PDA

View Full Version : 36 Benchmarks, results from 140985 to 160041



Sverre
03-18-2017, 04:54 AM
Never done this test before. Left it on Infinite and watched some movies. After coming back I went over the results and found quite large differences (se log under). I can understand some differences, but a difference of almost 20000 seems large. And the 3 first tests are also the 3 highest, almost 7000 higher than number 4, what happened here?

---

Sat Mar 18 2017 02:45:53 Image Editing: 180874 Time: 28.13 Encoding: 249369 Time: 38.449 OpenCL: 83627K Samples/sec: 2150 Heavy Multitasking: 193074 Time: 50.654 System Score: 160010
Sat Mar 18 2017 02:49:57 Image Editing: 177821 Time: 28.613 Encoding: 252110 Time: 38.031 OpenCL: 83627K Samples/sec: 2153 Heavy Multitasking: 193510 Time: 50.54 System Score: 160041
Sat Mar 18 2017 02:54:05 Image Editing: 180694 Time: 28.158 Encoding: 257679 Time: 37.209 OpenCL: 83627K Samples/sec: 2065 Heavy Multitasking: 178431 Time: 54.811 System Score: 158382
Sat Mar 18 2017 02:58:32 Image Editing: 134159 Time: 37.925 Encoding: 243399 Time: 39.392 OpenCL: 63941K Samples/sec: 1516 Heavy Multitasking: 174185 Time: 56.147 System Score: 141132
Sat Mar 18 2017 03:02:54 Image Editing: 135419 Time: 37.572 Encoding: 239664 Time: 40.006 OpenCL: 73784K Samples/sec: 1784 Heavy Multitasking: 174104 Time: 56.173 System Score: 140985
Sat Mar 18 2017 03:07:09 Image Editing: 142788 Time: 35.633 Encoding: 250732 Time: 38.24 OpenCL: 73784K Samples/sec: 1790 Heavy Multitasking: 185473 Time: 52.73 System Score: 148437
Sat Mar 18 2017 03:11:29 Image Editing: 144418 Time: 35.231 Encoding: 244261 Time: 39.253 OpenCL: 63941K Samples/sec: 1493 Heavy Multitasking: 171898 Time: 56.894 System Score: 143341
Sat Mar 18 2017 03:15:55 Image Editing: 131837 Time: 38.593 Encoding: 247087 Time: 38.804 OpenCL: 67222K Samples/sec: 1612 Heavy Multitasking: 176639 Time: 55.367 System Score: 142251
Sat Mar 18 2017 03:20:17 Image Editing: 140922 Time: 36.105 Encoding: 248464 Time: 38.589 OpenCL: 70503K Samples/sec: 1752 Heavy Multitasking: 177148 Time: 55.208 System Score: 145158
Sat Mar 18 2017 03:24:37 Image Editing: 146009 Time: 34.847 Encoding: 251126 Time: 38.18 OpenCL: 63941K Samples/sec: 1523 Heavy Multitasking: 183166 Time: 53.394 System Score: 148272
Sat Mar 18 2017 03:29:01 Image Editing: 137621 Time: 36.971 Encoding: 245745 Time: 39.016 OpenCL: 63941K Samples/sec: 1477 Heavy Multitasking: 179663 Time: 54.435 System Score: 143954
Sat Mar 18 2017 03:33:19 Image Editing: 136060 Time: 37.395 Encoding: 253617 Time: 37.805 OpenCL: 73784K Samples/sec: 1841 Heavy Multitasking: 180123 Time: 54.296 System Score: 146139
Sat Mar 18 2017 03:37:37 Image Editing: 148221 Time: 34.327 Encoding: 243066 Time: 39.446 OpenCL: 73784K Samples/sec: 1826 Heavy Multitasking: 173900 Time: 56.239 System Score: 144985
Sat Mar 18 2017 03:41:50 Image Editing: 148074 Time: 34.361 Encoding: 256521 Time: 37.377 OpenCL: 73784K Samples/sec: 1804 Heavy Multitasking: 183383 Time: 53.331 System Score: 150683
Sat Mar 18 2017 03:46:12 Image Editing: 139404 Time: 36.498 Encoding: 244541 Time: 39.208 OpenCL: 73784K Samples/sec: 1834 Heavy Multitasking: 171817 Time: 56.921 System Score: 142629
Sat Mar 18 2017 03:50:30 Image Editing: 143980 Time: 35.338 Encoding: 241986 Time: 39.622 OpenCL: 73784K Samples/sec: 1833 Heavy Multitasking: 174471 Time: 56.055 System Score: 143798
Sat Mar 18 2017 03:54:45 Image Editing: 142059 Time: 35.816 Encoding: 255027 Time: 37.596 OpenCL: 73784K Samples/sec: 1834 Heavy Multitasking: 184118 Time: 53.118 System Score: 148990
Sat Mar 18 2017 03:59:05 Image Editing: 147666 Time: 34.456 Encoding: 238270 Time: 40.24 OpenCL: 73784K Samples/sec: 1843 Heavy Multitasking: 170537 Time: 57.348 System Score: 142807
Sat Mar 18 2017 04:03:19 Image Editing: 148528 Time: 34.256 Encoding: 253664 Time: 37.798 OpenCL: 73784K Samples/sec: 1848 Heavy Multitasking: 183279 Time: 53.361 System Score: 150056
Sat Mar 18 2017 04:07:41 Image Editing: 144086 Time: 35.312 Encoding: 245455 Time: 39.062 OpenCL: 77065K Samples/sec: 1870 Heavy Multitasking: 182136 Time: 53.696 System Score: 146772
Sat Mar 18 2017 04:12:00 Image Editing: 143219 Time: 35.526 Encoding: 246890 Time: 38.835 OpenCL: 73784K Samples/sec: 1798 Heavy Multitasking: 174524 Time: 56.038 System Score: 144847
Sat Mar 18 2017 04:16:12 Image Editing: 148407 Time: 34.284 Encoding: 258876 Time: 37.037 OpenCL: 73784K Samples/sec: 1833 Heavy Multitasking: 184337 Time: 53.055 System Score: 151594
Sat Mar 18 2017 04:20:35 Image Editing: 140188 Time: 36.294 Encoding: 241901 Time: 39.636 OpenCL: 73784K Samples/sec: 1829 Heavy Multitasking: 182571 Time: 53.568 System Score: 144854
Sat Mar 18 2017 04:24:54 Image Editing: 145110 Time: 35.063 Encoding: 244548 Time: 39.207 OpenCL: 73784K Samples/sec: 1830 Heavy Multitasking: 173063 Time: 56.511 System Score: 144369
Sat Mar 18 2017 04:29:15 Image Editing: 139160 Time: 36.562 Encoding: 239574 Time: 40.021 OpenCL: 73784K Samples/sec: 1806 Heavy Multitasking: 174877 Time: 55.925 System Score: 142091
Sat Mar 18 2017 04:33:41 Image Editing: 136711 Time: 37.217 Encoding: 232267 Time: 41.28 OpenCL: 77065K Samples/sec: 1872 Heavy Multitasking: 181037 Time: 54.022 System Score: 141357
Sat Mar 18 2017 04:38:00 Image Editing: 149783 Time: 33.969 Encoding: 248059 Time: 38.652 OpenCL: 77065K Samples/sec: 1872 Heavy Multitasking: 184702 Time: 52.95 System Score: 149489
Sat Mar 18 2017 04:42:17 Image Editing: 148706 Time: 34.215 Encoding: 242145 Time: 39.596 OpenCL: 73784K Samples/sec: 1833 Heavy Multitasking: 175180 Time: 55.828 System Score: 145196
Sat Mar 18 2017 04:46:32 Image Editing: 147162 Time: 34.574 Encoding: 249863 Time: 38.373 OpenCL: 73784K Samples/sec: 1828 Heavy Multitasking: 183324 Time: 53.348 System Score: 148776
Sat Mar 18 2017 04:50:56 Image Editing: 137845 Time: 36.911 Encoding: 242017 Time: 39.617 OpenCL: 73784K Samples/sec: 1804 Heavy Multitasking: 180339 Time: 54.231 System Score: 143739
Sat Mar 18 2017 04:55:17 Image Editing: 148554 Time: 34.25 Encoding: 243307 Time: 39.407 OpenCL: 73784K Samples/sec: 1831 Heavy Multitasking: 183074 Time: 53.421 System Score: 147422
Sat Mar 18 2017 04:59:36 Image Editing: 148676 Time: 34.222 Encoding: 243473 Time: 39.38 OpenCL: 73784K Samples/sec: 1802 Heavy Multitasking: 172246 Time: 56.779 System Score: 144787
Sat Mar 18 2017 05:03:56 Image Editing: 139393 Time: 36.501 Encoding: 241974 Time: 39.624 OpenCL: 73784K Samples/sec: 1836 Heavy Multitasking: 173564 Time: 56.348 System Score: 142421
Sat Mar 18 2017 05:08:15 Image Editing: 146232 Time: 34.794 Encoding: 238311 Time: 40.233 OpenCL: 73784K Samples/sec: 1838 Heavy Multitasking: 178780 Time: 54.704 System Score: 144519
Sat Mar 18 2017 05:12:33 Image Editing: 140145 Time: 36.305 Encoding: 253302 Time: 37.852 OpenCL: 73784K Samples/sec: 1814 Heavy Multitasking: 176387 Time: 55.446 System Score: 146147
Sat Mar 18 2017 05:16:58 Image Editing: 134248 Time: 37.9 Encoding: 255054 Time: 37.592 OpenCL: 73784K Samples/sec: 1830 Heavy Multitasking: 172965 Time: 56.543 System Score: 144255

Arne Saknussemm
03-18-2017, 08:35 AM
Nothing happened here...move along folks ;)

Normal variation for benchmark runs....especially since your PC will have been doing varied tasks in the background...

The initial higher scores are just normal benchmarking...optimum runs after a fresh start...optimum run after second or third run

Other possibilities...temperature saturation of cooling...and aliens :D

Sverre
03-18-2017, 12:05 PM
Nothing happened here...move along folks ;)

Normal variation for benchmark runs....especially since your PC will have been doing varied tasks in the background...

The initial higher scores are just normal benchmarking...optimum runs after a fresh start...optimum run after second or third run

Other possibilities...temperature saturation of cooling...and aliens :D

Ok, just thought the difference was a bit high. Did a test over night and got 160319 on the first run, does this mean that there's no point in running it 2-3 times in a row?

Plenty of aliens up here in Norway! ;)

Why alchemist? Doesn't sound like it has anything to do with overclocking, or computers in general?

Can't log on to RealBench, isn't my account there the same as here?

Nodens
03-18-2017, 12:29 PM
Like Arne said, your system is doing stuff in the background. That chart of yours just shows that the background activity of your system creates load spikes every now and then. All perfectly normal.

Regarding the logging in, ASUS changed the networking part on their end to SSL (a good thing really) so versions prior to the current beta can't authenticate users or upload.

Arne Saknussemm
03-18-2017, 12:41 PM
does this mean that there's no point in running it 2-3 times in a row?

Actually I think second run is still better...see tips and tweaks thread https://rog.asus.com/forum/showthread.php?35793-ROG-Benchmarking-Tweaks-amp-Tips

Zka17
03-18-2017, 01:33 PM
It may be an optimization thing but second run is always the best for me! It was on all systems tried... (then I never run it more than 3 times without restarting...)

Nodens
03-18-2017, 01:40 PM
One word, caching :)

Arne Saknussemm
03-18-2017, 01:59 PM
http://digitalbloggers.com/christinegilmartin/files/2014/08/light-bulb-moment1st.jpg

Zka17
03-18-2017, 02:03 PM
One word, caching :) - good to know! :cool:

Now just one question: Is it enough to run the first test (video editing - that is showing the most differences between first and second runs) or better to run all the benchmark for that caching to happen?

Nodens
03-18-2017, 02:20 PM
Running just the encoding yours cache benefits will be isolated to that test alone. For example the images used in the GIMP test will not be cached in RAM as they would be read for the first time..

Zka17
03-18-2017, 02:31 PM
OK... I was hoping that I found a shortcut... :(

Sverre
03-18-2017, 02:57 PM
Over 191.000 following the suggestions, not too bad, that's #11 on the list with the same setup as me (although I wonder about some of the setups: 6900K at 1200MHz, GTX 980ti's, even a 980 (no ti) and a 1070, although some have higher frequency on the RAM). BTW, where are the frequencies for GPU Core and Memory, aren't those some of the most important for this test (I'm at 2151/1251)?

Zka17
03-18-2017, 03:02 PM
I found that the RB is taking the actual numbers your CPU is running... if you have Windows Power on something else than High, the CPU will downclock... - that's what could be the explanation for those high scores/low clock.

Sverre
03-18-2017, 03:02 PM
Over 191.000 following the suggestions, not too bad, that's #11 on the list with the same setup as me (although I wonder about some of the setups: 6900K at 1200MHz, GTX 980ti's, even a 980 (no ti) and a 1070, although some have higher frequency on the RAM). BTW, where are the frequencies for GPU Core and Memory, aren't those some of the most important for this test (I'm at 2151/1251)?

That was, over 191.000 on the 3 tests I did, but that was on 2.44, 2.54b was terrible (140.000'ish, several tests).

I did not run the suggested bat-file, not sure if it's ok to run it in Windows 10 (64 pro)?

Arne Saknussemm
03-18-2017, 03:13 PM
The scoring formula has changed between versions...no point comparing between versions...

Sverre
03-18-2017, 04:23 PM
The scoring formula has changed between versions...no point comparing between versions...


Then, maybe, those results should be removed, or at least we should be given the opportunity to decide which version we want to compare with (default should be the version we run ourselves).

Nodens
03-18-2017, 06:12 PM
Formula tweaking is not finished yet. Old versions will not be able to upload anyhow and results from 2.53b will be removed.

Sverre
03-18-2017, 08:50 PM
Formula tweaking is not finished yet. Old versions will not be able to upload anyhow and results from 2.53b will be removed.

I saved my 2.44, closed it, started 2.54b, nad uploaded the 2.44 result. No problem. Hope that isn't a work around.

Sverre
03-18-2017, 08:53 PM
OC'ed my card to 2190/1425 and it runs good in RealBench, but it worries me that this is a frequency I can only dream about in Valley (2151) and Fire Strike (2124).

OC'ed CPU (5960x) to 4.6GHz (1.35V) without problem and running it on 4.7GHz (1.399V) now. Also a frequency that isn't very stable on other benchmarks.

Arne Saknussemm
03-18-2017, 09:18 PM
There are a whole lot of you out there who seem incapable of grasping the point/value/raison d'etre of this benchmark/stresstest.

Despite the name being a rather big clue!

Yes, you can run some legacy benchmark like P95 and knock over your CPU...maybe at lower voltage than RB...that is not the point. The point of this test, is to stress ALL PC subsystems at the same time with a combination of REAL WORLD programs...thus giving you a much more holistic way of assuring stability in everyday computing. With the advantage of not drawing degrading levels of current through the CPU nor breaking the thermal limits of the CPU, nor the power limits.

It is also offering you choice...if you prefer a stress test that only stresses one thing and promotes electron tunneling at the same time...stick with it.

Sverre
03-18-2017, 10:43 PM
There are a whole lot of you out there who seem incapable of grasping the point/value/raison d'etre of this benchmark/stresstest.

Despite the name being a rather big clue!

Yes, you can run some legacy benchmark like P95 and knock over your CPU...maybe at lower voltage than RB...that is not the point. The point of this test, is to stress ALL PC subsystems at the same time with a combination of REAL WORLD programs...thus giving you a much more holistic way of assuring stability in everyday computing. With the advantage of not drawing degrading levels of current through the CPU nor breaking the thermal limits of the CPU, nor the power limits.

It is also offering you choice...if you prefer a stress test that only stresses one thing and promotes electron tunneling at the same time...stick with it.

Why so angry? Isn't it a valid question? Reading different forums you get different answers, from people I would expects know what they are talking about (checking their credentials; not easy with a name like yours, but you seem to know what you're talking about when I check out your posts here, on this forum). What I find most interesting is the different opinions on the validation of the data output from the CPU; corruption of data that will render it, sooner or later, worthless. I would imagine Windows (10 64 pro) is checking the validity of data before it saves it, and that, that problem is actually taken care of?

I'm OC'ing as follows, now:
MB - Rampage V Extreme USB 3.1
CPU - Intel Core i7 Extreme Edition 5960X: 4x4,7GHz + 4x4,6GHz (1,399V)
GPU - ROG Strix GTX 1080 O8G Gaming: Core 2,190GHz Memory 1,426GHz
RAM - Corsair Dominator Platinum CMD16GX4M4B3200C15: 3,2GHz (1,6GHz)

Will do a stress test and see how it works out. Any suggestions on the length of the test, and if it should be done in Normal or Diagnostic Startup?

I have not played around with the RAM, which I don't know much about. Would like to find a page where there are detailed suggestions for OC'ing Rampage V, anyone know of any? Not talking about the usual 5 to 6 step basic OC-guide, more like - If you do this, and it crashes, do that, and maybe that.

Zka17
03-19-2017, 03:30 AM
If you want to stress-test, I don't see the point to boot in diagnostic mode... you're not after scores, just stability...
Regarding OC'ing, check out Raja's guides over the Rampage section... (usually they are stickies)

Arne Saknussemm
03-19-2017, 07:49 AM
Why so angry? Isn't it a valid question?

:rolleyes: Ascribing emotions to written posts is always an interpretive event...likely to say more about the reader than the writer :p

Forgive me but I don't see any questions in this post


OC'ed my card to 2190/1425 and it runs good in RealBench, but it worries me that this is a frequency I can only dream about in Valley (2151) and Fire Strike (2124).

OC'ed CPU (5960x) to 4.6GHz (1.35V) without problem and running it on 4.7GHz (1.399V) now. Also a frequency that isn't very stable on other benchmarks.

Just statements that seem to indicate a misunderstanding of stress testing...I am simply trying to orient you to the correct view/use of this Benchmark/Stresstest.

I tried to reply to you in your other thread on some of the points you repeat here https://rog.asus.com/forum/showthread.php?91765-Overclocking-vs-stability

I simply don't know where you are getting some ideas from. It simply seems like some base assumptions or ideas are wrong and then of course conclusions will be wrong.

In general this whole stress testing thing and in particular that P95 is the go-to stability test, is nothing more than an internet meme.

Stability simply means the system is stable doing what you want it to do. If you want P95 stability then you test with P95....but this does not mean stability even for gaming, as we have seen here a thousand times over since P95 stresses only a single part of the PC at worst or one and a bit systems if you chose the right tests.

That is why RB came about....as a more realworld stress test for gaming/encoding/multimedia stability...the most stressful things that a regular computer user does with their machine. It is enormously useful in this regard and though it cannot cover all use cases it is by far the best test, for regular PC users, out there.

As far as Overclocking and benchmarking go...the point is, there is no ultimate stress test. There is no stress test that will give you numbers to run for all different use cases or benchmarks. If you bench Firestrike you have to find the OC for Firestrike...and even this differs between Firestrike and Firestrike Extreme and Firestrike Ultra. For Valley you will have to find an entirely different OC etc etc etc


Any suggestions on the length of the test, and if it should be done in Normal or Diagnostic Startup?

If you are testing for system stability for daily use it should be done in your normal daily OS environment...with none of the tweaks applied...such as setting windows for max performance etc...just your normal set up. Benchmarking for maximum score...then apply all the tweaks you want.

Length is really up to you...another internet meme is 8 hours for RB at the moment....given my past experiences with setting up OCs I'm happy with one or two hours.

There isn't really any guide of that type...there are of course no recipes for this stuff. Since the components are always different it is pretty much up to you to find out what works best for your system. Raja's guides are certainly the starting point...some users post up their settings https://rog.asus.com/forum/showthread.php?51063-The-Haswell-E-Overclocking-Thread-and-Overclocking-Guide&p=543169&viewfull=1#post543169...again as orientation not recipe...voltages will always have to your own...

Start a thread...no shortage of help here :)

JustinThyme
03-19-2017, 03:51 PM
Arne angry? LMAO. One of the most level headed posters on this forum. Me on the other hand, I often come across less than understood but that's who I am.

I have to agree 100% with Arne's assessment here. Ive run prime 95 in the past for 24 hours to prove stability only to have my machine crash once using it for something other than pounding the CPU repeatedly with the same instruction set that does nothing more than heat it up. Prime 95s only purpose in my book may be to test you cooling, that about as far as it goes.

Length of stress testing is subjective, mostly subjective to epeen size. If nothing else you should have seen in your first post here that on a cold boot you may run for 1000 hours uninterrupted then not make it 5 minutes later on. This is a tool like any other stress test. Actual stability will vary from one machine and configuration to the next. When over clocking I use the bench first. If it doesnt pass the bench I dial it down or tweak until it does then move onto the stress test for an hour or so. Once it makes it past the hour I call it good. If its not Ill find out later and may have to dial it back a notch.

OCing is not an exact science. It has high levels of probabilites but nothing is exact. Chips and memory modules vary, voltage control from one MOBO (same model even) to the next will vary, OS installed, background apps running etc all play a part.

Bottom line, this tool is just to help you get the best out of what you have and uses a set of instructions that closely resembles actual use instead of those designed to see how well your CPU can be used as a space heater.

Sverre
03-19-2017, 05:33 PM
Arne angry? LMAO. One of the most level headed posters on this forum. Me on the other hand, I often come across less than understood but that's who I am.

I have to agree 100% with Arne's assessment here. Ive run prime 95 in the past for 24 hours to prove stability only to have my machine crash once using it for something other than pounding the CPU repeatedly with the same instruction set that does nothing more than heat it up. Prime 95s only purpose in my book may be to test you cooling, that about as far as it goes.

Length of stress testing is subjective, mostly subjective to epeen size. If nothing else you should have seen in your first post here that on a cold boot you may run for 1000 hours uninterrupted then not make it 5 minutes later on. This is a tool like any other stress test. Actual stability will vary from one machine and configuration to the next. When over clocking I use the bench first. If it doesnt pass the bench I dial it down or tweak until it does then move onto the stress test for an hour or so. Once it makes it past the hour I call it good. If its not Ill find out later and may have to dial it back a notch.

OCing is not an exact science. It has high levels of probabilites but nothing is exact. Chips and memory modules vary, voltage control from one MOBO (same model even) to the next will vary, OS installed, background apps running etc all play a part.

Bottom line, this tool is just to help you get the best out of what you have and uses a set of instructions that closely resembles actual use instead of those designed to see how well your CPU can be used as a space heater.

It took exactly 1 second before I crashed in Albion Online, a game that doesn't require anything from the graphics card (I've played 2 hours now, and the temp is still 33C, which is pretty much the same as it does on almost no load), twice; and that was after I clicked on my character, to enter the world. So, no, this test isn't correct, either.

He might not have been angry, but the way he replied looked to me as angry, maybe even hostile; probably just bad choice of words.

I do know a little bit about OC'ing, I have played around with it for 4 or 5 years, but I've never found a real good guide for it, it's mostly the usual 5 or 6 steps. And, of course, I've done stability testing.

I just reset it to the OC that works with Valley, that works with even the heaviest games (at least I haven't crashed yet).

Arne Saknussemm
03-19-2017, 06:11 PM
One of the most level headed posters on this forum

Man...I thought the hood in my avatar hid that...
63308

Menthol
03-19-2017, 07:03 PM
Sometimes it takes a different level, be prepared is what I say

63309

Nodens
03-20-2017, 12:14 AM
It took exactly 1 second before I crashed in Albion Online, a game that doesn't require anything from the graphics card (I've played 2 hours now, and the temp is still 33C, which is pretty much the same as it does on almost no load), twice; and that was after I clicked on my character, to enter the world. So, no, this test isn't correct, either.


Have in mind that crashes can happen for reasons not related to OC. Driver issues, game engine bugs, OS corruption etc etc.

That said if you can pass RB ST it's 99.99% certain that you won't crash in any regular user application. The exception to that (the 0.01%) is GPU overclocking and games. Because the part of the ST that stresses the GPU is actually Compute load and not regular 3D load, there may be cases where instability is not detected on GPU overclocking. If you are overclocking your GPU you should ideally stress test the GPU with something else after you pass RB ST.

This is what RB ST is all about. It is possible to pass the ST and fail on some hardcore test that calculates whatever. All a stress test does is prove you can pass that particular test. Nothing more. RB's ST loads all available subsystems at the same time and in an asynchronous way creating load spikes and drips on the different subsystems. It is the closest thing to testing everything. Prime95 on the other hand calculates an FFT algorithm that can fit into modern CPU's cache entirely again and again and again. This has nothing to do with a normal load that can happen on your system. You can pass that test for hours only to have your system crash when bulk transfers happen on the PCIe bus for example. So you should ask yourself what you want to do with your system. If you want to use it for medical research and fold proteins, then stress test with that. If you want to calculate prime numbers, go ahead and test your ability to do so. Etc etc etc.
Same concept is behind the benchmark as well. It allows you to see how fast your PC is on a specific real world task. See real performance gains. In contrast to other benchmarks which also run a synthetic algorithm and give you a score or a FLOPS measurement that means nothing to a user. Yeah my CPU can go 83131263 GFLOPS..what does that mean? How does that translate to real, measurable performance? How much faster will my system actually be if I upgrade? This is what RB aims to do. You can see it when your OC shaves 20 secs off the encoding. It shows you what you are actually gaining or what you could gain by upgrading to a new platform and CPU, etc etc

Arne Saknussemm
03-24-2017, 07:43 AM
Sometimes it takes a different level, be prepared is what I say

63309

Is he staring or is he stairing?
Anywho...that haircut works on so many levels...a real step up from the average haircut.
Maybe he's Roger Bannister's step-son?
Favourite music...dubstep?
Maybe he asked the barber to step aside?...or maybe..."I need a haircut but I'm in a hurry so step on it!"

:rolleyes:

Nodens
03-24-2017, 01:26 PM
Rofl, so many bad puns in the same post ahahahahaha! :P