Do AMD's Radeon HD 7000s Trade Image Quality For Performance?

[PCgamer81]

When the game first loads up, it takes a second for the textures to fall in place. After that, I'm good to go.

If you are having any trouble after that, I suggest you look elsewhere.

There is no such thing as a Hard-Drive bottleneck. It doesn't exist. Because if it did, to be frank, SSDs would also be a bottleneck (even DDR2 is multiple times faster than solid state memory).

Your gross misunderstanding of an SSD's (or HDD's) role in gaming can be better rectified by understanding the nature of RAM and the Drive when it comes to the game.

When the game is loading (and maybe a second or two after it loads) all the data that is needed is pulled from the Hard Drive and cached into the RAM. Any data used and processed from then on out will be contingent not upon SSD speed, but on RAM size and speed.

Again (for dummies), the RAM provides the cache, and the GPU takes over from there. The only time an SSD would factor in the equation is at the very beginning (when the game loads), or if a RAM bottleneck is present, in which case the system would look to the SSD to do the caching - and your frame rate would crumble because it would be like a librarian having to run across the library pulling data from random bookshelves, as opposed to say pulling data from a table. RAM bottlenecks are nasty, thank goodness those days are over!.

Anyone with even a basic understanding of systems and how they work would understand this.

So, for the last time, it is IMPOSSIBLE for your drive to factor in to actual in-game frame rate. Absolutely impossible.

Good day!

 

[mapesdhs]

[citation][nom]PCgamer81[/nom]... So, for the last time, it is IMPOSSIBLE for your drive to factor in to actual in-game frame rate. Absolutely impossible.Good day![/citation]

Except for those games that continue to load data as one moves around the game world, which
can cause stuttering, an effect that is absolutely lessened by using an SSD. This is definitely the
case with Stalker and Oblivion, very likely true with other modern complex titles, especially if
one is using texture packs and other detail-enhancing mods.

Pointless denying it when there's plenty of evidence to prove otherwise. When I run the Stalker
benchmark from a HDD, there is significant stutter early on. Run off an SSD, there is still a little
but of stutter at the beginning, but much less compared to the HDD. Later in the benchmark,
more stutter occurs when run from a HDD, but none when run from an SSD.

You're correct to say an SSD can't (or shouldn't) enhance max frame frates, but they certainly
can prevent stutter during game play due to data loading.

Very few games load all the data they need at the start of a game and then never load anything
again. Indeed, these days that would be an extremely odd way of coding a game given the close
relationship between versions for the PC and console editions.

Ian.

 

[silverblue]

Not every game precaches. MMOs don't necessarily cache everything in the event it's actually going to appear on-screen, for example. Also, not everybody has 16GB of RAM. Additionally, I have heard of paging and the effects it has on system performance.

Discussion over... gracious of me, considering your tone.

 

[hanskey]

You should all stop trolling each other and read and respond with something other than snark if you want to convince other people that you are correct. Name calling is neither helpful nor necessary to make a convincing case for your viewpoint so please stop.

 

[PCgamer81]

You can't troll with facts.

All I have ever said is that SSDs do not improve in-game performance. Because they do not.

As badly as the sheep want to justify a foolish investment on their part, attempting to rework in their own minds how computers work is to no avail.

And thumbing down my comments do nothing to put wasted money back in their sad pockets.

But at the risk of coming across as a troll, I will humbly bow out.

Last post in this thread by me, any and all replies to me will not be read by these wise eyes.

Good day!

 

[silverblue]

[citation][nom]PCgamer81[/nom]All I have ever said is that SSDs do not improve in-game performance. Because they do not.[/citation]

All I was trying to convey was that if your game is loading a fair amount of data during gameplay, chances are it will detrimentally affect your framerate as the game cannot show what it hasn't loaded; the effect would be amplified if your system lacks sufficient RAM to cope with both the system and game simultaneously. mapesdhs says that having an SSD in this instance would reduce the effect due to the faster data access and transfer rates.

I really don't know what the rest of your post is about, though. Personally, I have no purchase to justify (other than the 1TB Samsung F3 HD103SJ that I bought before prices broke through the ceiling). The only thing that got to me was being spoken down to; it seems as if as long as you don't throw about insults, this would be acceptable. Ah well, if you're not going to take part in this thread anymore then I believe that's your loss, but it is your choice.

With that spec of yours, I'd feel more inclined to keep the refresh rate at 75Hz, enable vsync and forget 120Hz. Sounds mad, but I'd rather reduce the chance of image tearing whilst keeping a buttery smooth framerate, all without compromising visual fidelity.

 

[blazorthon]

[citation][nom]PCgamer81[/nom]You can't troll with facts.All I have ever said is that SSDs do not improve in-game performance. Because they do not.As badly as the sheep want to justify a foolish investment on their part, attempting to rework in their own minds how computers work is to no avail.And thumbing down my comments do nothing to put wasted money back in their sad pockets.But at the risk of coming across as a troll, I will humbly bow out. Last post in this thread by me, any and all replies to me will not be read by these wise eyes.Good day![/citation]

World of Warcraft, on my computers, takes a while to load in-game textures (like it or not, many games load textures during gameplay). With an SSD, this would stop meaning that even once I'm in the game, some things would not render for a few more seconds. Furthermore, some games stutter instead of just not rendering things at once (especially some first-person shooter games), reducing minimum frame rates. An SSD also fixes that. So, once agian, YES, an SSD CAN help in-game frame rates. You, yet again, are making assumptions about something you know nothing about. Tom's did articles PROVING that an SSD CAN improve gameplay. My whole point was that it's not a big difference unless you are multi-tasking (when the hard drive is busy, it gets a lot slower with loading textures and such) and whether you like it or not, you are easily proven wrong about this just as much as your other argument.

Several games (especially first-person shooter games) read and/or write hundreds of MB to several GB of data every time you play a map. Having an SSD for the times when the read/write usage spikes stops the stutter caused by the storage.

Storage can provide a bottleneck. SSDs don't bottleneck it because they are fast enough to end the bottleneck, whereas hard drives often are not. It doesn't take system RAM to solve a fairly minor bottleneck. Just because an HDD can be a bottleneck ,that doesn't mean that it will take DDR2 or better to end it. The SSDs are enough for this.

Your argument is like saying that just because a Radeon 6450 is a bottleneck for 720p with maxed out settings and AA in BF3, that a 6850 is still a bottleneck here just because it's not a 6970.

As badly as you want to say that everyone with more money than you is stupid, you're wrong. This isn't someone else trying to justify stupid purchases, this is YOU trolling people because YOU are ignorant of what higher quality parts can do, or you are simply jealous and refuse to admit that something you either can't afford or can't justify buying is better than what you bought. Your zeal on this despite other people who know more about the subject telling you that you are wrong and then telling you WHY you are wrong is ridiculous. I don't know how you managed to convince yourself that you are wise and know anything about this.

I may not be wealthy enough to buy most of this stuff, but I am fairly well-informed and I do keep up with technology even if I can't afford the best of it. Even if you're not here to read any of this, you're still wrong and at the very least, something should be said so that anyone else who reads this isn't misled by your lies. Also, you can troll with facts, however, you're trolling with lies nonetheless, so that doesn't matter. It's just another thing for you to be wrong about.

 

[PCgamer81]

Most people who get SSDs use them for their OS.

The general consensus is that they don't have enough of a positive effect on gameplay to merit the cost that would be required to install games to the SSD.

You also have to consider that SSDs are still multiple times slower than RAM so if their is any stuttering (most likely due to a RAM bottleneck), SSDs will not remedy it - only improve it, albeit barely. I have a standard 7200 RPM HDD and I do not experience any sub 60 framerate drops whatsoever, in fact not once. Of course, I am not bottlenecked by RAM.

 

[blazorthon]

[citation][nom]PCgamer81[/nom]Most people who get SSDs use them for their OS.The general consensus is that they don't have enough of a positive effect on gameplay to merit the cost that would be required to install games to the SSD.You also have to consider that SSDs are still multiple times slower than RAM so if their is any stuttering (most likely due to a RAM bottleneck), SSDs will not remedy it - only improve it, albeit barely. I have a standard 7200 RPM HDD and I do not experience any sub 60 framerate drops whatsoever, in fact not once. Of course, I am not bottlenecked by RAM.[/citation]

An SSD that can work dozens of times faster for random access speed and several times faster for sequential speed is not just going to barely make a difference. Tom's did an article on this and proved that an SSD completely removed the stutter caused by slow hard drives. You don't seem to understand much about how each component factors in as bottlenecks at all, despite your claiming otherwise.

Such drops are generally not large and the problem is usually more about textures not loading fast enough, rather than frame rates dropping. RAM rarely provides much of a bottleneck because it has little effect on performance. So long as you have 1600MHz memory for Intel and 1866MHz or 2133MHz for AMD, it should never be a problem at all for gaming. Even if you don't, RAM doesn't provide large bottlenecks unless it's 1066MHz or lower. Of course, being single channel would also make the problem a little greater, but that's usually not something to worry about for gamers who almost always have at least two memory modules in their systems.

For games that can have minor problems with hard drives, SSDs remedy the situation completely, unless they are crap SSDs. Whether or not they are as fast as system RAM is irrelevant because they don't need to be that fast.

EDIT: Most people who get a 64GB or smaller SSD are getting it for their OS. Any larger, it is for other things too. Most SSD buyers are not buying 64GB SSDs right now (120GB or 128GB being the most common) because they want their SSD to benefit more than just the OS. Gamers with an SSD generally buy a 120GB or 128GB SSD or larger because that allows them to have some games on the SSD. Not all games benefit from an SSD because not all games are storage-bound, so not all games should be on an SSD, so very large capacities to fit many games aren't necessary. 120GB-128GB are considered the sweet spot.

 

[PCgamer81]

Games installed on an SSD will load faster, save faster, and if there's any in-game caching, will be improved on an SSD.

Other than that...the performance will be the same on an SSD as it would on an HDD, providing your RAM is tip-top. I have 16GB of DDR3, so I don't really have to worry about it.

The belief that an SSD affects any other aspect of gameplay betrays a fundamental ignorance as to how computers work in general.

 

[blazorthon]

[citation][nom]PCgamer81[/nom]Games installed on an SSD will load faster, save faster, and if there's any in-game caching, will be improved on an SSD.Other than that...the performance will be the same on an SSD as it would on an HDD, providing your RAM is tip-top. I have 16GB of DDR3, so I don't really have to worry about it.The belief that an SSD affects any other aspect of gameplay betrays a fundamental ignorance as to how computers work in general.[/citation]

You don't know much about these games at all. Many games load things off of the hard drive even during play. The hard drive is not always fast enough, so you can get stutter and untextured artifacts in some games for a short amount of time (WoW can be a fairly bad offender of this) as the textures and such are loading. Tom's has already shown us how an SSD solves this problem.

Having 16GB of RAM is absolutely useless for gaming, unless you are multi-tasking as you play. In fact, having more than one DIMM per channel (assuming that you have a 4x4GB module configuration) decreases performance slightly when compared to having only one module per channel (for a dual channel, 2x4GB) because of the additional load on the memory controller. Going over 8GB has no benefit for gaming unless you are multi-tasking, so that you have 16GB means that you either mistakenly believe that having 16GB helps you in gaming, or you do more than gaming on that system.

Gamers who know what RAM they should have know that they should only have the next step over 4GB for their system. For dual and quad channel, that's 8GB. For triple channel, that's 6GB. Going over this decreases performance slightly instead of improving it.

 

[PCgamer81]

I think I know where I have been misunderstanding you.

Here is a likely scenario that pretty much explains it:

Performance with HDD

min. 18
avg. 55
max. 77

Performance with SSD

min. 38
avg. 61
max. 77

The above scenario is extremely likely, of that I have no doubt.

I thought you were arguing that an SSD yields a higher maximum framerate on average.

I am fully aware that when caching errors occur in-game (and they do), the framerate dips most severely as a result. Such dips factor into the minimum framerate, and can significantly reduce the average framerate depending on how often such errors occur.

Are we on the same page?

 

[PCgamer81]

I could achieve higher and more stable clocks with 2x4GB rather than 4x4GB, no doubt about it. No argument from me there.

If I ever need more power or become unstable, I can always pull a couple of sticks.

I do use the RAM.

 

[silverblue]

[citation][nom]PCgamer81[/nom]The above scenario is extremely likely, of that I have no doubt.[/citation]
Exactly what we've been trying to say.

 

[Guest]

Okay after all this nonsense of offtopic SSD discussion.....

Was this texture issue fixed in the 12.4 WHQL driver then ????

 

[blazorthon]

[citation][nom]PCgamer81[/nom]I think I know where I have been misunderstanding you.Here is a likely scenario that pretty much explains iterformance with HDDmin. 18avg. 55max. 77Performance with SSDmin. 38avg. 61max. 77The above scenario is extremely likely, of that I have no doubt.I thought you were arguing that an SSD yields a higher maximum framerate on average.I am fully aware that when caching errors occur in-game (and they do), the framerate dips most severely as a result. Such dips factor into the minimum framerate, and can significantly reduce the average framerate depending on how often such errors occur.Are we on the same page?[/citation]

I suppose. I never meant that SSDs help the maximum frame rates. Well, I guess that this is cleared up now.

 

[PCgamer81]

I always look at increasing performance as increasing maximum fps. As I was reading around, it occurred to me the error I was making.

That is why I kept bringing up RAM bottlenecks (maximum fps being improved by an SSD).

Anyway, back to topic...

 

[dkraptor]

[citation][nom]nebun[/nom]got to love nvidia....why is it that most of AMD's graphics are blurry?[/citation]
[citation][nom]nebun[/nom]Almost forgot...AMD was caught lying again ...what a shame[/citation]
obviously, you didn't even bother to read the article. Go back to your facebook page.

 

[mstngs351]

Glad to see the fix was quick and doesn't seem to hit performance. I do feel that a card company lowering quality to get better frame rates is misleading and wrong. It doesn't appear that is what happened here but I can't help but wonder if maybe at the time the fuzzy drivers were launched if it would have been a bigger performance hit. Maybe, maybe not.

 

[mstngs351]

[citation][nom]TheRabidDeer[/nom]Is it just me or is toms heavily biased towards nvidia? We see tons of articles for the Nvidia 6xx but very few for the 7xxx. Nothing negative for nvidia, but an article like this for AMD's, which is already being fixed even though it is undetectable... and the fix doesnt even yield a real change in framerates.[/citation]
Nah I think you're just imagining things. Sometimes it's easy to do. The number of articles is usually fairly well split esp when you consider that there are more types of nvidia (not necessarily a good thing) cards out there. Keep in mind that at the moment Nvidia has just launched a new awaited product so naturally they will get more of the spotlight for a short time. Just like when AMD launches a new line. Also if you read this article carefully it wasn't negative. Simply factual.

 

[dmnwlv]

Try looking for any difference in image after changing certain settings, especially aa, in nvidia control panel. You will find no effect on some games like magicka and AoE Online. I wonder have they fixed that after feeding back awhile ago. (If i remember correctly, AoE Online can be effective only in global settings while Magicka totally no effect.)

 

[dmnwlv]

Oh and nvidia panel settings dnt take immediate effect without restarting the game which can be quite hard to fine tune settings.

 

[oxford373]

i wanted to talk about last year nvidias accusation to amd about catalyst A.I. it wasn't cheat at all because ATI since the first driver that equipped with catalyst A.I. amd told us about this optimization and toms hardware also told us about this optimization at ATI x700 review and toms hardware disabled catalyst A.I. at that review http://www.tomshardware.com/reviews/ati-hits-back,884-6.htmlso it wasn't a secret especially for old toms hardware reader like me or the person who wrote ATI X700 review
actually its just couple of frames and no body can notice the difference between catalyst A.I. enabled or disabled

 

[blazorthon]

[citation][nom]oxford373[/nom]i wanted to talk about last year nvidias accusation to amd about catalyst A.I. it wasn't cheat at all because ATI since the first driver that equipped with catalyst A.I. told us about this optimization and toms hardware also told us about this optimization at ATI x700 review and you disabled catalyst A.I. at that review so it wasn't a secret especially for old toms hardware reader like me or the person who wrote ATI X700 review at 1:01 PM - September 21, 2004 by Lars Weinand i think this is the final judgment about that accusation[/citation]

Article clearly states that the slight decrease in picture quality wasn't a performance optimization. Considering that it was fixed by a new driver which not only fixed the quality problem, but also fixed it without a decrease in performance, all this article did was tell us that there was a minor driver issue that AMD has since fixed and Tom's wanted to tell us about it and the role that they played in helping AMD to realize that they had a problem. Heck, do you think that AMD employees will take screen shots and blow them up several hundred percent and compare them to shots acquired similarly with other graphics card setups?

If they didn't see the problem during gameplay (and most buyers didn't either) and it wasn't really noticed until Tom's did this very article (even then, many people were still commenting about how they couldn't see the problem and I think that that is a testament to how extremely minor it was) that shows us the problem, then I think that its safe to say that it was completely incidental. That it was fixed without a performance drop tells us that it wasn't some *performance optimization*, but just a simple problem that was got by undetected because it was very small.

Lets also not pretend that Nvidia wasn't on the *performance optimization* bandwagon those years ago too.

 

[oxford373]

for me as i wrote up catalyst A.I. was never a cheat but nvidia cheated many times and they continue to do it over and over again ,when nvidia acquired agiea physx they did everything can to cripple CPU-based PhysX mode like x87 instruction set which physx is the only application in the world still use this very old one and they made it single threaded actually they cant cripple CPU-based PhysX mode more than that even if they want to,and every game with( the way meant to be played) logo still suspicious especially if nvidia's cards outperforms amd's .