| View previous topic :: View next topic |
| Author |
Message |
ggrobot Elite Member

Joined: 28 May 2004 Posts: 53617
|
Posted: Mon Aug 10, 2015 9:11 pm Post subject: Win10 vs. Win8.1 vs. Win7 Performance [40070] |
|
|
The chaps over at TechSpot have pitted Windows 10 vs. Windows 8.1 vs. Windows 7 in a three-way deathmatch to decide which OS is the performance king. Here's an excerpt:
All three operating systems were tested using a fresh install with all the updates applied and then our test software installed. The exact sam
Read more...
Source: GGMania headlines
GGMania.com - Daily Gaming and Tech news |
|
| Back to top |
|
 |
Csimbi Elite Member

Joined: 05 Mar 2010 Posts: 5358 Location: The bright side of the dark side
|
Posted: Mon Aug 10, 2015 10:10 pm Post subject: |
|
|
Hardly surprising that Win 7 is faster.
What surprises me is that the difference is not that big.
To me, the most meaningful test is WinRAR.
That's what I use to tune my system - the memory clocks and timings to be specific.
Setting better memory clocks and timings can give you quite a boost and it's easily visible in the WinRAR benchmark in seconds. |
|
| Back to top |
|
 |
Tom Elite Member

Joined: 07 Jun 2004 Posts: 4289
|
Posted: Mon Aug 10, 2015 10:27 pm Post subject: |
|
|
Same OS different skin, not surprising at all.
Who deals with WinRAR files? Oh ya, those into pirating and downloading pirated titles. Yes, WinRAR is important. And look at that huge time difference between all 3. Wheewwww!!! |
|
| Back to top |
|
 |
zarlin Junior Member

Joined: 26 Jun 2010 Posts: 180
|
Posted: Mon Aug 10, 2015 10:37 pm Post subject: |
|
|
| Tom wrote: | Same OS different skin, not surprising at all.
Who deals with WinRAR files? Oh ya, those into pirating and downloading pirated titles. Yes, WinRAR is important. And look at that huge time difference between all 3. Wheewwww!!! |
Well it looks like you know it  |
|
| Back to top |
|
 |
gx-x Elite Member

Joined: 02 Jul 2007 Posts: 2655
|
Posted: Tue Aug 11, 2015 12:06 am Post subject: |
|
|
| Tom wrote: | Same OS different skin, not surprising at all.
Who deals with WinRAR files? Oh ya, those into pirating and downloading pirated titles. Yes, WinRAR is important. And look at that huge time difference between all 3. Wheewwww!!! |
you are missing the point...
WinRAR is just the quickest way to see how changes and what changes affected the system and by how much. Though with DDR3 today, moving away from SPD latencies is limited at best and changes in WinRAR speed are minimal if any at all. CPU speed is going to affect it way more.
As for it's use, well, that's up to you. I use it for a lot of other things other than unpacking stuff from internet. |
|
| Back to top |
|
 |
Tom Elite Member

Joined: 07 Jun 2004 Posts: 4289
|
Posted: Tue Aug 11, 2015 2:37 am Post subject: |
|
|
Yes, I know. The other sentence is just sarcasm. note the Wheewwww
 |
|
| Back to top |
|
 |
Csimbi Elite Member

Joined: 05 Mar 2010 Posts: 5358 Location: The bright side of the dark side
|
Posted: Tue Aug 11, 2015 1:29 pm Post subject: |
|
|
| gx-x wrote: | | Tom wrote: | Same OS different skin, not surprising at all.
Who deals with WinRAR files? Oh ya, those into pirating and downloading pirated titles. Yes, WinRAR is important. And look at that huge time difference between all 3. Wheewwww!!! |
you are missing the point...
WinRAR is just the quickest way to see how changes and what changes affected the system and by how much. Though with DDR3 today, moving away from SPD latencies is limited at best and changes in WinRAR speed are minimal if any at all. CPU speed is going to affect it way more.
As for it's use, well, that's up to you. I use it for a lot of other things other than unpacking stuff from internet. |
@Tom
Dunno, I have archived all my junk with RAR ever since the DOS days.
Private stuff, too. It's the best all-around archiver. Nowadays there are better compressors, but they are not that much better...
@gx-x
Yep, that's the point. No need for any other benchmark tool.
Downclocking the memory allows for lower latencies.
Combine those two properly and you'll be surprised that your 1600MHz memory outperforms the 2100MHz one!
I don't know about CPUs becoming the bottlenecks. I'm thinking the next bottlenext is the I/O between the memory and the CPU as well as the memory itself.
Why?
While CPUs don't get much faster (we've been around 4Ghz for four years now), they do get more cores. More cores mean more parallel processing -> more CPUs will need to be fed from the same memory (!)
That will put quite a load on both memory and the bus. |
|
| Back to top |
|
 |
Tom Elite Member

Joined: 07 Jun 2004 Posts: 4289
|
Posted: Tue Aug 11, 2015 3:03 pm Post subject: |
|
|
| With the cost of hard drives and memory these days, who needs to compress? Waste of time and a more serious situation if corruption occurs. If your archive is affected you are fucked. In my case, I could lose a bunch of files but not an archive holding hundreds or thousands of files. I got a bunch of TB's in a NAS configuration. |
|
| Back to top |
|
 |
gx-x Elite Member

Joined: 02 Jul 2007 Posts: 2655
|
Posted: Tue Aug 11, 2015 3:10 pm Post subject: |
|
|
Well actually, the archiver's tool speed also depends on algorithm and engine itself. 7zip is actually often faster than winRAR, for example.
as for memory, stepping down a notch will yield you what SPD will itself tell you. Usually, for example, if you have 1600MHz 9-9-9-24 memory, going down to 1333 will set it to 8-8-8-22. And yes, it's a better option, but differences are not that big in real operation, even in synthetic benchmarks they are not that big. Like I said, CPU will matter more, but also, I don't think I would notice weaker CPU in tasks like using the browser and listening to music in the same time. By this I mean if you were to give me [email protected] over my current one @3.8, or give i5 @4.7GHz, or [email protected] things don't matter for mundane tasks.
As shown in the review, on a nicely balanced system, all results are pretty much the same, confirming what I wrote couple of weeks back, that, you can sit back, relax and not care about win10 until they sort out some things and until DX12 games (if you play games) start coming out. |
|
| Back to top |
|
 |
Csimbi Elite Member

Joined: 05 Mar 2010 Posts: 5358 Location: The bright side of the dark side
|
Posted: Tue Aug 11, 2015 8:32 pm Post subject: |
|
|
| Tom wrote: | | With the cost of hard drives and memory these days, who needs to compress? Waste of time and a more serious situation if corruption occurs. If your archive is affected you are fucked. In my case, I could lose a bunch of files but not an archive holding hundreds or thousands of files. I got a bunch of TB's in a NAS configuration. |
Cost of hard drives are still high.
That's why I add recovery record.
I am actually downclocking my 2133 Mhz modules to 1666Mhz and I can get 7-7-7-24. I can tell you, you can feel that Windows is a bit more snappy. I have a 2600K behind it though. I bet it would not make a difference on those i5s and i3s (no offense).
I could downclock it to 800Mhz with the same timing and change 2T to 1T, but the speed did not change. Kinda logical.
We will see. I read about the 'improvements' Skylake won't be bringing despite my expectation, so I just put in an order for a 4790K. |
|
| Back to top |
|
 |
gx-x Elite Member

Joined: 02 Jul 2007 Posts: 2655
|
Posted: Tue Aug 11, 2015 9:01 pm Post subject: |
|
|
I think it's placebo because you want to see it. Reviews and extensive testing has been done and performance differences on high end CPU's and next to non-existent. You'd even hardly notice i5 @4.4 over i7 @4.4, let alone 0.5% improvement in the memory bandwidth department.
Also, no offense meant. In using firefox, chrome, watching movies, doing regular "desktop" crap I didn't notice any improvement when I switched from i3 2100 that was fixed @3.1GHz to intel i5 that is now @3.8GHz. I did however notice it in games of course, photoshop, aftereffects and some other apps, but apps, that don't even bother to utilize the cpu over 15-20% have nothing to gain. It's like comparing a McLaren P1 and Lada Niva on red light. Sure, one looks better but both are doing 0 mph. :d
PS. Save your money on 4790K. It's not worth it. you need new mobo too and all that for on average ~5% faster CPU? I get it, you are itching, but there just isn't anything that much better than Sandy I7. Unless you are willing to go 6 cores route...wait for 8 core i7 at normal price. |
|
| Back to top |
|
 |
zarlin Junior Member

Joined: 26 Jun 2010 Posts: 180
|
Posted: Tue Aug 11, 2015 10:07 pm Post subject: |
|
|
| Tom wrote: | | With the cost of hard drives and memory these days, who needs to compress? Waste of time and a more serious situation if corruption occurs. If your archive is affected you are fucked. In my case, I could lose a bunch of files but not an archive holding hundreds or thousands of files. I got a bunch of TB's in a NAS configuration. |
With the cost of hard drives and memory these days, who needs to compress
not you...i guess
Waste of time
in your eyes it is
I got a bunch of TB's in a NAS configuration
Good for you ........ goooood for you  |
|
| Back to top |
|
 |
Csimbi Elite Member

Joined: 05 Mar 2010 Posts: 5358 Location: The bright side of the dark side
|
Posted: Wed Aug 12, 2015 12:34 am Post subject: |
|
|
@gx-x
Dunno. When the WinRAR score jumps from ~1700 to ~1900 (single core), that's 10%. For free. Enable multi-core and you will see some serious boost.
If you check this, the 2600K ticks at 3.8Ghz, while the 4790K ticks at 4.4Ghz.
That's 13%, without overclocking.
Its memory bandwidth is also higher.
So why do you say that it's only 5% faster? |
|
| Back to top |
|
 |
gx-x Elite Member

Joined: 02 Jul 2007 Posts: 2655
|
Posted: Wed Aug 12, 2015 6:54 am Post subject: |
|
|
| Csimbi wrote: | @gx-x
Dunno. When the WinRAR score jumps from ~1700 to ~1900 (single core), that's 10%. For free. Enable multi-core and you will see some serious boost.
If you check this, the 2600K ticks at 3.8Ghz, while the 4790K ticks at 4.4Ghz.
That's 13%, without overclocking.
Its memory bandwidth is also higher.
So why do you say that it's only 5% faster? |
every godamn test on internetz says that clock for clock it's 5% faster AT BEST.
1700 1900 single core? It can only be lower in multicore, not higher, but you tested it, you would have posted it if it were higher. You choose a scenario where memory would make up for the use of cores.
Dude, please....i7 2600Kis good cpu. Go buy whatever. I told you that you can save money, you don't need it. But hey, it's your money.
"If you check this, the 2600K ticks at 3.8Ghz, while the 4790K ticks at 4.4Ghz. "
what the hell does this mean or prove? You can't get your 2600K to 4.4? You can't up the core voltage to 2.2/3V? Amazing... Clock for clock was always ASSUMED at my end, I am sorry if that made confusion. Clock for Clock differences are minimal at best. You are better off investing in 5xxx series and getting 6 core i7, 4xxx is waste of money at this point. |
|
| Back to top |
|
 |
Csimbi Elite Member

Joined: 05 Mar 2010 Posts: 5358 Location: The bright side of the dark side
|
Posted: Wed Aug 12, 2015 11:59 am Post subject: |
|
|
But who'd run it clock for clock?
You can overclock the 2600K to where the 4790K runs at stock.
The 2600K will be left behind when you overclock the 4790K.
Hence, the clock for clock comparisons are useless.
I also like that it uses less power. It pays for it in the long run. |
|
| Back to top |
|
 |
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
Powered by phpBB © 2001, 2666 phpBB Group
|
|