them all we'll see in how many I'm below, but let me anticipate it, it's sure to have to do with 3Dmark, because of course, he tests everything everywhere, but only puts the ones that interest him, lest he get caught with X fps in some game and be exposed...

I'm following his script, the one that interests him, when it ends, I'll present mine and we'll see the excuses flow like a river...

Then you can cry about the boost of the 970 as much as you want, the result will be the same.

Calm down a bit, I'm talking about your last comment, or is it not possible to comment here in "duel of who has the biggest"? :facepalm:

About Heaven, man, look how good, a 9%? But what about everything else where the differences are simply ridiculous? The issue is that you're not taking into account that ciclito already put those GTX 970s at much higher frequencies, especially the OTHER one that was the first to get rid of, I'm not saying this to give him credit, it's that I've seen this commented on other forums and some results with OC records, when he says that the "bad" one is the one that was left, he's not saying it as a last resort.

What do you want me to tell you, when you put a result of Last Light without tessellation in the result (that's what it says whether you like it or not in your screenshot, I don't have to trust someone's word over what the html result says that the game saves, yes, if it's for a sparrow copy and stories, then either you have the game correctly, or better yet, don't even mention it), well, I didn't even take it into account to look at the following results... :troll: But even so, after reviewing it (besides the fact that Last Light is not a very promaxwell game precisely):

I'm surprised that you say that in LL there is "almost a 6%" difference in your favor, when it's more like a meager 5% (specifically, 5.272108%), something that from a mathematical language when applying rounding doesn't make any sense to do it upwards (if it were 5.51% another bird would be singing). Don't inflate results.

Anyway, are you really debating over differences that are basically insignificant? Tell me what use you have for such "thick differences" in a 780 ti with an "up to 9% faster" than a 970, and both pushed to the limit.

Seriously, what real gameplay extra do you get with that "up to" that doesn't happen even half the time, what options will you be able to activate that can't be done on a 970. There, even having more VRAM (even if it's "small") determines more than the other.

You give me a stable difference of 15-20% and not of "up to" and we'll start talking about differences that are noticeable in performance. It's not rocket science, eh?, serious things start at no less than 30-35%, and if we talk about 50% it's when you really IMPROVE with a graphics change. What I'm doing is pointing out that the "debate" doesn't make sense and even less so with the heat that's on both sides, for differences that aren't worth putting up statues for, precisely.

But don't you see that there are more differences between both graphics due to the game used than by the average power they show (and here you'll see why I say that about Last Light)? An image like this:

It depends much more on the game compared than on anything else.

In Crysis 3, your 780 will surely sodomize the 970 (exaggeration, but it will get a good advantage), the opposite if we talk about Far Cry 4 (note, the game was configured to run well on a 770 without VRAM problems), Battlefield 4 runs much better than the 3, etc.

That is, if you get stupid by making "war" it's enough to throw at each other the games that run best on each of the two graphics as "argument-counterargument" to infinity.

I really don't care if the 780 ti is the "fastest", since they weren't even close in price tag. Even less for these average differences. But I already told you that I have no problem with the 780 Ti outperforming the 970. That's why I see as ridiculous that eagerness to compete when the differences are scarce, do you need that tone of "annihilation" when talking about such similar results? It's a bit comical, on both sides.

Relax. We're talking about graphics with similar performance, and no one's life depends on it or a fortune is won in the form of a bet.

[Javisoft]

@wwwendigo:

Tranquilízate un poco que estoy hablando de tu último comentario, ¿o es que no se puede comentar aquí en "duelo de quién la tiene más gorda" que os marcáis? :facepalm:

Sobre el Heaven, hombre, mira qué bien, un 9% ¿pero qué pasa en todo lo demás donde las diferencias son simplemente ridículas? El tema que no estás teniendo en cuenta es que ciclito ya puso esas GTX 970 a frecuencias mucho mayores, sobre todo la OTRA que fue la primera que se deshizo, no lo digo para darle la razón, es que esto ya se lo ví comentado por otros lares y algún que otro resultado con récords de OC, cuando te dice que la "mala" es la que le quedaba, no lo dice como recurso de última hora precisamente.

Yo qué quieres que te diga, cuando has puesto un resultado de Last Light sin teselación en el resultado (es lo que pone te guste o no en tu captura, yo no tengo porqué confiar en la palabra de alguien por encima de lo que pone el html del resultado que guarda el juego, sí, si es por una copia sparrow e historias, pues entonces o se tiene el juego correctamente, o mejor ya ni comentarlo), pues como que ni lo tuve en cuenta mirar más los resultados siguientes… :troll: Pero aún así, tras revisarlo (además de que Last Light no es un juego demasiado promaxwell precisamente):

Me sorprende que digas que en LL hay "casi un 6%" de diferencia a tu favor, cuando más bien es un 5% escaso (en concreto, un 5,272108%), algo que desde el lenguaje matemático a la hora de aplicar redondeos no tiene ningún sentido hacerlo hacia arriba (si fuera un 5,51% otro gallo cantaría). No me infles resultados.

Sea como sea, ¿realmente estáis debatiendo por diferencias básicamente insignificantes? Dime de qué te sirve tan "gruesas diferencias" en una 780 ti con un "hasta un 9% más rápida" que una 970, y ambas puestas a reventar.

En serio, qué jugabilidad real extra obtienes con ese "hasta" que no llega ni la mitad de las veces, qué opciones podrás activar que no se pueda en una 970. Ahí hasta determina más que una tenga más VRAM (aunque sea siendo "cojita") que la otra.

Tú dame un 15-20% de diferencia estable y no de "hasta" y ya empezamos a hablar de diferencias perceptibles de rendimiento. Tampoco para tirar cohetes, ¿eh?, las cosas serias empiezan en no menos de un 30-35%, y ya si hablamos de un 50% es cuando se mejora DE VERDAD en un cambio de gráficas. Yo lo que hago es señalar que el "debate" no tiene sentido y menos en lo encendido que está por ambas partes, por diferencias que no son para poner estatuas, precisamente.

¿Pero no veis que hay más diferencias entre ambas gráficas por el juego usado que por la potencia media que demuestren (y aquí verás porqué digo lo de Last Light)? Una imagen como muestra:

Depende mucho más del juego comparado que de cualquier otra cosa.

En Crysis 3 seguro que tu 780 sodomiza a la 970 (exageración, pero le sacará buena ventaja), lo contrario si hablamos de Far Cry 4 (ojo, estaba configurado el juego para ir bien en una 770 sin problemas de VRAM), Battlefield 4 va bastante mejor que el 3, etc.

Que vamos, si os ponéis tontos al hacer "la guerra" sólo basta con arrojaros los juegos que mejor van en cada una de las dos gráficas como "argumentación-contraargumentación" hasta el infinito.

A mí realmente me da un poco igual que sea la 780 ti la "más rápida", ya que ni siquiera han estado ni cerca en tag de precio. Menos por estas diferencias medias. Pero ya te digo que no tengo ningún problema en que la 780 Ti le saque la cabeza a la 970. Por eso veo ridículo esa afán de competencia cuando las diferencias son escasas, ¿hace falta ese tono de "aniquilación" cuando hablamos de resultados tan parecidos? Es que es un poco cómico, por ambas partes.

Relax. Que estamos hablando de gráficas parecidas en rendimiento, y no le va la vida a nadie ni se gana una fortuna en forma de apuesta.

Alucino cada dia mas contigo, se te cae la mascara de entendido macho … enserio crees que la puntuacion que da es sin teselación? en very high? enserio? quieres hacerme creer que el no iba a tope en el test?

Perdona pero no soy idiota, como digo, 970 a 1620 estables en mi mano, codo con codo contra mi 780Ti, media de diferencia 7%... Cuando me muestres una 970 que te aguante eso, por favor, avisame, que no es facil, tu amigo ciclito consiguio 1800 mhz sin voltmod XDDDDDDDD, claro y con 1v :facepalm:

Lo que te estoy mostrando es la verdad al 100%, insisto, yo no miento, que quedo por debajo, fantastico, no hay problema, pero la verdad por delante, se puede decir lo mismo de algunos?

Claramente no, no pensaras que me voy a tragar un Heaven a 1570 mhz dando esa puntuacion... que no tio, que hay mil foros con scores para comparar, será casualidad que nunca rankea en los scores oficiales? Si es tan tan tan bueno, por que no se le ve?

Se la mide constantemente alardeando de hardware y puntuacion pero no entra a competir?

En fin, ni seré el primero ni el ultimo que lo pilla falseando datos y con fantasmadas...

Luego, le hace test, capturas, videos, hasta a su ombligo y me estas diciendo que me está poniendo resultados de la gpu " mala "? Cuando tengo datos aqui que son inferiores, equivalentes y superiores con la G1? Mira, va a tope, como voy yo, la falsa modestia para otros...

Yo estoy con una grafica que no es buena, con peor sistema de refrigeracion y en plena ola de calor, lo que está haciendo la pequeña de mi derecha, es echarle unos "·$%&/(? con una media de 95º que es para quitarse el sombrero y no los 60º maximo que tendria su Ichill y tirando por lo alto.

Me pides un 15-20% de mas, cuando la 980Ti le mete a la 780Ti un 25% de media, vamos lo que me estas pidiendo es imposible sin LN2 y totalmente absurdo para la comparacion INJUSTA, de momento claro.

Yo lo que te estoy mostrando es que una grafica es mas rapida que la otra y no por un misero porcentaje si quitas 3Dmark, donde compitien en inferioridad por el propio test, aun asi por delante y sin excusas...

No se que vienes a reclamar, si la 980 le mete un 10% a la 970 de media y nadie discute que la primera es mas rapida y con un frametime mas constante, que como bien reclamas y se ha visto ya, es una " diferencia perceptible de rendimiento ".

Si esperas tus porcentajes magnificos, que ninguna review ni usuario jamas ha mostrado, vuelve al pais de la piruleta, por que eso aqui no va a pasar a no ser que meta nieve carbonica o LN2, y de momento, lo descarto, siento desilusionarte ¬¬

En vez de meter mas madera, que es lo unico que haces y encima con paridas como la teselación off, coge tu 970, que creo que tienes y aporta datos, si no la tienes, como tu bien decias en otro hilo, estas hablando de oidas, no hace falta que diga mas no?

Pues eso a seguir bien ;D y buen rollo .

[Javisoft]

Well, seeing that I'm not going to get much more out of the air with the heat, I've decided that I'm going to push the card to its limit. To do this, I've put a nickel-plexy block on it, an EK backplate and a fugipoly ultra extreme thermalpad.

The configuration for all the tests was:

3930K @ 4.8 ghz
Asus rampage IV Extreme
2X4 Corsair Dominator GT 2133 Cas7
Raid 0 840 Pro
Evga Gtx 780Ti SC
EK Supremacy EVo block
Alphacool D5+Top
DDC+Top+Depo
EK Bahia Depo
TFC Xchanger 360
EK XT45 360
TFC Xchanger 120
7 X Corsair SP120HP

First results:

• 6.4 fps ( + 4,21% vs previous, no data for the 970 ¬¬ )

• 1.83 fps ( + 2.95% vs previous and + 8,38% vs 970 oc )

• 0.44% vs previous ;D + 2,23% vs 970 oc

Tomorrow I'll try to run the others, but seeing the lack of increase in the super 3Dmark... There's an average increase of 7% in games but practically none in firestrike, fantastic indeed :ugly:...