Sorry, PC World, I know you're a big mainstream magazine and website, and I'm just a little independent blogger, but this article of yours blows. The opening discussion of SLI, Eyefinity, and multi-display support has more holes in the than my spaghetti strainer. Call it nit-picking if you want, I call it writing a halfway intelligent article. I'm insulted that you get paid to produce this pathetic, mindless drivel.
My two year old son does better fact checking than you. He knows damn well whether his cup has juice, water, or milk in it, and he can tell the difference between cereal, cookies, and cake.
Where to begin? How about the part where you say you need SLI to run more than two displays with Nvidia cards? How about just needing two graphics cards. SLI is running multiple identical graphics cards together for a 3D performance boost. You can run two dissimilar cards together, not linked with a bridge, and use the extra outputs.
But hey, enough about the fun details. How about this fact. I'm the next best thing to an EVGA fanboy. The only thing I'm missing is the compulsion to suggest their products when they'd be inappropriate, or otherwise not be objective about things like price/performance. That means I invariably have Nvidia cards. And I STILL say your benchmarks are over the top biased, moronic, pathetic, idiotic, and otherwise pure and utter garbage.
Crysis 2? Dirt? A grand total of two games and a couple of synthetic benches? What the hell are you smoking? I would be embarrassed to call that research. And that's completely ignoring the part where you picked games that nvidia tends to do well on. And the part where you benched the 560 Ti against a 6870, when a 6950 is much closer to the same price point. HD 6950 goes pretty much blow for blow against a 560 Ti.
Which card do I personally prefer? 560 Ti, definitely. But your methodology is trash. Of course, you turn around and show your bipartisan idiocy with declaring the HD 6970 superior to the GTX 570 based on features. If you're running multi-display gaming, whether it's 3DSurround or Eyefinity, you probably don't want 5760x1080 resolution running at low details, or at 3 frames per decade. Which is what's going to happen if you run single card Eyefinity. Maybe with iPhone's for the displays.
Last but not least... GTX 570 vs HD 6970 is somehow the "Mainstream gamer" market? That's hysterical. Last I checked, mainstream is usually... mainstream. The best gaming hardware census out there, Steam, shows those two cards combining to make up roughly 5% of DX11 capable GPUs. That, of course, is ignoring all the DX9 and DX10 cards still in use since console ports have killed the hardware industry. GTX 570 makes up .99% of all GPUs, 6970 falls under "Other", so who knows how much they make up. Last time I checked, that's not mainstream.
Article is laughable, guys. But at least it doesn't sound like advertising, since it's stupid all around. Or if it does, it's attempting to advertise direct competitors in the same article. Which would be in keeping with the quality you showed.
Tech tips, rants, and possibly occasional benchmarks and product reviews. Occasional Enthusiast PC component discussion, mockery of tech stuff, and vehement opinion, all from the rather unique viewpoint of an ex-Infantry, Stay-At-Home-Dad Geek. If you like it, subscribe to the feed.
Showing posts with label SLI. Show all posts
Showing posts with label SLI. Show all posts
Wednesday, October 26, 2011
Saturday, October 15, 2011
Anti Aliasing, GPU Murder, Justified, or No?
Anti-Aliasing, multi-sampling, 8x, 16x? Analytical? FSAA? What the hell is this shit, why can't I max it, and what am I looking for that makes it worth turning my GPU into an EZ-Bake oven? It can be kind of annoying when you're beating your framerate into the floor for something that's pretty hard to spot specifically unless you know what to look for.
Lucky for you, you don't need to look any further to find at least some of the answers. Simply put, AA is designed to make angled line on square pixels not look like a staircase. Since each pixel can only have one RGB value at any given time, you can't just have it be half and half with two different colors. The way Anti-aliasing fixes this, roughly, is to take raw texture data for some or all of the pixels involved, and kind of blend the colors in a way that lends itself to both sides of the line, making the transition appear smooth.
One of the reasons this is difficult to spot for a lot of people is pixel density. Most displays have fairly high density, which basically just refers to how much area the pixels are packed into. 22" 1920x1080 has the same number of pixels to play with as 30" 1920x1080. That means that the bigger your display, the larger the pixel, which can exaggerate the staircase shape on angled lines.
The fun thing here, is that the better your pixel density, the less you need AA. You might need 4x or so at 22" 1080p, but at 32", you should be trying to max it for lines that look similarly smooth. The different varieties and multiples involved, like 8xMSAA, or 4xSSAA are just different methods that can be used to determine what the resulting pixel data will be.
These three images were originally taken in 1080p from Dragon Age 2, DX11, Highest quality everything, with the only thing changing in each shot being the level of Anti-Aliasing, as seen in the captions.
Now that I've explained anti-aliasing, spotting it should be easy, right?
I'm guessing a lot of you out there actually have to work really hard to spot it, right? Like I said, the smaller the pixel, the smaller the square shape, the smoother the line, even without anti-aliasing.
On my 32" TV, I can barely see the difference between 4x and 8x at desk viewing distance.
So, now that we've gone over exactly how big of a difference it makes, you can see that trying to max this may not really matter for your overall gaming experience. Given the massive GPU horsepower needed for high AA with decent quality textures, it can be an expensive prospect to truly max every game that comes out.
But, if you can live without something you can't see or can barely see, you can generally get by with a lot less power. The exact difference will vary a good bit from game to game, but frequently, if you're getting slightly jerky FPS at 8xAA, dropping to 4xAA will get you fairly smooth, in my experience. That's a big performance difference for a barely visible change.
Now obviously, this is one of those things where everyone needs to draw their own conclusions. I'll keep using insane PCs, because getting performance out of them is half the fun. But for people wanting performance on a budget, check the AA used in benchmarks, and you might just find that you'll be ok with a card that doesn't look quite as pretty if all you see is how long the bars are.
Lucky for you, you don't need to look any further to find at least some of the answers. Simply put, AA is designed to make angled line on square pixels not look like a staircase. Since each pixel can only have one RGB value at any given time, you can't just have it be half and half with two different colors. The way Anti-aliasing fixes this, roughly, is to take raw texture data for some or all of the pixels involved, and kind of blend the colors in a way that lends itself to both sides of the line, making the transition appear smooth.
One of the reasons this is difficult to spot for a lot of people is pixel density. Most displays have fairly high density, which basically just refers to how much area the pixels are packed into. 22" 1920x1080 has the same number of pixels to play with as 30" 1920x1080. That means that the bigger your display, the larger the pixel, which can exaggerate the staircase shape on angled lines.
The fun thing here, is that the better your pixel density, the less you need AA. You might need 4x or so at 22" 1080p, but at 32", you should be trying to max it for lines that look similarly smooth. The different varieties and multiples involved, like 8xMSAA, or 4xSSAA are just different methods that can be used to determine what the resulting pixel data will be.
![]() |
| 0xAA |
![]() |
| 4xaa |
![]() | |
| 8xAA |
These three images were originally taken in 1080p from Dragon Age 2, DX11, Highest quality everything, with the only thing changing in each shot being the level of Anti-Aliasing, as seen in the captions.
Now that I've explained anti-aliasing, spotting it should be easy, right?
I'm guessing a lot of you out there actually have to work really hard to spot it, right? Like I said, the smaller the pixel, the smaller the square shape, the smoother the line, even without anti-aliasing.
On my 32" TV, I can barely see the difference between 4x and 8x at desk viewing distance.
So, now that we've gone over exactly how big of a difference it makes, you can see that trying to max this may not really matter for your overall gaming experience. Given the massive GPU horsepower needed for high AA with decent quality textures, it can be an expensive prospect to truly max every game that comes out.
But, if you can live without something you can't see or can barely see, you can generally get by with a lot less power. The exact difference will vary a good bit from game to game, but frequently, if you're getting slightly jerky FPS at 8xAA, dropping to 4xAA will get you fairly smooth, in my experience. That's a big performance difference for a barely visible change.
Now obviously, this is one of those things where everyone needs to draw their own conclusions. I'll keep using insane PCs, because getting performance out of them is half the fun. But for people wanting performance on a budget, check the AA used in benchmarks, and you might just find that you'll be ok with a card that doesn't look quite as pretty if all you see is how long the bars are.
Labels:
AMD,
benchmark,
computer,
Crossfire,
directx,
Enthusiast,
fps,
framerate,
GPU,
graphics,
memory,
RAM,
review,
SLI,
vertical sync
Practical DX11, What It Does, and What It Means
As we all know, the current generation of Microsoft's ubiquitous and ambiguous DirectX runtime is DX11. Just to make that sound less like techno-babble designed to keep software engineers in a job, DirectX is the overall software package for handling graphics in Windows. (There's also some sound stuff, but that's outside the scope of this article.) Since DirectX includes features for 3D gaming, it can be kind of important to gamers and enthusiasts to have hardware supporting at least recent versions.
Most games right now, mind you, only require DX9 to play, due to the red-headed step brother, consoles. DX10, well, yeah, that got released at some point, and did some stuff, and nobody really gave a shit. DX11, on the other hand, has this wonky tessellation, which mostly sounds like it fell out of a science fiction novel, possibly as some really painful way to die.
Basically, to explain the portions of tessellation that really made sense beyond "ooh shiny" without two or three different degrees in graphics design, software, and who knows what else, it's just a different way of representing shapes. Tessellation, roughly, lets you break shapes down into smaller shapes. The primary use for it is for smoothing out certain kinds of detail.
Roughly, tessellation breaks shapes down into tiny triangles, letting you make what's called a displacement map. To make this make sense, do you remember the little plexiglass box toy with a whole lot of pins that could slide up and down, so when you set it on top of something, from the top it would be a 3-D representation of the object?
Roughly, this is what the displacement mapping does in 3-D rendering. Not specifically, but it lets you take a shape, and break it down into smaller shapes that can be worked with like this. That way, as you get closer to something, it can render it's way toward a model, instead of trying to render all the models and textures in full detail all the way out to max viewing distance.
I hope this makes some sense, this is all sounding smarter in my head than it looks once I type it, but I can't decide if lack of caffeine is affecting my reading or my writing. Maybe I'm just crazy. Or all of the above.
Don't get me wrong, this is a very basic and generalized explanation, which by nature will be somewhat wonky. If you want to read the fancy version, you can check it out here.
Ok, so, here's the question, what's this mean for me? Smoother textures and better poly mapping, making things look better all around, for starters. Pretty simple, and nice. The second major thing you'll see out of it is textures being able to look good close in without soaking up a ton of GPU resources at a long distance. No, this doesn't sound huge, but what it does is allow an increased view distance, since it can avoid rendering so much at a distance, meaning less stuff pops out of thin air.
I hope this has cleared this up reasonably well for some people, I know it's not that useful for some, but understanding graphics can help you understand what sort of system you personally need for the graphics quality you desire. I'm hoping to do some other similar ones over the next few days, so stay tuned.
Most games right now, mind you, only require DX9 to play, due to the red-headed step brother, consoles. DX10, well, yeah, that got released at some point, and did some stuff, and nobody really gave a shit. DX11, on the other hand, has this wonky tessellation, which mostly sounds like it fell out of a science fiction novel, possibly as some really painful way to die.
Basically, to explain the portions of tessellation that really made sense beyond "ooh shiny" without two or three different degrees in graphics design, software, and who knows what else, it's just a different way of representing shapes. Tessellation, roughly, lets you break shapes down into smaller shapes. The primary use for it is for smoothing out certain kinds of detail.
Roughly, tessellation breaks shapes down into tiny triangles, letting you make what's called a displacement map. To make this make sense, do you remember the little plexiglass box toy with a whole lot of pins that could slide up and down, so when you set it on top of something, from the top it would be a 3-D representation of the object?
Roughly, this is what the displacement mapping does in 3-D rendering. Not specifically, but it lets you take a shape, and break it down into smaller shapes that can be worked with like this. That way, as you get closer to something, it can render it's way toward a model, instead of trying to render all the models and textures in full detail all the way out to max viewing distance.
I hope this makes some sense, this is all sounding smarter in my head than it looks once I type it, but I can't decide if lack of caffeine is affecting my reading or my writing. Maybe I'm just crazy. Or all of the above.
Don't get me wrong, this is a very basic and generalized explanation, which by nature will be somewhat wonky. If you want to read the fancy version, you can check it out here.
Ok, so, here's the question, what's this mean for me? Smoother textures and better poly mapping, making things look better all around, for starters. Pretty simple, and nice. The second major thing you'll see out of it is textures being able to look good close in without soaking up a ton of GPU resources at a long distance. No, this doesn't sound huge, but what it does is allow an increased view distance, since it can avoid rendering so much at a distance, meaning less stuff pops out of thin air.
I hope this has cleared this up reasonably well for some people, I know it's not that useful for some, but understanding graphics can help you understand what sort of system you personally need for the graphics quality you desire. I'm hoping to do some other similar ones over the next few days, so stay tuned.
Labels:
AMD,
benchmark,
computer,
Crossfire,
directx,
display,
Enthusiast,
fps,
framerate,
gamer,
GPU,
graphics,
JingleHell,
memory,
monitor,
overclock,
PC,
RAM,
SLI,
tesselation
Saturday, October 8, 2011
Multi-GPU and Multi-Monitor Un-Confused
Eyefinity, SLI, 3DSurround, 3DVision, and CrossfireX. Multiple monitors. What does it all mean? What's the difference, in this epic pile of confusing terminology, and what do I need to make this work?
SLI and CrossfireX, for starters. These are multiple-GPU configurations designed for extra graphics power on a single display. These are the ones most people think of first in multi-GPU configurations, and the terms most likely to be generically thrown in as a placeholder value. SLI is the nVidia variant, CrossfireX is AMD. CrossfireX only requires multiple PCIE slots, SLI requires an SLI ready motherboard, which, as of recently, can include AMD motherboards.
3DVision: This is using funky technology and woogy glasses to make things pop out of the screen, like a bad '80s sci-fi flick. This hasn't been implemented all that well in a lot of games. It does, however, occasionally get mixed up with 3DSurround, due to the stupid similarity in naming.
3DSurround and Eyefinity are the methods of playing games across multiple displays. Not gaming on one and having a browser or other stuff going in a spare display, that's just a normal multiple display configuration. Some AMD cards will support Eyefinity off a single card, but it's pretty stupid. No nVidia solutions do that, which is really perfectly reasonable, since playing across multiple displays doesn't make a whole lot of sense if you're having to drop the settings to "fuzzy blob" to get the extra screen real estate. By the way, use an odd number of displays per row for this, you don't want lines in the middle of the game area.
Multiple monitor setups just include using two monitors, with at most one for gaming. This is pretty much supported by almost everything these days, since it's not all that harsh for non-3D use. It can be handy if you actually have a use for the screen real estate, but for a lot of people it's basically just a fancy thing.
Remember, for SLI, Eyefinity, 3DSurround, and CrossfireX, you need identical cards. You can't SLI a GTX 560Ti and a GT 9800. You could use the 9800 for dedicated PhysX, or for extra monitor outs, but you can't get the two different cards running SLI for alternate frame rendering.
No, this isn't long, really, but it's something that drives me nuts occasionally.
SLI and CrossfireX, for starters. These are multiple-GPU configurations designed for extra graphics power on a single display. These are the ones most people think of first in multi-GPU configurations, and the terms most likely to be generically thrown in as a placeholder value. SLI is the nVidia variant, CrossfireX is AMD. CrossfireX only requires multiple PCIE slots, SLI requires an SLI ready motherboard, which, as of recently, can include AMD motherboards.
3DVision: This is using funky technology and woogy glasses to make things pop out of the screen, like a bad '80s sci-fi flick. This hasn't been implemented all that well in a lot of games. It does, however, occasionally get mixed up with 3DSurround, due to the stupid similarity in naming.
3DSurround and Eyefinity are the methods of playing games across multiple displays. Not gaming on one and having a browser or other stuff going in a spare display, that's just a normal multiple display configuration. Some AMD cards will support Eyefinity off a single card, but it's pretty stupid. No nVidia solutions do that, which is really perfectly reasonable, since playing across multiple displays doesn't make a whole lot of sense if you're having to drop the settings to "fuzzy blob" to get the extra screen real estate. By the way, use an odd number of displays per row for this, you don't want lines in the middle of the game area.
Multiple monitor setups just include using two monitors, with at most one for gaming. This is pretty much supported by almost everything these days, since it's not all that harsh for non-3D use. It can be handy if you actually have a use for the screen real estate, but for a lot of people it's basically just a fancy thing.
Remember, for SLI, Eyefinity, 3DSurround, and CrossfireX, you need identical cards. You can't SLI a GTX 560Ti and a GT 9800. You could use the 9800 for dedicated PhysX, or for extra monitor outs, but you can't get the two different cards running SLI for alternate frame rendering.
No, this isn't long, really, but it's something that drives me nuts occasionally.
Labels:
computer,
Crossfire,
display,
DPI,
Enthusiast,
fps,
framerate,
gamer,
GPU,
JingleHell,
memory,
monitor,
PC,
RAM,
SLI
Saturday, October 1, 2011
Airflow, Cases, Coolers, or DIY Wind Tunnel Kits.
Ok, so I've been thinking about a rant on this topic for a while anyway, but a good friend's predicament made me decide to go ahead and do it. For the sake of this post, we'll call him "Thomas", to avoid any potential embarrassment.
If you've ever seen a high end Enterprise workstation, or an older PC, you might remember something being vaguely different from a lot of PC's these days. If you can guess what I'm talking about, raise your hand. Then feel like a moron, since nobody can see your hand. The thing I'm talking about, of course, is the fact that those computers don't move enough air to achieve escape velocity. Holy shit, crazy talk, I know. No turbofan sitting next to your desk? How dare I blaspheme the pantheon of PC case manufacturers that way?
Don't get me wrong, there's absolutely nothing wrong with a case with a decent amount of airflow. But a decent amount of airflow and a decent amount of moving air that doesn't do a damn thing are entirely different animals. My friend "Thomas" is a great example of this. He has some cheap-ass gamer case that Ali-Baba and the CyberPower Thieves stuck him with. It has two front 120mm intakes, top-mounted isolated PSU intake, optional (in this case installed) 140mm top exhaust, the radiator for his (ugh) Corsair H50 mounted as an exhaust in the back, and one or two side 120mm fans. That's a metric asston of air moving around, in case you hadn't noticed. Somehow, though, he can't keep SLI GTX460's very cool. Granted, that's not always easy, but it's not insanely difficult, either, and he's hitting dangerous temps under load.
So. why am I bringing this up? Well, airflow is complicated, and I'm not going to claim to know everything (or even close) about aerodynamics. What I do know, though, is that a lot of things people think they know about PC cooling are very very wrong. A popular myth: "Get a bigger case to keep things cooler." Horribly wrong. Get a case that's got room for your components. Large amounts of dead airspace with no components create a path of least resistance for airflow that doesn't serve a purpose.
There's a couple of (general) rules to follow when trying to figure out your airflow.
1: Know your components. Nothing more important than knowing your configuration's needs. Granted, these can change later on with upgrades and whatnot, but it's still good to know.
2: Keep dead air space to a minimum within reason. Don't cramp your components, trapping heat is bad. But once your components fit reasonably, too much more space makes for bad airflow to where you need it.
3: No competing airflow. Avoid cases where your PSU is going to be fighting your CPU cooler or GPU airflow.
Here's the thing. Air follows the path of least resistance. Hot air rises, and heat is bad stuff. Because of this, if you can stand the slightly higher noise, External Exhaust graphics cards are good stuff, since they blow the hot air out of the case, unlike radial coolers that just get the heat away from the GPU and let it rise up through the CPU cooler. Negative pressure cases are ok for systems with limited restriction on airflow, but if there's a lot of stuff in there, actively directing the airflow with a positive pressure system is frequently better.
Don't forget to prioritize cooling. RAM doesn't need cooling as much as your CPU, so don't stick some retarded fan over your memory that thrashes airflow for your CPU. If you have a side panel, push air into your GPU if it's external exhaust, or pull air away before it can rise if it's radial cooling. If you're using a closed loop liquid cooler, I don't care how much it helps your CPU temps to use the radiator as an intake, just say no to pulling heat into your case.
Above all, apply common sense. The more directly cool air gets to your components, and hot air gets away from them, the better. Always. Avoid dead air, and avoid competing airflow. Avoid pointless restriction.
Remember, a lot of moving air doesn't mean good cooling. A space shuttle launch moves a shitload of air, and that's not cold at all.
If you've ever seen a high end Enterprise workstation, or an older PC, you might remember something being vaguely different from a lot of PC's these days. If you can guess what I'm talking about, raise your hand. Then feel like a moron, since nobody can see your hand. The thing I'm talking about, of course, is the fact that those computers don't move enough air to achieve escape velocity. Holy shit, crazy talk, I know. No turbofan sitting next to your desk? How dare I blaspheme the pantheon of PC case manufacturers that way?
Don't get me wrong, there's absolutely nothing wrong with a case with a decent amount of airflow. But a decent amount of airflow and a decent amount of moving air that doesn't do a damn thing are entirely different animals. My friend "Thomas" is a great example of this. He has some cheap-ass gamer case that Ali-Baba and the CyberPower Thieves stuck him with. It has two front 120mm intakes, top-mounted isolated PSU intake, optional (in this case installed) 140mm top exhaust, the radiator for his (ugh) Corsair H50 mounted as an exhaust in the back, and one or two side 120mm fans. That's a metric asston of air moving around, in case you hadn't noticed. Somehow, though, he can't keep SLI GTX460's very cool. Granted, that's not always easy, but it's not insanely difficult, either, and he's hitting dangerous temps under load.
So. why am I bringing this up? Well, airflow is complicated, and I'm not going to claim to know everything (or even close) about aerodynamics. What I do know, though, is that a lot of things people think they know about PC cooling are very very wrong. A popular myth: "Get a bigger case to keep things cooler." Horribly wrong. Get a case that's got room for your components. Large amounts of dead airspace with no components create a path of least resistance for airflow that doesn't serve a purpose.
There's a couple of (general) rules to follow when trying to figure out your airflow.
1: Know your components. Nothing more important than knowing your configuration's needs. Granted, these can change later on with upgrades and whatnot, but it's still good to know.
2: Keep dead air space to a minimum within reason. Don't cramp your components, trapping heat is bad. But once your components fit reasonably, too much more space makes for bad airflow to where you need it.
3: No competing airflow. Avoid cases where your PSU is going to be fighting your CPU cooler or GPU airflow.
Here's the thing. Air follows the path of least resistance. Hot air rises, and heat is bad stuff. Because of this, if you can stand the slightly higher noise, External Exhaust graphics cards are good stuff, since they blow the hot air out of the case, unlike radial coolers that just get the heat away from the GPU and let it rise up through the CPU cooler. Negative pressure cases are ok for systems with limited restriction on airflow, but if there's a lot of stuff in there, actively directing the airflow with a positive pressure system is frequently better.
Don't forget to prioritize cooling. RAM doesn't need cooling as much as your CPU, so don't stick some retarded fan over your memory that thrashes airflow for your CPU. If you have a side panel, push air into your GPU if it's external exhaust, or pull air away before it can rise if it's radial cooling. If you're using a closed loop liquid cooler, I don't care how much it helps your CPU temps to use the radiator as an intake, just say no to pulling heat into your case.
Above all, apply common sense. The more directly cool air gets to your components, and hot air gets away from them, the better. Always. Avoid dead air, and avoid competing airflow. Avoid pointless restriction.
Remember, a lot of moving air doesn't mean good cooling. A space shuttle launch moves a shitload of air, and that's not cold at all.
Monday, September 26, 2011
Benchmarking for Fun (Or for Info.)
So if you frequent tech forums, you might hear something along the lines of "...only matters in synthetic benchmarks." or "You can't tell the difference without a benchmark." And, if you like the idea of having a slightly overpowered system, you may just want to know how to go about those benchmarks. Even if that's not the case, you may be wondering what will help your performance in a particular game the most, or wanting to find out just how effective your CPU or GPU is for a certain game.
If you read a lot of reviews, you've probably seen all the methodology, all the different names of software, and seen tons of numbers with pretty bars and graphs. Well, I hate to tell you this, but generally, the pretty bars and graphs don't just come with the software. Luckily, it's the numbers that matter anyway.
So, for starters, lets discuss benching for fun. The most commonly used benches for gaming performance are Futuremark's 3DMark series. The "standard" comparison points are the default settings, which you can get in the free trial version. Lets you get a better database of scores going.
If you're going to bench for fun, the most important thing to remember is that you need a consistent set of benching processes. The best way to do this is to have a secondary account on your PC with the absolute minimum of automatic services, to make sure memory use is consistent. Keep drivers updated, although you usually won't want to use Beta drivers unless you need a specific feature. (Always test beta drivers prior to sustained use, some graphics drivers have been known to cause thermal issues.)
So, now that you have your consistent stuff, you get a baseline. That's just at your normal settings, how does it work. After that, you can try tweaking things to see if they gain you performance. Whatever you decide to tweak, make sure you test with the same processes as before. It really can make a huge difference in scores, or enough to skew data to the point of unreliability.
Generally, the GPU is the easiest thing to tweak, since there's simple GUI based OC utilities that work reasonably well, like MSI Afterburner, or (to a point) EVGA Precision. Remember to push it up a little at a time, and test for stability before starting the benchmark. If you raise your GPU OC and suddenly lose performance, you either need to raise your GPU voltage (at your own risk), or lower the clock back down a bit. Generally, in graphics cards, RAM clock will be less performance for the voltage than core clock and shaders, so you generally won't want to bother with it.
The CPU, is, of course, one of those things people really think about overclocking, and a lot of people are scared of it. If you do your homework, have the right components, and do it carefully, it can be a perfectly safe way to gain performance. You only void your warranty if you do damage that can definitely be attributed to overclocking, so usually you're safe if you keep the OC within the CPU manufacturers specified voltage range.
This isn't an overclocking guide, but I will say that in recent i7's, it's usually best to raise clock to the desired level first, then see if you can squeeze hyperthreading back on safely. Hyperthreading will raise your score in 3DMark, but less than most clock speed gains. RAM speed and timings can also affect it, as can IMC clock, and basically all the usual culprits.
But now, I'm sure, you're wondering about the benching for information thing. After all, E-peenery is fun, but there's only so much you can say about it. If you want to learn something through benchmarking, you're going to need a few things.
1: Consistent Monitoring. Be it FRAPS for FPS/Frametimes, HWMonitor for temps, or something else, you must use software. Eyeballing it is worse than useless.
2: Consistent Playthrough. You need a replay, or a specific save file, or something that you always use for benchmarking. If you're getting an FPS recording of different things, it doesn't tell you much.
3: Objectivity. This is the key. If you aren't objective, you're useless. You need to be willing to get results you didn't want or expect.
Well, as for methodology, it depends a bit, based on what exactly you're wanting to test, but essentially you need to isolate a component. For CPU/RAM testing, graphics settings not requiring the CPU should be turned down, to include resolution. This ensures that the limiting factor will be your CPU. For RAM, you just get a baseline at a "Typical" setting, like CL9 1333 for DDR3, and work from there, using the same CPU clock.
If you want to isolate GPU, you crank the desired graphics settings, turn down anything you can involving the CPU, and hope you aren't trying to isolate the graphics card on an RTS. Won't happen unless you have a really screwy rig.
If you want to monitor temps, you need a consistent ambient temperature, and consistency in everything except what you're testing. Including what you use to heat up the PC, and how long you run it prior to measuring.
I know this hasn't really been long on details, but it's really more to get you thinking the right way.
If you read a lot of reviews, you've probably seen all the methodology, all the different names of software, and seen tons of numbers with pretty bars and graphs. Well, I hate to tell you this, but generally, the pretty bars and graphs don't just come with the software. Luckily, it's the numbers that matter anyway.
So, for starters, lets discuss benching for fun. The most commonly used benches for gaming performance are Futuremark's 3DMark series. The "standard" comparison points are the default settings, which you can get in the free trial version. Lets you get a better database of scores going.
If you're going to bench for fun, the most important thing to remember is that you need a consistent set of benching processes. The best way to do this is to have a secondary account on your PC with the absolute minimum of automatic services, to make sure memory use is consistent. Keep drivers updated, although you usually won't want to use Beta drivers unless you need a specific feature. (Always test beta drivers prior to sustained use, some graphics drivers have been known to cause thermal issues.)
So, now that you have your consistent stuff, you get a baseline. That's just at your normal settings, how does it work. After that, you can try tweaking things to see if they gain you performance. Whatever you decide to tweak, make sure you test with the same processes as before. It really can make a huge difference in scores, or enough to skew data to the point of unreliability.
Generally, the GPU is the easiest thing to tweak, since there's simple GUI based OC utilities that work reasonably well, like MSI Afterburner, or (to a point) EVGA Precision. Remember to push it up a little at a time, and test for stability before starting the benchmark. If you raise your GPU OC and suddenly lose performance, you either need to raise your GPU voltage (at your own risk), or lower the clock back down a bit. Generally, in graphics cards, RAM clock will be less performance for the voltage than core clock and shaders, so you generally won't want to bother with it.
The CPU, is, of course, one of those things people really think about overclocking, and a lot of people are scared of it. If you do your homework, have the right components, and do it carefully, it can be a perfectly safe way to gain performance. You only void your warranty if you do damage that can definitely be attributed to overclocking, so usually you're safe if you keep the OC within the CPU manufacturers specified voltage range.
This isn't an overclocking guide, but I will say that in recent i7's, it's usually best to raise clock to the desired level first, then see if you can squeeze hyperthreading back on safely. Hyperthreading will raise your score in 3DMark, but less than most clock speed gains. RAM speed and timings can also affect it, as can IMC clock, and basically all the usual culprits.
But now, I'm sure, you're wondering about the benching for information thing. After all, E-peenery is fun, but there's only so much you can say about it. If you want to learn something through benchmarking, you're going to need a few things.
1: Consistent Monitoring. Be it FRAPS for FPS/Frametimes, HWMonitor for temps, or something else, you must use software. Eyeballing it is worse than useless.
2: Consistent Playthrough. You need a replay, or a specific save file, or something that you always use for benchmarking. If you're getting an FPS recording of different things, it doesn't tell you much.
3: Objectivity. This is the key. If you aren't objective, you're useless. You need to be willing to get results you didn't want or expect.
Well, as for methodology, it depends a bit, based on what exactly you're wanting to test, but essentially you need to isolate a component. For CPU/RAM testing, graphics settings not requiring the CPU should be turned down, to include resolution. This ensures that the limiting factor will be your CPU. For RAM, you just get a baseline at a "Typical" setting, like CL9 1333 for DDR3, and work from there, using the same CPU clock.
If you want to isolate GPU, you crank the desired graphics settings, turn down anything you can involving the CPU, and hope you aren't trying to isolate the graphics card on an RTS. Won't happen unless you have a really screwy rig.
If you want to monitor temps, you need a consistent ambient temperature, and consistency in everything except what you're testing. Including what you use to heat up the PC, and how long you run it prior to measuring.
I know this hasn't really been long on details, but it's really more to get you thinking the right way.
Labels:
benchmark,
computer,
Crossfire,
Enthusiast,
fps,
framerate,
gamer,
GPU,
heat sink,
hz,
JingleHell,
liquid cool,
memory,
monitor,
overclock,
PC,
RAM,
SLI,
Tech
Tuesday, September 13, 2011
Enthusiast PC Parts and You.
Today, I thought I'd write on a topic that often gets glossed over. We all know that "enthusiast" PC components tend to involve premiums for insignificant gains, factory OC's, and branding. As such, I thought I'd write something informative on the topic.
First things first, I'm going to give you the two rules of enthusiast components.
1: Make an informed decision.
2: If you have to ask if it's a good idea, you should do a lot more research before paying more for it.
Multi-GPU Configurations
This is something that I think lots of people don't understand, and it's kind of stigmatized beyond it's issues at this point.
1: False Savings. Unless you use a dual-GPU single-PCB card (which isn't a savings by any stretch), you have to buy a better motherboard. No matter how you package the GPU's, you need better cooling and more PSU. In the case of two mid-range GPU's to top the performance of a single GPU, this will end up making it cost more, and be more of a pain to work with.
2: Driver issues and scaling. The scaling is dramatically improved in recent years, with games with decent SLI/CrossfireX support easily hitting 180%+ of a single card solution. The fact is, though, the scaling is variable, especially due to driver issues. Remember, if you buy a new release, the drivers may not work well with it day one. In that case, the driver problems will scale much faster than the GPU's, causing shit framerates compared to a single better card.
3: Tinkering. Some people see this as a feature, I'm one of them. I love tinkering endlessly with things to try and squeeze barely noticeable improvements out of my system. If you do, multi-GPU is awesome. Custom profiles tend to be the best way to get results, and this can give you hours benching a particular game and tweaking things trying to get 1-2 more FPS.
4: Noise. The increased heat from multiple cards and less airflow comes with another drawback. GPU fans get really really loud if they get loaded heavily, particularly if there's insufficient airflow, which tends to be the case with multiple dual-slot cards jammed close together on a motherboard.
You'll notice I haven't given any real definitive reason not to SLI/Crossfire? That's because it isn't objectively bad. It's objectively more work, hotter, noisier, and more power consuming, but if those things are worth it to you personally, that's perfectly fine.
Enthusiast Motherboards
For the sake of this discussion, I'm not just talking about the ability to overclock. When I say enthusiast motherboard, we're talking about feature-loaded. Tons of extra USB3 and SATA 6Gb/s, BIOS loaded with extra features, heavy chipset cooling, multiple PCI-e x16 slots for SLI/Crossfire.
Obviously, the motherboard is important. You don't want a junk one if you plan to overclock. But you can overclock without paying for features you don't need. If you aren't planning to go multi-GPU, don't pay extra for more PCI-e than you need. If you aren't going to use 6Gb SATA, don't pay for it. You don't need it for HDD's, so unless you're looking at SSD's, you can avoid getting a lot of the ports.
Good chipset and VREG cooling is handy if you plan to do a heavy OC, but for lower OC's, pretty much any board built to OC will be fine on cooling.
RAM
This is one of the big ones. You can easily pay 2-3x more for memory that gives, best case, 4-5% better performance. Given that that money could go into more CPU or GPU for bigger gains, that doesn't make sense unless those are already top end.
Getting memory working properly at better timings than CL7, or faster clock than 1600Mhz can be pretty involved, and due to weird voltage dependencies, even those numbers can be tricky, depending on your CPU clock. And again, you aren't seeing huge gains from it. if you really want to OC your memory, get a low voltage kit, not one that's been tested at higher clocks. The low voltage will let you OC the memory yourself, and they usually don't mention that above a certain clock in their testing, they were running higher VDIMM, sometimes unhealthily so.
Heat spreaders on RAM are pure gimmick factor. Servers and high end workstations don't tend to have them, why should your gaming rig? Unless you're being downright abusive, given that those chips are usually rated in the 90C range, the odds of your memory melting before your CPU or VREG are slim to none. The main two things heatsinks on memory do is block CPU coolers and look pretty. Actually, let's broaden that to all memory cooling.
Factory OC GPU's
Unless you're the epitome of lazy, these are almost always nothing but wasted money, the exception being certain cards where the only way to get certain non-reference cooling is to buy the factory OC. Overclocking a GPU is done through software with a GUI in the OS, takes little to no time, and you can find what should be stable numbers by looking at what the factory OC's run at. Not much else on this one.
Cooling
Here's the thing. If your PC was intended to have plumbing, the case would come with Drano and a plunger. It doesn't, so keep it where it belongs. The biggest problem with liquid cooling (ignoring the fact that it still bottoms out at ambient temperature, just like air), is the fact that if it dies, your PC isn't going to run until you fix it. Yous CPU can't work with no cooling, and a liquid cooler with no liquid is no cooling. A heatsink loses a fan, well, all you have to do is drop your OC down for a while and run it passive until you can get a new fan.
Spending extra money on thermal paste is downright insane for 99.99% of users. Why? Because, paying an extra $10 for 2-3C cooler at best doesn't make any sense when you could pay an extra $10 for a better aftermarket heatsink than you originally planned to get and lower temps more, or more quietly. Thermal paste will almost never be the key to making your temps comfortable for long term use, especially if you aren't shooting for records.
First things first, I'm going to give you the two rules of enthusiast components.
1: Make an informed decision.
2: If you have to ask if it's a good idea, you should do a lot more research before paying more for it.
Multi-GPU Configurations
This is something that I think lots of people don't understand, and it's kind of stigmatized beyond it's issues at this point.
1: False Savings. Unless you use a dual-GPU single-PCB card (which isn't a savings by any stretch), you have to buy a better motherboard. No matter how you package the GPU's, you need better cooling and more PSU. In the case of two mid-range GPU's to top the performance of a single GPU, this will end up making it cost more, and be more of a pain to work with.
2: Driver issues and scaling. The scaling is dramatically improved in recent years, with games with decent SLI/CrossfireX support easily hitting 180%+ of a single card solution. The fact is, though, the scaling is variable, especially due to driver issues. Remember, if you buy a new release, the drivers may not work well with it day one. In that case, the driver problems will scale much faster than the GPU's, causing shit framerates compared to a single better card.
3: Tinkering. Some people see this as a feature, I'm one of them. I love tinkering endlessly with things to try and squeeze barely noticeable improvements out of my system. If you do, multi-GPU is awesome. Custom profiles tend to be the best way to get results, and this can give you hours benching a particular game and tweaking things trying to get 1-2 more FPS.
4: Noise. The increased heat from multiple cards and less airflow comes with another drawback. GPU fans get really really loud if they get loaded heavily, particularly if there's insufficient airflow, which tends to be the case with multiple dual-slot cards jammed close together on a motherboard.
You'll notice I haven't given any real definitive reason not to SLI/Crossfire? That's because it isn't objectively bad. It's objectively more work, hotter, noisier, and more power consuming, but if those things are worth it to you personally, that's perfectly fine.
Enthusiast Motherboards
For the sake of this discussion, I'm not just talking about the ability to overclock. When I say enthusiast motherboard, we're talking about feature-loaded. Tons of extra USB3 and SATA 6Gb/s, BIOS loaded with extra features, heavy chipset cooling, multiple PCI-e x16 slots for SLI/Crossfire.
Obviously, the motherboard is important. You don't want a junk one if you plan to overclock. But you can overclock without paying for features you don't need. If you aren't planning to go multi-GPU, don't pay extra for more PCI-e than you need. If you aren't going to use 6Gb SATA, don't pay for it. You don't need it for HDD's, so unless you're looking at SSD's, you can avoid getting a lot of the ports.
Good chipset and VREG cooling is handy if you plan to do a heavy OC, but for lower OC's, pretty much any board built to OC will be fine on cooling.
RAM
This is one of the big ones. You can easily pay 2-3x more for memory that gives, best case, 4-5% better performance. Given that that money could go into more CPU or GPU for bigger gains, that doesn't make sense unless those are already top end.
Getting memory working properly at better timings than CL7, or faster clock than 1600Mhz can be pretty involved, and due to weird voltage dependencies, even those numbers can be tricky, depending on your CPU clock. And again, you aren't seeing huge gains from it. if you really want to OC your memory, get a low voltage kit, not one that's been tested at higher clocks. The low voltage will let you OC the memory yourself, and they usually don't mention that above a certain clock in their testing, they were running higher VDIMM, sometimes unhealthily so.
Heat spreaders on RAM are pure gimmick factor. Servers and high end workstations don't tend to have them, why should your gaming rig? Unless you're being downright abusive, given that those chips are usually rated in the 90C range, the odds of your memory melting before your CPU or VREG are slim to none. The main two things heatsinks on memory do is block CPU coolers and look pretty. Actually, let's broaden that to all memory cooling.
Factory OC GPU's
Unless you're the epitome of lazy, these are almost always nothing but wasted money, the exception being certain cards where the only way to get certain non-reference cooling is to buy the factory OC. Overclocking a GPU is done through software with a GUI in the OS, takes little to no time, and you can find what should be stable numbers by looking at what the factory OC's run at. Not much else on this one.
Cooling
Here's the thing. If your PC was intended to have plumbing, the case would come with Drano and a plunger. It doesn't, so keep it where it belongs. The biggest problem with liquid cooling (ignoring the fact that it still bottoms out at ambient temperature, just like air), is the fact that if it dies, your PC isn't going to run until you fix it. Yous CPU can't work with no cooling, and a liquid cooler with no liquid is no cooling. A heatsink loses a fan, well, all you have to do is drop your OC down for a while and run it passive until you can get a new fan.
Spending extra money on thermal paste is downright insane for 99.99% of users. Why? Because, paying an extra $10 for 2-3C cooler at best doesn't make any sense when you could pay an extra $10 for a better aftermarket heatsink than you originally planned to get and lower temps more, or more quietly. Thermal paste will almost never be the key to making your temps comfortable for long term use, especially if you aren't shooting for records.
Subscribe to:
Posts (Atom)



