Lorem ipsum dolor sit amet

[I/O]

Firstly, the only one who said anything about extra VRAM making a GPU faster is you. 

Now your just trolling. I have said from the beginning that Vram count does not make a GPU faster LOL you said  extra Vram makes a card faster not the OC LOL and this has been the basis of this argument that you have already lost. Many others here will back me up on this one. Crysis 3 never goes above 3GB Vram even when using Ultra settings 8xMSAA and at the settings used in your review you posted they were using SMAA which is allot less vram intensive

[BiG StroOnZ]

Now your just trolling. I have said from the beginning that Vram count does not make a GPU faster LOL you said  extra Vram makes a card faster not the OC LOL and this has been the basis of this argument that you have already lost. Mant other here will back me up on this one.

 

I never said that VRAM makes the GPU faster. Not once, I even asked for you to quote me saying that. Which you couldn't do. Instead you changed your argument to the next. It's not the VRAM itself that makes the card faster. Hur dur. It's the 3GB of VRAM becoming a bottleneck in certain scenarios that makes a card with more VRAM perform better (a la 4GB R9 290X and 6GB 780). This has been explained numerous times to you. Every single time it has been exclaimed you change your tune to the next topic and try to find something else to nit pick about.

 

Nobody is backing you up, you are all alone on this. You've been proven many times over wrong and you still won't accept the loss. Even when someone came in and told you, you lost. You still won't accept the lost. LMAO. You are the king of trolls. 

You ignore the data that proves you wrong and start off spewing some more nonsense. What's going to be next? Is it the OC, because I already proved you wrong about it being the OC. You even proved yourself wrong about it being the OC. LMAO. Remember that? Happened like twenty minutes ago when you kindly posted that graph. 

[I/O]

It's the 3GB of VRAM becoming a bottleneck in certain scenarios that makes a card with more VRAM perform better (a la 4GB R9 290X and 6GB 780).

OK now we are communicating a little better here as it seemed like you thought that the extra Vram was making the card faster. Anyway I already showed that at the settings used with was 1600P and SMAA that Crysis 3 does not even come close to using up 3GB Vram in fact it only uses 1.6GB Vram at those setting. So given the data presented we can rule out any Vram bottle necking in this given scenario unless we can prove that Crysis 3 uses more than 3GB Vram with the settings used in the graphs you posted.

[BiG StroOnZ]

OK now we are communicating a little better here as it seemed like you thought that the extra Vram was making the card faster. Anyway I already showed that at the settings used with was 1600P and SMAA that Crysis 3 does not even come close to using up 3GB Vram in fact it only uses 1.6GB Vram at those setting. So given the data presented we can rule out any Vram bottle necking in this given scenario unless we can prove that Crysis 3 uses more than 3GB Vram with the settings used in the graphs you posted.

 

That's because Crysis 3 has a command that limits it's VRAM usage: r_TexturesStreamPoolSize 

 

At default it is set to 512MB, if you set it to 1024MB the game will naturally utilize more VRAM. 

 

[I/O]

That's because Crysis 3 has a command that limits it's VRAM usage: r_TexturesStreamPoolSize 

 

At default it is set to 512MB, if you set it to 1024MB the game will naturally utilize more VRAM. 

 

I would like you to prove that Crysis 3 uses more than 3GB of Vram at the settings you are talking about in the graph you posted.

[BiG StroOnZ]

I would like you to prove that Crysis 3 uses more than 3GB of Vram at the settings you are talking about in the graph you posted.

 

If we are going to play that game, then I would like you to prove that Crysis 3 doesn't use more than 3GB of VRAM at those settings on the "Safeties Off" level. Which is where Hard OCP tested. According to the graph you are using, we don't know what level they use therefore we cannot take that graph as definitive information for how much VRAM is used. This is because as the author said in the benchmarks you are using:

 

 

 

Please keep in mind though that memory utilization will vary a bit per level, it's not a very precise thing to measure.

[I/O]

If we are going to play that game, then I would like you to prove that Crysis 3 doesn't use more than 3GB of VRAM at those settings on the "Safeties Off" level. Which is where Hard OCP tested. According to the graph you are using, we don't know what level they use therefore we cannot take that graph as definitive information for how much VRAM is used. This is because as the author said in the benchmarks you are using:

The point was that you claimed on the settings used in the benchmark you posted that the performance increase was due to the extra Vram. I proved that the game at those settings did not even use up close to 3GB Vram therefore eliminating the Vram as the reason for the gain in framerates. Once again there is no proof that Crysis 3 uses more than 3GB vram at the settings used however I proved that it in fact uses only around 1.6GB Vram with the settings used so we can mark that metric of the list.

[BiG StroOnZ]

The point was that you claimed on the settings used in the benchmark you posted that the performance increase was due to the extra Vram. I proved that the game at those settings did not even use up close to 3GB Vram therefore eliminating the Vram as the reason for the gain in framerates. Once again there is no proof that Crysis 3 uses more than 3GB vram at the settings used however I proved that it in fact uses only around 1.6GB Vram with the settings used so we can mark that metric of the list.

 

I also proved to you that it wasn't the OC, you actually proved yourself that it wasn't the OC by providing graphs for us.

 

So if it's not the OC, and you are claiming it's not the extra VRAM (because you are saying there is no bottleneck because you are taking your graphs as definitive data).

 

Then whats the answer? Was it magic? A primitive voodoo shtick? What could it be?

[I/O]

I also proved to you that it wasn't the OC, you actually proved yourself that it wasn't the OC by providing graphs for us.

 

So if it's not the OC, and it's not the extra VRAM.

 

Then whats the answer? Was it magic? A primitive voodoo shtick? What could it be?

Doesn't matter what it was it could have been a driver, memory speed but the point was made that extra Vram was not the cause of the few extra FPS. I would stick to more reputable sites in the future like anandtech IMHO.

[BiG StroOnZ]

Doesn't matter what it was it could have been a driver, memory speed but the point was made that extra Vram was not the cause of the few extra FPS. I would stick to more reputable sites in the future like anandtech IMHO.

 

Well we know it's not the Driver and we know it's not the Memory:

 

 

Same drivers, same memory speed. 

 

Firstly it wasn't just a "a few extra fps" 

 

In Far Cry 3, difference of 6fps. In Tomb Raider a difference of 9 fps. In Crysis 3 difference of 7 fps. In Battlefield 4 a difference of 11 fps. 

 

Lastly in Watch Dogs it was a difference of 3 fps, and they were allowed to play at higher settings than the 3GB 780. So not only did it get more fps, it got more fps at higher settings. 

 

Hard OCP is as reputable as it gets. 

 

It does matter what the cause of the additional fps was because that's this entire debate. Every time you got proven wrong about what your reason was for the additional fps you changed your tune. Here you are changing your tune again. Now the site isn't reputable. The only conclusion that we have is the one that they gave us, which is because of the additional memory. Which only means one reason, 3GB of VRAM is a bottleneck in the scenarios they tested. It's not that difficult understand. But you want to go off another website's benchmarks that aren't even valid because they are not even the same test scenarios. 

[I/O]

Every time you got proven wrong about what your reason was for the additional fps you changed your tune.

I proved my standing and it still stands that the increases in Vram did no account for the increase in framrate because it did not max out the Vram buffer on the 3GB card. I WIN ! YES VICTORY IS MINE ALL MINE MUHAHA !

[BiG StroOnZ]

Your examples of VRAM usage are invalid because they are not the same test scenarios (not the same level as on the Hard OCP review). The review you used said VRAM usage varies between levels in that game. Also, you only used one game to disprove the benchmark as valid when they tested multiple games. Sorry you did not win you are not excused. They tested other games where the results were the same, 6-11 fps average increase on all games tested (and on Watch Dogs being able to play at higher settings in addition to having higher fps). Meanwhile, you are only going by one benchmark from another reviewer (who tested VRAM usage on an unknown map) for one singular game, that being Crysis 3. Doesn't explain the results on the other games and still doesn't disprove their results on Crysis 3 being that on your benchmark we don't know what map they played on and the reviewer himself said different maps result in different VRAM usage.

 

All we know is that the card with the extra VRAM performed better, and that the clocks between the Strix and the regular 780 aren't that much of a difference for the extra performance to be directly related to the clocks. It usually takes an overclock twice that amount to achieve the same results. Is it possible that the extra VRAM allows for smoother loading of textures? Resulting in a better experience?

 

Taken from the review:

 

 

 

In Watch Dogs the new ASUS STRIX GTX 780 OC 6GB offered a better gameplay experience at 1600p and 1080p. Thanks to the added memory we were able to run at "Ultra" textures all the way up to 1600p without issue. This made the game run smoother indoors and especially outdoors as you drive through the city with fast loading of textures. With the standard GTX 780 we had to use the lower setting of "High" textures which were noticeably inferior.

 

 

 

 

In other places when we cranked up the AA setting to 8X MSAA, even at 1080p, the ASUS STRIX GTX 780 OC 6GB proved to be smoother and better compared to the standard GeForce GTX 780.

[Wherestheanykey]

I would get the 6gb version and oc it past the ti version,the extra vram will help with 4k if you decide to get it in the future.