• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Thief Video Card Performance Preview @ [H]

FrgMstr

Just Plain Mean
Staff member
2FA
Joined
May 18, 1997
Messages
58,306
Thief Video Card Performance Preview - Thief, being the first game to support AMD's TruAudio, as well as Mantle API technology is here. Out of the gate the two compelling technologies of the game are missing. Therefore we will take a look today at how the game performs under D3D between AMD and NVIDIA so you will know what to expect out of the game in this Preview.
 
I figured Thief would have Mantle/TruAudio support at launch. It's a little disappointing to have to wait for it, as the technology matures I'm sure this won't always be the case.
 
I'm looking forward to the 4K scores

Fixed! Thanks. Kyle
 
Last edited by a moderator:
Thanks for the preview.

To give an indication of whether Mantle can help, it would have been nice to see a CPU use graph.
If any core is saturating under DX, Mantle should help.
 
I can't see how Mantle would help in this type of game , there not a huge amount of AI or object count as its like a SP mode with limited FOV , LOD .
 
There are 3 camps that could benefit from mantle.
Those with uneven framerate in XFire, those who game at 120Hz and those with the lowest end gfx cards.
 
I ended up grabbing this yesterday. Disregarding what you think of the game, one thing I did notice is that it pushes my GPU usage through the roof. It's at 90-100% almost all of the time. My video card fan is wide open. That's pretty unusual.
 
There are 3 camps that could benefit from mantle.
Those with uneven framerate in XFire, those who game at 120Hz and those with the lowest end gfx cards.
Also those who have multiple graphics cards.

Basically any situation where the CPU is a limitation will benefit.
 
I mentioned Crossfire :), but yeah, any situ where the CPU is being thrashed.
For this game it would more commonly be a slower CPU with dual gfx cards.
 
Are you guys reviewing only AMD sponsored titles these recent times or what? No offense but you ignored Call Of Duty Ghosts, Assassin's Creed 4, Batman Origins and Need For Speed Rivals. All very important and famous titles, they are advanced in their graphics and technology too! Heck, you paid attention to the lowly Splinter Cell Black List. I would hate for my favorite tech site to lag behind like that.
 
"We did also try enabling SSAA, and found the "Low" setting to be playable on all four video cards at 2560x1600"

I mean really? testing GTX780 ti and 290x only with FXAA? that's not [H] at all, who needs 120 fps, you had to include the SSAA instead of FXAA when comparing 780 Ti vs 290X.
 
Are you guys reviewing only AMD sponsored titles these recent times or what? No offense but you ignored Call Of Duty Ghosts, Assassin's Creed 4, Batman Origins and Need For Speed Rivals. All very important and famous titles, they are advanced in their graphics and technology too! Heck, you paid attention to the lowly Splinter Cell Black List. I would hate for my favorite tech site to lag behind like that.
The lack of an Assassin's Creed IV review is especially puzzling. AC4 uses a lot of modern technologies and can also bring graphic cards to their knees.
"We did also try enabling SSAA, and found the "Low" setting to be playable on all four video cards at 2560x1600"

I mean really? testing GTX780 ti and 290x only with FXAA? that's not [H] at all, who needs 120 fps, you had to include the SSAA instead of FXAA when comparing 780 Ti vs 290X.
In all fairness, it does say that it's a "Preview." I'm sure they just want to give people interested in this game a taste of the graphics and performance of this game while the full review is still being written/edited.
 
Would have liked to see the results at 1080p instead of a "these cards ran it fine", how well did they run it? I personally have a R9 270x (well two of them) and a 24" 1080p monitor. All of my gaming, and most of everyone's gaming, is done at that res. I get that we shouldn't have to worry about it, but can we see some results please? If you were trying to simply be the first out with a review, it's a little lackluster. I have been on this site since inception and have always come here first for reviews, if I have to wait a little longer for some more data then I'm ok with that.
 
Are you guys reviewing only AMD sponsored titles these recent times or what? No offense but you ignored Call Of Duty Ghosts, Assassin's Creed 4, Batman Origins and Need For Speed Rivals. All very important and famous titles, they are advanced in their graphics and technology too! Heck, you paid attention to the lowly Splinter Cell Black List. I would hate for my favorite tech site to lag behind like that.

We assessed each game and determined it wasn't worth our time. I am only glad I was able to get my money back on NFS Rivals.
 
"We did also try enabling SSAA, and found the "Low" setting to be playable on all four video cards at 2560x1600"

I mean really? testing GTX780 ti and 290x only with FXAA? that's not [H] at all, who needs 120 fps, you had to include the SSAA instead of FXAA when comparing 780 Ti vs 290X.

This was an ap2ap preview test, we generally stick to one setting so you can compare all the cards in the evaluation. We chose the highest possible in-game settings with FXAA. This, by coincidence, is also the same setting we find playable in Tomb Raider, granted with TressFX.

When we preform highest playable testing, that section will show what was playable. In fact, I went above and beyond preview testing by finding out what WAS playable on all 4 cards and going and posting that in the review with the sentence you just quoted. When we do our full evaluation, that will be included, but we see no reason to do a full evaluation until TruAudio and Mantle are in the game.
 
We assessed each game and determined it wasn't worth our time. I am only glad I was able to get my money back on NFS Rivals.
You found out Assassin's Creed 4 not worth your time? I mean I can relate with your assessment regarding Batman and NFS, but AC4? that's a little strange don't you think?

What about Ghosts then? an incredibly popular and graphically intense game? you pick Thief above all of these triple A titles? is Thief worth more of your time than AC4 and Ghosts?
 
Are you guys reviewing only AMD sponsored titles these recent times or what? No offense but you ignored Call Of Duty Ghosts, Assassin's Creed 4, Batman Origins and Need For Speed Rivals. All very important and famous titles, they are advanced in their graphics and technology too! Heck, you paid attention to the lowly Splinter Cell Black List. I would hate for my favorite tech site to lag behind like that.
Aside from AC4, none of those games are really pushing the technological boundaries any. And I heard Rivals was locked to 30fps, so what's the point in testing that game?
 
Would have liked to see the results at 1080p instead of a "these cards ran it fine", how well did they run it? I personally have a R9 270x (well two of them) and a 24" 1080p monitor. All of my gaming, and most of everyone's gaming, is done at that res. I get that we shouldn't have to worry about it, but can we see some results please? If you were trying to simply be the first out with a review, it's a little lackluster. I have been on this site since inception and have always come here first for reviews, if I have to wait a little longer for some more data then I'm ok with that.

This Preview follows the format of every single game performance preview we've done in the past. I spent all day, and night, squeezing in every card I could for comparison. As much was put in this preview as I could muster in a day. All Performance Previews have followed this day 1 coverage format.
 
Aside from AC4, none of those games are really pushing the technological boundaries any. And I heard Rivals was locked to 30fps, so what's the point in testing that game?
Ghosts had Tessellation and Global Illumination, Direct Compute Fur, even PhysX. the game is really heavy and can't be maxed out (1080p @60fps) even on a Titan class GPU.
 
You found out Assassin's Creed 4 not worth your time? I mean I can relate with your assessment regarding Batman and NFS, but AC4? that's a little strange don't you think?

What about Ghosts then? an incredibly popular and graphically intense game? you pick Thief above all of these triple A titles? is Thief worth more of your time than AC4 and Ghosts?

The last AC game we evaluated was AC3 in 2012. It was pretty much a waste of time even then, and yes, is still with Black Flag.

Ghosts was also a turd, and not worth our time. Looking back in my emails I informed Kyle about NFS Rivals and COD Ghost on 11/19 both games were not worth it. We had more important things to do.

Believe me, I want a game we can add to our gaming suite, or replace an old game with, but none of those games listed are it, sorry.
 
Ghosts had Tessellation and Global Illumination, Direct Compute Fur, even PhysX. the game is really heavy and can't be maxed out (1080p @60fps) even on a Titan class GPU.

Past Call of Duty games have all been the same for us, and pretty much equal a waste of time, Ghost was no exception. The game is also very CPU dependent, rather than GPU dependent, it scales better with CPU, rather than GPU, and this is not good for a GPU dependent test. We passed on it, and aren't looking back.
 
It's OK, I just wanted HOCP to test all triple A titles out there, especially the demanding ones .. but If you guys are truly busy who can blame you?

Good luck with your future benchmarks, I hope you guys are ready for Watch Dogs and Witcher 3! :)
 
I mentioned Crossfire :), but yeah, any situ where the CPU is being thrashed.
For this game it would more commonly be a slower CPU with dual gfx cards.

Also those who have multiple graphics cards.

Basically any situation where the CPU is a limitation will benefit.

But do you really need CrossFire in a game like this that is already performing so well? How much performance is "enough" when you can play at the maximum settings already?
 
I can't see how Mantle would help in this type of game , there not a huge amount of AI or object count as its like a SP mode with limited FOV , LOD .

Exactly, that's why I'm dubious, right now, how it can help further when I can already play the game maxed out on $150 video cards.
 
But do you really need CrossFire in a game like this that is already performing so well? How much performance is "enough" when you can play at the maximum settings already?
Nope, you don't, I was just adding on to the comment.

Frankly I don't care about Mantle in this game, but I would love to see TrueAudio in action as that was one of the reasons I opted to upgrade to 290Xs.
 
This was an ap2ap preview test, we generally stick to one setting so you can compare all the cards in the evaluation. We chose the highest possible in-game settings with FXAA. This, by coincidence, is also the same setting we find playable in Tomb Raider, granted with TressFX.

When we preform highest playable testing, that section will show what was playable. In fact, I went above and beyond preview testing by finding out what WAS playable on all 4 cards and going and posting that in the review with the sentence you just quoted. When we do our full evaluation, that will be included, but we see no reason to do a full evaluation until TruAudio and Mantle are in the game.

Thanks, sorry if my post was kind off agressive :eek:
 
Ghosts had Tessellation and Global Illumination, Direct Compute Fur, even PhysX. the game is really heavy and can't be maxed out (1080p @60fps) even on a Titan class GPU.

Honestly, the game still looks like it has textures straight from the year 2006. The game does not have impressive graphics, although MP is semi mindless fun at times. Keep in mind COD: ghosts uses a modified Quake 3 engine. Yes, you heard correct. Modified quake 3 engine. And the textures are AWFUL up close despite the tessellation and all of that shit. Tessellation doesn't magically transform awful low resolution textures into something they aren't - the assets in COD: Ghosts just aren't up to par with modern AAA titles.

So I completely agree with Brent in this respect. It's just not a game worth benchmarking IMO. That said, i've had a bit of stupid fun in MP. But it certainly is not an amazing game, and the graphics are really not good - particularly the textures.
 
"We did also try enabling SSAA, and found the "Low" setting to be playable on all four video cards at 2560x1600"

I mean really? testing GTX780 ti and 290x only with FXAA? that's not [H] at all, who needs 120 fps, you had to include the SSAA instead of FXAA when comparing 780 Ti vs 290X.

To add to my earlier statement: When you enable SSAA, at that point, you are pretty much testing only the performance of SSAA, because it is the most demanding thing in the game, and takes the focus away from the game's graphics settings themselves. Before you do that, it is important to know how the graphics in the game burden the GPU. Do the hard contact shadows cause a GPU performance burden, does the parrallax occlusion mapping cause a burden, does tessellation decrease performance. It is important to know how much the graphics options themselves impact the game performance first. In this game we found out they do not, and performance is very high even with everything enabled. That is very important information.
 
Exactly, that's why I'm dubious, right now, how it can help further when I can already play the game maxed out on $150 video cards.

Yes, Where this type of low level driver would help out a lot is in a game/sim of a flight/helo combat Sim .
Not only would it help in not having to instance objects to reduce calls because of longer rendering distances but also the damage that can be kept . something instancing is not great at .
A updated Falcon4 would be really good example of game/sim but to bad there not popular and tend not to be to graphical because of need for long viewing distances .
 
But do you really need CrossFire in a game like this that is already performing so well? How much performance is "enough" when you can play at the maximum settings already?
Its not a matter of need, its what people have in their machines.
For those running slower cards in crossfire, it should give a boost, especially if they have a slower CPU as well.
It has been shown that even if GPU limited there can still be a performance increase.
 
Its not a matter of need, its what people have in their machines.
For those running slower cards in crossfire, it should give a boost, especially if they have a slower CPU as well.
It has been shown that even if GPU limited there can still be a performance increase.

Read this again:

Brent_Justice said:
how it can help further when I can already play the game maxed out on $150 video cards.

So far, Mantle seems aimed at making AMD's low-end CPU products not suck so bad. HardOCP doesn't test low-end stuff. HardOCP doesn't test old stuff.
 
Honestly, the game still looks like it has textures straight from the year 2006. The game does not have impressive graphics, although MP is semi mindless fun at times. Keep in mind COD: ghosts uses a modified Quake 3 engine. Yes, you heard correct. Modified quake 3 engine. And the textures are AWFUL up close despite the tessellation and all of that shit. Tessellation doesn't magically transform awful low resolution textures into something they aren't - the assets in COD: Ghosts just aren't up to par with modern AAA titles.

So I completely agree with Brent in this respect. It's just not a game worth benchmarking IMO. That said, i've had a bit of stupid fun in MP. But it certainly is not an amazing game, and the graphics are really not good - particularly the textures.

I played the multiplayer trial on Steam this weekend. Colossal waste of time and bandwidth. Graphics weren't all that great, and while the framerate was good on my system, I still had issues with sound stutters/pops, and stutter in general. Even in the menus, it stuttered. So glad I got to try this game. Not sure what happened... Black Ops 2 ran just fine on my system...
 
Read this again:

So far, Mantle seems aimed at making AMD's low-end CPU products not suck so bad. HardOCP doesn't test low-end stuff. HardOCP doesn't test old stuff.

You quoted me to make a statment against, then brought up exactly the reason.
People with dual cards and slow CPUs will suffer more due to the Xfire overhead.

They dont need to test old stuff, giving a % metric of approximate improvement when CPU limited in D3D would help people.
Very specific testing isnt going to happen, granted.
 
Awesome work as always, but I hope you guys didn't put too much work into this. Thief looks to be an epic flop and none of the enhancement features work, so I don't think you will want to be using this game in your benchmarks. They haven't even managed to get normal audio working, so it would be a stretch to use this to test audio enhancements. It would be like using Diakatana in a benchmark.
 
Are you guys reviewing only AMD sponsored titles these recent times or what? No offense but you ignored Call Of Duty Ghosts, Assassin's Creed 4, Batman Origins and Need For Speed Rivals. All very important and famous titles, they are advanced in their graphics and technology too! Heck, you paid attention to the lowly Splinter Cell Black List. I would hate for my favorite tech site to lag behind like that.

Batman should be excluded, WB flat out refused the fix that AMD offered them.

http://www.extremetech.com/extreme/...surps-power-from-developers-end-users-and-amd
 
I played the multiplayer trial on Steam this weekend. Colossal waste of time and bandwidth. Graphics weren't all that great, and while the framerate was good on my system, I still had issues with sound stutters/pops, and stutter in general. Even in the menus, it stuttered. So glad I got to try this game. Not sure what happened... Black Ops 2 ran just fine on my system...

Yeah, it definitely has some performance weirdness. And I understand it has physx and tessellation but honestly tessellation cannot fix textures that are ridiculously low resolution with the Q3 engine. They just need to start over from scratch and get a new engine for that game, or something, Q3 engine (even if modified) in 2014 is just a travesty. And yeah, Black Ops 2 is a better and more fun game. I can appreciate COD for what it is, it's occasional mindless fun, but damn. The graphics are pretty terrible even with the DX11 features. And Ghosts still isn't as good as BO2 was, which was more fun in MP.

As far as thief goes. Graphics are great and all but the overriding problem with the game is that the game is.........NOT the essence of what made Thief a great franchise. This game is everything that is wrong with the gaming industry in current times, catering to the lowest common denominator and removing game mechanics that can be construed as difficult. The prior thief games were excellent. This one? Not feeling it. Reviews have been very mixed as well, it just isn't as good as the prior Thief games, which were VERY good. I can appreciate great graphics, but I appreciate great gameplay even more than that - Thief just isn't the game most long time franchise fans had hoped for. Which is surprising because Nixxes has such a great track record; they dropped the ball on this one IMO.

I love great graphics, but I try to not get TOO caught up in that at times because really - gameplay is what matters. The Thief reboot is just REALLY disappointing. Game is too linear and doesn't have the sense of constant danger that prior games have. Just so not what I was hoping for, and I pre-purchased the game. Never again will I pre-order. Ugh.
 
Last edited:
Back
Top