Win 8.1 preview performance

deasnutz

Limp Gawd
Joined
Feb 9, 2012
Messages
377
Just thought I'd share my experience, Win 8.1 clean install, no additional drivers...it includes 326.01.

I'm getting exceptional performance on 680 SLI, min fps in bf3 jumped 30 frames, much more stable average.

Hyperthreading DISABLED.
 
min fps in bf3 jumped 30 frames

lol sure it did. a thread like this is almost useless if you dont have direct comparisons and benchmarks. just saying your minimums went up by 30 fps sounds pretty silly.
 
Probably wouldn't apply to your rig, just sayin, no hate. I'm sure it's your pride and joy.
lol I am not emotionally attached to my pc. I could build a $5000 pc if I wanted to but just prefer the bang for buck approach for what little I game. plus you could not pay me to deal with any potential multi card issues.

and the point of the sarcastic reply was because you probably did not come remotely close to adding 30 fps to your minimum fps in BF 3 just for going to Windows 8.1. this thread is nothing like your title would indicate as you have no actual results to share.
 
Probably wouldn't apply to your rig, just sayin, no hate. I'm sure it's your pride and joy.
Considering most people on this forum have better "rigs" here than you, I find this comment of yours laughable.
Your inability to write in English language?
LOL.

Something was most likely wrong with the way his pc was configured, for him to see that kind of gain.
 
^^ Yea?...But can windows 8.1 play Crysis?

Which version?

I've had NO problem with the last two (Crysis 2 or Crysis 3) at a resolution of 1920x1080 and Extreme OR Ultra settings (either no AA or lowest AA) - and that's on a bone-stock Q6600 and GTX550 Ti (same drivers the OP referred to - I've updated since then simply to stay up to date, NOT due to any driver issues).

I get away with it because I have a 23" display.
 
Which version?

I've had NO problem with the last two (Crysis 2 or Crysis 3) at a resolution of 1920x1080 and Extreme OR Ultra settings (either no AA or lowest AA) - and that's on a bone-stock Q6600 and GTX550 Ti (same drivers the OP referred to - I've updated since then simply to stay up to date, NOT due to any driver issues).

I get away with it because I have a 23" display.
whoosh

and no you most certainly did not play Crysis 2 and Crysis 3 smoothly on extreme/ultra settings at 1920x1080 with a dinky gtx550 ti.
 
lmao at this thread. i don't know whats funnier the review of windows 8.1 with a bf3 fps increase and "HYPERTHREADING DISABLED", or all the replies dismissing it.
 
Just thought I'd share my experience, Win 8.1 clean install, no additional drivers...it includes 326.01.

I'm getting exceptional performance on 680 SLI, min fps in bf3 jumped 30 frames, much more stable average.

Hyperthreading DISABLED.

LOL this guy is Steve Ballmer. :D
 
When I booted up the Windows 8.1 preview a fairy nyph princess came out of my PC and made love to all day and night...I woke up the next day just fine...was it a dream? The memmory certainly is fading from my mind too quickly to be sure it was real, but, by all that is holy it felt so real! Crysis 3 runs about the same.
 
whoosh

and no you most certainly did not play Crysis 2 and Crysis 3 smoothly on extreme/ultra settings at 1920x1080 with a dinky gtx550 ti.

And why are you so quick to dismiss it?

In the case of Crysis 2, an AMD HD5770 can do it paired with the same CPU, and the HD5770 is admittedly slower than the GTX550Ti.

Further, note that I have either the lowest AA settings (Crysis 3) or no AA at all (Crysis 2). In both cases, it's due to my having a 1920x1080 TN display - therefore, additional AA is outright wasted. If you keep the AA low (or turn it off) you can turn other detail settings up (not down) to compensate - therefore Extreme LOD/minimal AA or Ultra LOD but no AA.

Crysis 3 - this is, admittedly, a VERY odd duck of a shooter in that it performs better with minimum AA than without AA (at least on the Ti) and I have not been able to puzzle out why.

Crysis 2 - are you using non-standard textures (which are generally designed for GPUs with larger amounts of RAM, such as GTX660/HD6850 and other 2GB or larger GPUs)? If so, then comparing workloads to a 1GB GPU using standard textures (such as those included with the HD Texture Pack, included with the Maximum Edition Pack) is NOT apples to apples. I have nothing against those with more powerful (read - larger RAM loadouts) GPUs wanting larger textures, as the standard-sized textures leave such GPUs under-utilized - however, it makes GPU comparison harder, not easier, as it introduces another variable. For those of us with 1GB, the Crytek-supplied HD Texture Pack is just fine.
 
you have to be kidding. you CANT run Crysis 2 on extreme and Crysis 3 on very high settings at 1080 with a gtx550 ti even with no AA and that is FACT. all that gibberish and nonsense you just posted does not change the FACT that a gtx550 ti is not strong enough to run those settings at 1080. NOT EVEN CLOSE.
 
http://www.guru3d.com/articles_pages/msi_geforce_gtx_770_gaming_review,18.html
This link is only using fxaa and very high settings. Every card in the chart is faster than a 550ti and the bottom of the chart averages 17fps. That's not playable and a 550ti is even worse.

And what CPU are the GPUs In Question being tested with? (That is NOT mentioned.)

And you (and the link) DID mention FXAA (which is above both MSAA and no AA). Any AA (including, if not especially FXAA) does add more drag onto a GPU - I not only didn't dispute that, but stipulated it. (It's also why I questioned the oddity of Crysis 3's performance - normally, adding AA should have made performance worse, if not MUCH worse.) I traded AA for greater detail - which makes sense; why is that so hard to fathom? Why would I deliberately bog down gaming by adding AA - especially since the benefit of AA would otherwise be zero? Now, those of you with the heftier-RAM GPUs can add AA - often whether you use non-standard OR standard textures - since your GPUs have power to spare - which the chart also shows. (How does HD7770 fall into that category, though? It has the same amount of RAM as the GTX 550 Ti or GTX 560 Ti and has a narrower memory bus besides - 128-bit for HD7770 vs. 192-bit for 550Ti.) If anything, I would recommend that anyone with 1GB of GPU RAM dispense with AA in Crysis 2 altogether - because it is a drag/bog on performance, and increase the LOD instead. Any chart of apples-to-apples performance for Crysis 2 shows that AA is a decided drag on performance - yes, that even includes Kyle's apples-to-apples comparisons. On top of that, twenty-three inches is comparatively small for a display; further, of what benefit is AA on a TN-based display (which mine is)? I ditched AA in Crysis 2 because it made sense - if I could have in Crysis 3, I WOULD have.
 
And what CPU are the GPUs In Question being tested with? (That is NOT mentioned.)

And you (and the link) DID mention FXAA (which is above both MSAA and no AA). Any AA (including, if not especially FXAA) does add more drag onto a GPU - I not only didn't dispute that, but stipulated it. (It's also why I questioned the oddity of Crysis 3's performance - normally, adding AA should have made performance worse, if not MUCH worse.) I traded AA for greater detail - which makes sense; why is that so hard to fathom? Why would I deliberately bog down gaming by adding AA - especially since the benefit of AA would otherwise be zero? Now, those of you with the heftier-RAM GPUs can add AA - often whether you use non-standard OR standard textures - since your GPUs have power to spare - which the chart also shows. (How does HD7770 fall into that category, though? It has the same amount of RAM as the GTX 550 Ti or GTX 560 Ti and has a narrower memory bus besides - 128-bit for HD7770 vs. 192-bit for 550Ti.) If anything, I would recommend that anyone with 1GB of GPU RAM dispense with AA in Crysis 2 altogether - because it is a drag/bog on performance, and increase the LOD instead. Any chart of apples-to-apples performance for Crysis 2 shows that AA is a decided drag on performance - yes, that even includes Kyle's apples-to-apples comparisons. On top of that, twenty-three inches is comparatively small for a display; further, of what benefit is AA on a TN-based display (which mine is)? I ditched AA in Crysis 2 because it made sense - if I could have in Crysis 3, I WOULD have.


I hear what you're saying personally and I'm not suggesting you're lying, but this is just one of those things people will need proof or GTFO kind of responses.

Also there is "playable" settings and "able to play" settings. A game running at 1 FPS is playable, but you're not able to play it realistically. The definition of playable seems to be lost in this entire argument. What kind of Min, Avg, Max FPS are we talking about on the Q6600/550 combo being discussed. One mans FPS may be more tolerable than another mans. Just because people say 30FPS min and 60FPS recommended doesn't mean some people don't play games at less than 30FPS. I think we've all done it at some point because we had to make due.
 
Probably wouldn't apply to your rig, just sayin, no hate. I'm sure it's your pride and joy.

Hey, don't be a dick towards people. Especially that "I'm sure it's your pride and joy" remark. Not everyone needs two 680s in SLI. I know for a fact that I don't and that I will be perfectly fine with my two 760s that will arrive soon - they do everything that I need them to.

The guy has a 660Ti, so what? I was going to buy a 660Ti before the 7xx series came out. It gets 60FPS minimum on a friend's rig on ultra settings in Battlefield 3. And seeing as the vast majority of people use 60Hz monitor.... there's no point in spending 800$ on two 680s if you just want to play BF3 occasionally, now is there?

Also run a proper analysis with benchmarks or it's BS. Nobody I know running 8.1 has 30fps increases just for using 8.1, maybe 1-3fps and I'd believe it. Maybe you had some program that was hogging all of the graphics memory or downclocking them to save power that got wiped out with the clean install? Your drivers could have been FUBAR'd before 8.1 by some glitch or maybe a messed up beta?
 
And what CPU are the GPUs In Question being tested with? (That is NOT mentioned.)

And you (and the link) DID mention FXAA (which is above both MSAA and no AA). Any AA (including, if not especially FXAA) does add more drag onto a GPU - I not only didn't dispute that, but stipulated it. (It's also why I questioned the oddity of Crysis 3's performance - normally, adding AA should have made performance worse, if not MUCH worse.) I traded AA for greater detail - which makes sense; why is that so hard to fathom? Why would I deliberately bog down gaming by adding AA - especially since the benefit of AA would otherwise be zero? Now, those of you with the heftier-RAM GPUs can add AA - often whether you use non-standard OR standard textures - since your GPUs have power to spare - which the chart also shows. (How does HD7770 fall into that category, though? It has the same amount of RAM as the GTX 550 Ti or GTX 560 Ti and has a narrower memory bus besides - 128-bit for HD7770 vs. 192-bit for 550Ti.) If anything, I would recommend that anyone with 1GB of GPU RAM dispense with AA in Crysis 2 altogether - because it is a drag/bog on performance, and increase the LOD instead. Any chart of apples-to-apples performance for Crysis 2 shows that AA is a decided drag on performance - yes, that even includes Kyle's apples-to-apples comparisons. On top of that, twenty-three inches is comparatively small for a display; further, of what benefit is AA on a TN-based display (which mine is)? I ditched AA in Crysis 2 because it made sense - if I could have in Crysis 3, I WOULD have.

What does the size of your monitor have to do with anything? 1920x1080 is available on everything from 17" to 27" and all would perform identically with regard to FPS. Also I'd think AA would be more important on a TN display since they tend to have faster response times resulting in less pixel blurring that would make any aliasing more noticeable.
 
I hear what you're saying personally and I'm not suggesting you're lying, but this is just one of those things people will need proof or GTFO kind of responses.

Also there is "playable" settings and "able to play" settings. A game running at 1 FPS is playable, but you're not able to play it realistically. The definition of playable seems to be lost in this entire argument. What kind of Min, Avg, Max FPS are we talking about on the Q6600/550 combo being discussed. One mans FPS may be more tolerable than another mans. Just because people say 30FPS min and 60FPS recommended doesn't mean some people don't play games at less than 30FPS. I think we've all done it at some point because we had to make due.

That's my point exactly, Liger. Usually I can't consider a game even remotely playable if framerates are 15 fps or less - it's also a floor I've made no secret of having. However, I don't have superhuman eyesight where I can tell framerates above 30 fps, and especially not on a bargain/value TN display. Some folks push the detail settings higher because they can - I don't have an issue with that. However, making the assumption that pushing the envelope is the ONLY way to go in that area is being insulting and (worse) sounds like the sort of e-peen wavery that starts flamewars - I don't, and won't, get into that.
 
What does the size of your monitor have to do with anything? 1920x1080 is available on everything from 17" to 27" and all would perform identically with regard to FPS. Also I'd think AA would be more important on a TN display since they tend to have faster response times resulting in less pixel blurring that would make any aliasing more noticeable.

Because adding AA is certainly a drag on GPU performance - any GPU test from anywhere makes that plain. AA is all about increased detail - that I agree with. However, adding AA with a 1 GB GPU, or any GPU for that matter, typically also involves reducing settings elsewhere in-game and having the increased AA compensate for that settings reduction. (Only higher-end GPUs can combine max detail with AA at any level - again, not one GPU test worth anything says otherwise, and I don't, either.) Instead of trading detail for AA, I went the opposite direction - I kept both detail and resolution high, but turned AA off.

1920x1080 is critical to me for three reasons - it's my monitor's display ceiling (it can go no taller), it's also my desktop resolution, and it's my preferred GAMING resolution. (In the case of flat-panels, while they are still, by and large multi-resolution capable, there is still a preferred resolution. Gamers should have a GPU that can keep up with their display's preferred resolution - that is something I have said myself - repeatedly. For the first time in years, I'm not held back from gaming at my display's preferred resolution in most games, if not all games.) If I had a better (in terms of resolution, screen size, et. al.) display, i would doubtless want a better GPU (something in the GTX660 or above area) - that makes sense. With a better GPU (again, GTX660 is the example) I easily COULD max the settings AND add AA without taking too many lumps performance-wise. (I didn't dispute that, either.) However, I simply questioned the value of adding AA at the expense of adding detail - I was able to add increased detail settings without taking a major hit in terms of frame rates in Crysis 2, which I could not do by adding AA, even at the expense of lowering detail settings. (Basically, AA + lower detail hurt worse than higher detail but no AA.) That is, in fact, pretty standard for pixel-based game engines, and CryEngine 3 as Crysis 2 implemented it is bog-standard in that regard.

That is also why the alarms went off in my Crysis 3 testing - it didn't follow that pattern. I actually started with the same settings as Crysis 2 - Extreme (not Ultra) LOD, but no AA. Unlike Crysis 2, however, adding 1x SMAA increased (not decreased) framerates (same detail level). However, it's the only game which displays such abnormal behavior. So far, there is nothing that can explain this oddity - not from nVidia or Crytek. The most I can hope for eventually is something coming out of Crytek that explains the performance differences between the CryEngine 3 implementations in both games - in the meantime, I've simply hit a wall.
 
And what CPU are the GPUs In Question being tested with? (That is NOT mentioned.)

And you (and the link) DID mention FXAA (which is above both MSAA and no AA). Any AA (including, if not especially FXAA) does add more drag onto a GPU - I not only didn't dispute that, but stipulated it. (It's also why I questioned the oddity of Crysis 3's performance - normally, adding AA should have made performance worse, if not MUCH worse.) I traded AA for greater detail - which makes sense; why is that so hard to fathom? Why would I deliberately bog down gaming by adding AA - especially since the benefit of AA would otherwise be zero? Now, those of you with the heftier-RAM GPUs can add AA - often whether you use non-standard OR standard textures - since your GPUs have power to spare - which the chart also shows. (How does HD7770 fall into that category, though? It has the same amount of RAM as the GTX 550 Ti or GTX 560 Ti and has a narrower memory bus besides - 128-bit for HD7770 vs. 192-bit for 550Ti.) If anything, I would recommend that anyone with 1GB of GPU RAM dispense with AA in Crysis 2 altogether - because it is a drag/bog on performance, and increase the LOD instead. Any chart of apples-to-apples performance for Crysis 2 shows that AA is a decided drag on performance - yes, that even includes Kyle's apples-to-apples comparisons. On top of that, twenty-three inches is comparatively small for a display; further, of what benefit is AA on a TN-based display (which mine is)? I ditched AA in Crysis 2 because it made sense - if I could have in Crysis 3, I WOULD have.

Fxaa is the lightest form of aa in c3. Has almost 0 impact on performance. Msaa is much much worse in c3. Huge performance hit.
 
And what CPU are the GPUs In Question being tested with? (That is NOT mentioned.)

And you (and the link) DID mention FXAA (which is above both MSAA and no AA). Any AA (including, if not especially FXAA) does add more drag onto a GPU - I not only didn't dispute that, but stipulated it. (It's also why I questioned the oddity of Crysis 3's performance - normally, adding AA should have made performance worse, if not MUCH worse.) I traded AA for greater detail - which makes sense; why is that so hard to fathom? Why would I deliberately bog down gaming by adding AA - especially since the benefit of AA would otherwise be zero? Now, those of you with the heftier-RAM GPUs can add AA - often whether you use non-standard OR standard textures - since your GPUs have power to spare - which the chart also shows. (How does HD7770 fall into that category, though? It has the same amount of RAM as the GTX 550 Ti or GTX 560 Ti and has a narrower memory bus besides - 128-bit for HD7770 vs. 192-bit for 550Ti.) If anything, I would recommend that anyone with 1GB of GPU RAM dispense with AA in Crysis 2 altogether - because it is a drag/bog on performance, and increase the LOD instead. Any chart of apples-to-apples performance for Crysis 2 shows that AA is a decided drag on performance - yes, that even includes Kyle's apples-to-apples comparisons. On top of that, twenty-three inches is comparatively small for a display; further, of what benefit is AA on a TN-based display (which mine is)? I ditched AA in Crysis 2 because it made sense - if I could have in Crysis 3, I WOULD have.
look you seem like a nice guy but you really just have no idea what you are talking about for most of what you post here.
 
Guys, I just installed Window 8.1 Preview on my potato. I get minimum 160 fps on Derp City 6: Rollin' Derp Style
 
Back
Top