|
On July 12 2011 09:11 Rachnar wrote:+ Show Spoiler + to max the ressources sc2 draws, storm/colossi fire and all lasers and stuff use a LOT, so try and micro a toss army vs a maxed out other toss a cliking to use the most
No. Micro=bad. Repeatability=good. Anyways, those would be the things he'd want to use to test if his GPU was causing frame drops.
When you're trying to benchmark, you want to be able to closely repeat the test for as close as possible to identical results, and for CPU, high unit count is good, so max zerglings vs max zealots will cause a fairly hefty resource drain that can be a-moved for a fairly similar situation twice in a row.
|
ah ok, but in that case i still think maxed out colossi would to it too ^^ it's just a cliking too :p
|
I just don't think it's the GPU at 1280x1024. The CPU should also be fine, but it's kind of hard to say what situation unless we see it.
But if there's something wrong with the CPU like it's locked or something. Then again, he also said it's not going above 70% usage...
If only you had dual screens, could play SC2 on one screen, have monitoring apps on another screen, and take a screenshot -_-
|
On July 12 2011 09:36 Rachnar wrote: ah ok, but in that case i still think maxed out colossi would to it too ^^ it's just a cliking too :p
Maxed out colossi have two major issues for testing it. One, they're supply heavy units. You can't pack as many on the field at a time, making it less stress on the CPU. For two, they kill things FAST, so the test duration would be shorter.
With Zerglings vs Zealots, it's going to take a bit longer, giving more time to see if the effects are similar or better with different levels of CPU settings.
It's basically just efficiency and level of strain. You go for a really harsh scenario to make problems most likely to show themselves, and you maximize the duration of the stress for a sustained test.
|
Do you lag in singleplayer games vs the AI? If you don't lag there or in replays its likely internet, but we'd have to know that first.
|
Without reading any other posts in the thread, I can say with 100% certainty that it is lagging because of your processor. Memory speeds affect performance very negligibly, and your 460 is slightly better than my video card, and I run SC2 on 1920x1200 on high/ultra. I have a core 2 duo at 3.65GHZ overclocked, and I notice that SC2 hits 100% on both cores in large battles and lags. 1280x1024 is one of the very lowest resolutions you can hit, and your graphics card is probably barely even heating up.
If the C2D is lagging at 3.7GHZ, you can bet an Athlon X4 3.8GHZ will too. At the settings you're using, I'd recommend Intel, as amd has nothing that compete higher than what you already have(extra cores mean nothing in terms of a lot of games)
|
On July 12 2011 09:58 Element)LoGiC wrote: Without reading any other posts in the thread, I can say with 100% certainty that it is lagging because of your processor. Memory speeds affect performance very negligibly, and your 460 is slightly better than my video card, and I run SC2 on 1920x1200 on high/ultra. I have a core 2 duo at 3.65GHZ overclocked, and I notice that SC2 hits 100% on both cores in large battles and lags. 1280x1024 is one of the very lowest resolutions you can hit, and your graphics card is probably barely even heating up.
If the C2D is lagging at 3.7GHZ, you can bet an Athlon X4 3.8GHZ will too. At the settings you're using, I'd recommend Intel, as amd has nothing that compete higher than what you already have(extra cores mean nothing in terms of a lot of games) Uhm he doesn't have a athlon x4. Its a phenom 955, which is quite a bit faster IIRC. And so what if he wants to use AMD? AMD cpus are often cheaper.
|
On July 12 2011 10:12 Coriolis wrote:Show nested quote +On July 12 2011 09:58 Element)LoGiC wrote: Without reading any other posts in the thread, I can say with 100% certainty that it is lagging because of your processor. Memory speeds affect performance very negligibly, and your 460 is slightly better than my video card, and I run SC2 on 1920x1200 on high/ultra. I have a core 2 duo at 3.65GHZ overclocked, and I notice that SC2 hits 100% on both cores in large battles and lags. 1280x1024 is one of the very lowest resolutions you can hit, and your graphics card is probably barely even heating up.
If the C2D is lagging at 3.7GHZ, you can bet an Athlon X4 3.8GHZ will too. At the settings you're using, I'd recommend Intel, as amd has nothing that compete higher than what you already have(extra cores mean nothing in terms of a lot of games) Uhm he doesn't have a athlon x4. Its a phenom 955, which is quite a bit faster IIRC. And so what if he wants to use AMD? AMD cpus are often cheaper.
AMD is currently being wrecked on everything except IGP performance. Now that the i3 2310 is out, I don't think they even have that slight edge they had where if you wanted a quad core CPU in the ~$400 range you went AMD... And that was barely an edge, since the i3 2100 basically smashed the contenders for it in almost everything that didn't just require 4 physical cores. This isn't really the place for this though.
Edit: I brainfarted, 2310 is definitely an i5, not i3, and it's tricky to squeeze into the $400 range without some good bargain sniping.
And the 955 vs a similarly clocked C2D is similar performance in anything that isn't taking advantage of the physical cores, so the rest of his point stands reasonably well.
|
BTW jinglehell, i was wondering how much more performance would i have from overcloking my x6 1090t ? cant do so right now but in theorie would it change much for games and all ?
|
On July 12 2011 10:30 Rachnar wrote: BTW jinglehell, i was wondering how much more performance would i have from overcloking my x6 1090t ? cant do so right now but in theorie would it change much for games and all ?
Results vary depending on the game's CPU requirements. You pretty much have to check on a game-by-game basis. Most non-RTS games you're likely to run into your GPU's limitations rather than your CPU unless you have a really poorly optimized rig. But that also depends on resolution and settings. There's no one answer, basically.
|
You probably get much of a difference from OC'ing the X6. Most times, you won't use all the cores.
But, if you already have the equipment, it's definitely worth it... little benefit it may be.
|
|
Lol, in addition, he's subtly telling you how much superior his i7 930 is to your six core Phenom.
I kid i kid + Show Spoiler +
|
Nah, I would have used this benchmark for that, since it has a 920, and a 930 is basically the same, but with a higher max multiplier. That one has results for more CPUs, but less games in total.
But it also has performance in other applications, so it's still nice. And the bench I posted earlier in the thread shows CPU performance for SC2 specifically.
|
Am I seeing that correctly? The X4 955 outperforming the X6 1090T?
Oh... lol... some of the colors look so similar. Nvm
|
Those colors make my eyes hurt. You have to go down the list systematically to see whats going on. They should've organized that chart differently.
At jingle below: Yes, it is .
|
No, they didn't use distinct enough of shades, that's an i7 980x coming in second place there.
The 955 is dead last or near the bottom on the tests.
Is it bad that nearly all of my bookmarks are some sort of benchmark results or articles related to computer parts?
|
lol idk depends on what you consider bad :p
and yeah it would be basicly for rts games (sc2, civ5, maybe bf3 when it comes out)
i guess i will overclock to 4ghz as i have a decent cooling system, and i've found a tutorial on how to do it exactly
|
On July 12 2011 11:35 Rachnar wrote: lol idk depends on what you consider bad :p
and yeah it would be basicly for rts games (sc2, civ5, maybe bf3 when it comes out)
i guess i will overclock to 4ghz as i have a decent cooling system, and i've found a tutorial on how to do it exactly
Did you find a tutorial, or someone's settings? Bear in mind, the method is more important than the specific settings, because unless your chips are from the same batch, mileage can and will vary significantly.
Hell, even within a batch it can vary a good deal.
I wouldn't shoot straight for 4Ghz, either, OC a small amount your first go at it, just to get more familiar with it, test for a couple hours stability, and see how your temps do, then see how high you can push stable without a VCore increase, test for longer stability, like overnight, and then work your way up from there in small increments.
You'll probably hit a wall where continuing to increase the clock isn't worth the increased voltage at some point. That's the part that really varies chip by chip.
And 4Ghz on a 1090T is probably a bit near extreme.
|
My system is worse than yours and I dont lag at all. Maybe you have to many bloated applications hogging resources. My build is below (your is way better than mine and i run on ultra)
AMD Phenom II x3 700e 2.4 GHz (OC'd to 2.8 GHz) Radeon HD 5670 512 MB @ stock settings Cheap 2GB DDR2 800 RAM 250 watt old PSU (pulled from a dell GX280)
So you really should have no issues. Maybe you should try re-installing your os. idk. :/
|
|
|
|