MSI GTX 1080 Gaming X Plus 11 Gbps 8 GB

[jabbadap]

Guru 3d did just that with an older MSI 1080 gaming x . It was about 5% increase at 4k for most games.

Does they have same max non-advertised boost clocks though(Well fast check from tpus reviews, gives 1974 MHz for plus and 1962 MHz non plus, so not that much different).

I am more interested in the GTX 1060. The 8 gb/s version saw MONSTER https://www.techpowerup.com/reviews/MSI/GTX_1060_Gaming_X/27.html performance gains when the memory was pushed to 9.5 gb/s. If the new version can push the memory to 10 gb/s and beyond, it will leave the RX 580 behind. Nvidia always had GPUs that were 1 or 2 levels above what AMD offers, but now it might be 5: Titan XP, 1080ti, 1080, 1070, 1060 9 GB/s, RX580. (Titan X seems obsolete after the release of the 1080ti, so I don't really count that)

Well Titan Xp is still the fastest of them all, only available as reference form nvidia themselves though and extreme price so more prosumer card than consumer. But if money is not the problem and one don't really care about warranties, liquid cooled Titan Xp:s are the ones to beat.

 

[Shatun_Bear]

Two different worlds really. Take a look at the OC Sapphire RX 580 Nitro+ review and you will see that it still uses a little more watts than this card for a lot less performance. Also, that Nitro+ was a $255 card. People buying in on a budget are going to take power draw into consideration with the Nitro+. People being able to afford $570 on a card......not as much.

In what world? That doesn't make any sense.

 

[Rosie Scenario]

I am more interested in the GTX 1060. The 8 gb/s version saw MONSTER https://www.techpowerup.com/reviews/MSI/GTX_1060_Gaming_X/27.html performance gains when the memory was pushed to 9.5 gb/s. If the new version can push the memory to 10 gb/s and beyond, it will leave the RX 580 behind. Nvidia always had GPUs that were 1 or 2 levels above what AMD offers, but now it might be 5: Titan XP, 1080ti, 1080, 1070, 1060 9 GB/s, RX580. (Titan X seems obsolete after the release of the 1080ti, so I don't really count that)

1060 here. Not a question of "defeat"... If ATI come out with a good one they have my cash. I can bench @10gbs now (not that I require it.) I love fastsync more than 200fps.

 

[intelzen]

you thought there would be real performance increases ?

then what was the point of this "upgrade"?

 

[rtwjunkie]

then what was the point of this "upgrade"?

By appealing with a perception of faster for those that haven't upgraded to a 1080 or above, aren't willing to pay for a 1080Ti and didn't think the previous 1080 offered enough increase from what they have.

Really, it's going after the same type of customer that AMD is appealing to with the RX 580.

 

[AtlasRush]

Drivers: NVIDIA: 378.92 WHQL
AMD: Catalyst 16.3.3 Beta

I hope it's a typo...

 

[TheGuruStud]

By appealing with a perception of faster for those that haven't upgraded to a 1080 or above, aren't willing to pay for a 1080Ti and didn't think the previous 1080 offered enough increase from what they have.

Really, it's going after the same type of customer that AMD is appealing to with the RX 580.

Same type or dumber?

 

[rtwjunkie]

Same type or dumber?

I would need to commission a study to get a good answer to that.

 

[EarthDog]

Settle down there.

Lets look at performance increase, perf/ watt DECREASE, and perf/$ decrease at their target resolution from TPU:

1.) Stock RX 480 vs Nitro 580 at 1440p: 10%, 23%, 10%
2.) Stock 1080 vs Gaming+ 1080 at 4k: 12%, 17%, 5%

Both have WORSE perf./watt and per/$, but the GTX 1080 less so. Also, it gets love for not being called the GTX 1180

The last part in bold there, spot on... thats the difference, they dont call it a different card when its not.

 

[TheGuruStud]

The last part in bold there, spot on... thats the difference, they dont call it a different card when its not.

Wouldn't that be a first in history?

 

[Nihilus]

Wouldn't that be a first in history?

GTX 460 was different AND was not re branded. Better in every way than the GTX 465 but wasn't called the GTX 560 SE. Similar deal with the RX 285. Also, AMD never rebranded the "Ghz edition" into a new lineup. Not sure why they are doing it now so much.

 

[Kanan]

Too bad you didn't include the MSI Gaming Non-Plus or at least another custom 1080 so we could compare directly. Honestly, nobody cares about reference Nvidia speeds compared to custom ones. Custom vs. custom is the main thing.

That's criticism on high niveau though - it's the one thing missing on TPU reviews. Custom vs. custom.

 

[W1zzard]

Drivers: NVIDIA: 378.92 WHQL
AMD: Catalyst 16.3.3 Beta

I hope it's a typo...

Yup, typo, forgot to update the year. Fixed in all affected reviews.

 

[intelzen]

Too bad you didn't include the MSI Gaming Non-Plus or at least another custom 1080 so we could compare directly. Honestly, nobody cares about reference Nvidia speeds compared to custom ones. Custom vs. custom is the main thing.

That's criticism on high niveau though - it's the one thing missing on TPU reviews. Custom vs. custom.

this... I understand why Wizz does not include custom cards (because he has reviewed buttloads of those - too much data)... but in this test "Msi ....Gaming X" vs "Msi... Gaming X Plus 11 Gbps" would be crucial, because 99% people will not look it up and conclude something like this: "damn, that new 11Gpbs gives a nice +10% overall performance increase" - and that is pure wrong (well that is what nvidia would like us to think)

 

[64K]

this... I understand why Wizz does not include custom cards (because he has reviewed buttloads of those - too much data)... but in this test "Msi ....Gaming X" vs "Msi... Gaming X Plus 11 Gbps" would be crucial, because 99% people will not look it up and conclude something like this: "damn, that new 11Gpbs gives a nice +10% overall performance increase" - and that is pure wrong (well that is what nvidia would like us to think)

It really is relevant when reviewing a non-reference 1080 with faster VRAM than the previous card.

I would also ask that when a review is made and the GPU is obviously marketed against the competitor, such as the RX 580 Nitro+ and the GTX 1060 6 GB, then one additional benchmark slot be added with non-reference results for comparison between the two for best bang for the buck.

 

[Shatun_Bear]

It really is relevant when reviewing a non-reference 1080 with faster VRAM than the previous card.

I would also ask that when a review is made and the GPU is obviously marketed against the competitor, such as the RX 580 Nitro+ and the GTX 1060 6 GB, then one additional benchmark slot be added with non-reference results for comparison between the two for best bang for the buck.

Yeah I notice in the 580 review there are the 480 benches there. I expected to see the old 1060 in this review.

 

[rtwjunkie]

Yeah I notice in the 580 review there are the 480 benches there. I expected to see the old 1060 in this review.

Thing is, the vast majority of GPU's that W1zzard tests he doesn't keep. In any given cycle there could be 2 to 3 versions of each model, including reference models. And since he retests everything each time he does a review, you can see how utterly exhausting it would be to include more than the reference models from each prior GPU.

 

[Kanan]

Thing is, the vast majority of GPU's that W1zzard tests he doesn't keep. In any given cycle there could be 2 to 3 versions of each model, including reference models. And since he retests everything each time he does a review, you can see how utterly exhausting it would be to include more than the reference models from each prior GPU.

I know bud, but he could simply enter the data he already has, no need to rebench the same card again. Unless it's different drivers and they have a real impact that is. But if the difference is negligible he can use the "old" data again.

 

[W1zzard]

I know bud, but he could simply enter the data he already has, no need to rebench the same card again. Unless it's different drivers and they have a real impact that is. But if the difference is negligible he can use the "old" data again.

It is different drivers and different games: https://www.techpowerup.com/reviews/MSI/GTX_1080_Gaming_X/

Using relative performance against a baseline card should be close enough though, or pick a game and look at relative differences

 

[Kanan]

It is different drivers and different games: https://www.techpowerup.com/reviews/MSI/GTX_1080_Gaming_X/

Using relative performance against a baseline card should be close enough though, or pick a game and look at relative differences

I'd simply view your old review and then compare custom vs custom - essentially what I say is to include it, so I don't have to search up that review and compare myself, it's simply for comfort purposes (for me) in the end. But if it's against your work schedule or so nvm. I would simply include it as far as possible, leaving aside the games it wasn't tested for in the newer review so that at least some of it is in the new review to compare side by side. In the end PCGH do a pretty good job on that but I guess they have more workforce as well, so I'm not exactly expecting you to do that, rather as far as it's possible to do for you.

/my 2 cents

 

[Vayra86]

Strangest thing is the difference between those two reviews though, have you looked at GTA V?

 

[W1zzard]

Strangest thing is the difference between those two reviews though, have you looked at GTA V?

Yeah different test scene, different test duration, possibly different settings too.

 

[Vayra86]

Yeah different test scene, different test duration, possibly different settings too.

Yeah if that is the case we definitely need to just use the data from the ref 1080 in the most recent review and go from there to compare.

 

[W1zzard]

980 ti vs 1080 stock: 33.6 vs. 32.4%, so really the same