SixFootDuo
Supreme [H]ardness
- Joined
- Oct 5, 2004
- Messages
- 5,825
Intel 4770K @ 4.5GHz ~ 125 bootstrap / 36 Multi ~ 2250mhz DDR3 w/ 2 x XFX 290x @ stock
Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
Hmm, managed to pull 22686 points on a GTX 780, but but it looks like I'm a bit CPU limited. Link: http://www.3dmark.com/sd/2173893
In the Combined Test, there's enough of a CPU bottleneck that the GPU actually drops back to base (rather than boosted) 3D clocks. There literally wasn't enough GPU load for it to bother staying at 1200 MHz.
Confirmed by using K-boost to force the graphics card to maintain 1200MHz for the duration of all tests. Score and framerate didn't change, so it's a CPU bottleneck for sure.
We're running the same video driver, right down the the sub-version detected by 3DMark... and yet it detects yours as "approved" and mine as "unapproved"?Here we go, with the right test. lol Got 27827 http://www.3dmark.com/3dm/3355765
That's what happens when the graphics in a test are so weak that it runs at 200 FPS. CPU huffs and puffs while the graphics card puts up its feet and waits for the rest of the system to catch uplol wow your 780 barely beats my 7970......now thats funny...
That's what this test is doing, effectively... it's so light on graphics that it will run your CPU flat-out if you have any kind of decent graphics card.about time they come out with a test that pushes the cpu as well
Huh? What are you talking about?Dude. You can't overclock in 3dmark anymore? Looks like they changed the rules.
That's what this test is doing, effectively... it's so light on graphics that it will run your CPU flat-out if you have any kind of decent graphics card.
If I could get a Core i7 3770k for cheap I'd swap to that (since it would allow me to keep my current mobo + RAM). Would probably improve my score considerably.
__________________
80%? You might want to double-check your charts. Test #1 and #2 will have lower CPU usage, but your CPU should be at 100% in Test #3 no matter what, and very-nearly 100% in Test #4 (unless you have a weak graphics card holding things back).No doubt about it...first i seen my cpu pushed to 80% in any graphics test ever....that i5 is killing you
We're running the same video driver, right down the the sub-version detected by 3DMark... and yet it detects yours as "approved" and mine as "unapproved"?
Huh? What are you talking about?
Look at the results link from my first post in this thread, my CPU and GPU are both overclocked: http://www.3dmark.com/sd/2173893
I was just noticing that, in this particular test, my GPU clock doesn't seem to matter much because I'm almost totally CPU-bound.
But it seems I was confused (takes a guy with a spine to admit when he's confused or wrong or mistaken) and they aren't really talking about video card overclocking, I guess? It seems to be more about optimizations from cell phone manufacturers.So overclocking isn't allowed now?
Overclocking by manufacturers is allowed provided that it applies equally to all apps, all of the time. Overclocking optimizations that are selectively applied to our benchmarks are forbidden.
Beside the point, there were plenty of examples of overclocked components already in this thread. Was just pointing-out that fact, since it plainly shows overclocking is allowed.Wasn't actually addressing you or whatever it is you happen to be feverishly arguing about atm.
They're talking about application-specific optimizations. So yes, Android phones that clock-up when they detect 3DMark (and ONLY 3DMark) wouldn't be allowed.But it seems I was confused (takes a guy with a spine to admit when he's confused or wrong or mistaken) and they aren't really talking about video card overclocking, I guess? It seems to be more about optimizations from cell phone manufacturers.
Yeah, I was noticing that. Comparing our physics scores (and only our physics scores, since they're totally CPU dependent):edit: Physics Score 8244 lol yea your cpu bottlenecked bad
Core i5 2500k @ 4.5 GHz = 8244 points
Core i7 4770k @ 4.5 GHz = 11973 points
Core i5 2500k @ 4.5 GHz = 8244 points
Xeon X5670 @ 4.2GHz = 12386 points (5 or 6 year old cpu lol)
That's a 45% increase in performance at the same clockspeed... which is pretty nuts. The only thing I can think of that would put such a massive gap between these two chips is if 3DMark seriously loves HyperThreading.
__________________
That is good for an older cpu but its not that old as even if you bought it on launch day its not quite 4.5 years old.fixed it for you lol fuck it be a while still...before in can upgrade
Well, just keep in mind, your performance depends on how threaded a given application is (your CPU with 12 threads is just barely ahead of a more-modern chip with only 8 threads in this bench, for example).fixed it for you lol fuck it be a while still...before in can upgrade
That is good for an older cpu but its not that old as even if you bought it on launch day its not quite 4.5 years old.
During the last two tests? It was pretty much pegged at 100% all the way through.what was you max cpu usage during the test...was just curious..mine never went over 85% at most....which in it self is pretty good...bf4 as an example uses like 20% and people actually bitch about it being cpu bottle necked...dam what the hell they running?
During the last two tests? It was pretty much pegged at 100% all the way through.
Also, there's a key distinction to be aware of when you're talking about CPU usage on multiple cores. Lets say you have a quad core processor, and a game only spawns 2 threads, and those 2 threads use every last scrap of CPU time they can get their hands on...
End result? You're CPU limited even though the CPU graph only shows 50% usage.
we can only hope they make games from this point forward using at least 8 threads...or even 12 for me
No OS currently in existence can run one thread on multiple cores. Applications have to be multi-threaded from the get-go.Or a Windows OS that scales programs up to the amount of available threads. Win 8.1 is a lot better than 7, but it still needs work.
No OS currently in existence can run one thread on multiple cores. Applications have to be multi-threaded from the get-go.
Even if you could force each instruction from a single thread to execute on a different core, you'd very-likely make the application run slower. You'd end up continually running into a condition where the instruction on one core needs the output of an instruction on another core, so everything would have to pause until the core handling the needed instruction finishes.
Well, the thing is, it's a fundamental problem with how the simplest bits of code execute. Here's an example:Doesn't mean that it has to stay this way. I think other ways of multithreading is going to be the future as it is obvious that better CPUs aren't the answer.
Well, the thing is, it's a fundamental problem with how the simplest bits of code execute. Here's an example:
1. I want to know the answer to "X + Y"
2. The answer to X is "A + B"
3. The answer to Y is "C + D"
4. I cannot execute "X + Y" until I've already executed "A + B" and "C + D"
Implications of this example scenario?:
- #2 and #3 can be run in parallel (at the same time) because they have no dependency on one another. These operations can be threaded.
- #4 requires the data from #2 and #3 in order to execute. There is absolutely no way to run #4 at the same time as #2 and #3, because it requires data that has not yet come into existence. Spawning an additional thread for #4 would be pointless, because it has to wait on #2 and #3 anyway.
The only way to work-around this is NOT using operations that have dependencies on other operations, but that's not always possible, and it's still up to the developer programming the application to implement (which circles right back around to what I said originally: applications have to be threaded from the get-go).
Even more issues with getting one thread to run across multiple cores can be found here: http://arstechnica.com/uncategorized/2006/07/7263-2/
80%? You might want to double-check your charts. Test #1 and #2 will have lower CPU usage, but your CPU should be at 100% in Test #3 no matter what, and very-nearly 100% in Test #4 (unless you have a weak graphics card holding things back).
I don't have a weak graphics card holding anything back, so I'm running straight into a CPU bottleneck in test #4. CPU usage at 100%, GPU usage so low that the card doesn't even bother boosting.
Yeah, I'm not all that worried about my "bottleneck" either, to be honest. Even in Test #4, the benchmark ran at 80+ FPS the whole way through. More than enough for my 60Hz monitors.is the card weak...in comparison to yours yea...but its not holding back any games i play at 1080p.....well except crysis 3...but i doubt even yours holds 60 fps on that game