VOGONS

Common searches


DOSBox 0.74 CPU Benchmark

Topic actions

Reply 21 of 196, by Tertz

User metadata
Rank Oldbie
Rank
Oldbie
smeezekitty wrote:

Timedemo 1: 23.8
Timedemo 2: 36.3
Timedemo 3: 46.3

Thanks, but your results look as not valid.
As fps look strange, - so it's required to follow the described testing process accurately.
If you have run 3 different demos: it's needed to run demo1 for 3 times by a command "timedemo demo1" and then to say an average result for this demo1. demo2 and demo3 are not obligatory.

upd: Meanwhile, I'll use the 1st number as close to what was expected.

Last edited by Tertz on 2015-05-18, 14:07. Edited 3 times in total.

DOSBox CPU Benchmark
Yamaha YMF7x4 Guide

Reply 22 of 196, by joacim

User metadata
Rank Newbie
Rank
Newbie

i5-4690 running at stock speeds. CPU mostly stayed at 3.9 GHz with some dips down to 3.8 GHz. 64-bit Windows 7.

demo1:
969 frames 17.5 seconds 55.3 FPS
969 frames 17.5 seconds 55.3 FPS
969 frames 17.4 seconds 55.8 FPS
55.47 average FPS

demo2:
985 frames 15.7 seconds 62.9 FPS
985 frames 15.6 seconds 63.2 FPS
985 frames 15.6 seconds 63.1 FPS
63.07 average FPS

demo3:
1090 frames 19.6 seconds 55.5 FPS
1090 frames 19.6 seconds 55.7 FPS
1090 frames 19.7 seconds 55.4 FPS
55.53 average FPS

Tests were ran after a fresh boot with the networking interface disabled. I didn't bother with disabling Virtual CloneDrive (Because I'm lazy!).

Let me know if you need more information from me, or if you want me to rerun with different settings. 😀

Reply 23 of 196, by Tertz

User metadata
Rank Oldbie
Rank
Oldbie
joacim wrote:

CPU mostly stayed at 3.9 GHz with some dips down to 3.8 GHz. <...> Let me know if you need more information from me, or if you want me to rerun with different settings.

If during measuring in demos there existed some monitoring GHz application like CPU-Z, then without this application your fps may to be a little higher on 1-2 fps.

DOSBox CPU Benchmark
Yamaha YMF7x4 Guide

Reply 24 of 196, by GeorgeMan

User metadata
Rank Oldbie
Rank
Oldbie

i5-2500K @ 3.7t (stock)
With internet enabled, no other programs running but also not disabled the background ones, which are only: Corsair Link, AMD Catalyst, Hard disk sentintel, Intel RST.

demo1: 41.3fps
demo2: 45.8fps
demo3: 40.9fps
average: 42.3fps

Windows 8.1 x64

😀

1. Athlon XP 3200+ | ASUS A7V600 | Radeon 9500 @ Pro | SB Audigy 2 ZS | 80GB IDE, 500GB SSD IDE2Sata, 2x1TB HDDs | Win 98SE, XP, Vista
2. Pentium MMX 266| Qdi Titanium IIIB | Hercules graphics & Amber monitor | 1 + 10GB HDDs | DOS 6.22, Win 3.1, 95C

Reply 25 of 196, by Tertz

User metadata
Rank Oldbie
Rank
Oldbie
GeorgeMan wrote:

which are only: Corsair Link, AMD Catalyst, Hard disk sentintel, Intel RST

Hardware monitoring software like Sentintel and Corsair Link take resources. These both could eaten up to 5-10% of fps. If you technically can to measure without them, please do it.

DOSBox CPU Benchmark
Yamaha YMF7x4 Guide

Reply 26 of 196, by SquallStrife

User metadata
Rank l33t
Rank
l33t

Core i5 2400 @ stock, Win8.1 x64, showing 3850.12MHz.

demo1: 45.4 46.1 45.8 avg 45.7
demo2: 51.1 51.5 51.4 avg 51.3
demo3: 45.8 45.8 45.7 avg 45.8

VogonsDrivers.com | Link | News Thread

Reply 27 of 196, by GeorgeMan

User metadata
Rank Oldbie
Rank
Oldbie
Tertz wrote:
GeorgeMan wrote:

which are only: Corsair Link, AMD Catalyst, Hard disk sentintel, Intel RST

Hardware monitoring software like Sentintel and Corsair Link take resources. These both could eaten up to 5-10% of fps. If you technically can to measure without them, please do it.

Actually I disabled ethernet, did a restart, closed them but the FPS in all tests were ~10% lower. Go figure... 😜

1. Athlon XP 3200+ | ASUS A7V600 | Radeon 9500 @ Pro | SB Audigy 2 ZS | 80GB IDE, 500GB SSD IDE2Sata, 2x1TB HDDs | Win 98SE, XP, Vista
2. Pentium MMX 266| Qdi Titanium IIIB | Hercules graphics & Amber monitor | 1 + 10GB HDDs | DOS 6.22, Win 3.1, 95C

Reply 28 of 196, by joacim

User metadata
Rank Newbie
Rank
Newbie
Tertz wrote:
joacim wrote:

CPU mostly stayed at 3.9 GHz with some dips down to 3.8 GHz. <...> Let me know if you need more information from me, or if you want me to rerun with different settings.

If during measuring in demos there existed some monitoring GHz application like CPU-Z, then without this application your fps may to be a little higher on 1-2 fps.

I did not run cpuz during measurements. I did a fourth run after with cpuz to see what the frequency was like during the demos.

Reply 29 of 196, by Tertz

User metadata
Rank Oldbie
Rank
Oldbie

Thanks all for your results.

GeorgeMan wrote:

Actually I disabled ethernet, did a restart, closed them but the FPS in all tests were ~10% lower.

Interesting. If in your experiments with applications having a negative load you'll get a higher fps, then feel free to post new results. Meanwhile I'll use the current ones. Maybe some software changes frequences or conditions when turbo mode is turned on, - this may be checked by CPU-Z.

DOSBox CPU Benchmark
Yamaha YMF7x4 Guide

Reply 30 of 196, by smeezekitty

User metadata
Rank Oldbie
Rank
Oldbie
Tertz wrote:
Thanks, but your results look as not valid. As fps look strange, - so it's required to follow the described testing process accu […]
Show full quote
smeezekitty wrote:

Timedemo 1: 23.8
Timedemo 2: 36.3
Timedemo 3: 46.3

Thanks, but your results look as not valid.
As fps look strange, - so it's required to follow the described testing process accurately.
If you have run 3 different demos: it's needed to run demo1 for 3 times by a command "timedemo demo1" and then to say an avarage result for this demo1. demo2 and demo3 are not obligatory.

upd: Meanwhile, I'll use the 1st number as close to what was expected.

You are right. I screwed up.
I read the time in seconds rather than the FPS on the last two time demos for some reason.

New run:
Timedemo 1:
23.8
24.1
24.0
avg: 24

Timedemo 2:
27.6
27.5
27.5
avg: 27.53

Timedemo3:
23.9
23.9
24.0

Reply 31 of 196, by kithylin

User metadata
Rank l33t
Rank
l33t

I'm trying to figure out how to make this work but this is the entire reason I never used dosbox instead of real dos PC's. I'm sitting here trying to figure out how to make it work.. I think I got all the configuration file settings right.. there's a ton of crap in there and it sounds like you want us to just input what you said in an empty file or something.. confused on that part.

Then I actually find the settings you want and get the program running and figured out how to mount c:\dos as c: in to the dosbox thing.. and when I extract files in windows, I try using DIR in the dosbox prompt and it shows no changes, no new files no new folders. So.. can't figure out how to run anything yet. I think I have to close dosbox and re-open it and re-mount it every single time I change contents of the folders with windows (outside of dosbox). In general dosbox is clunky and not straight-forward to use. I'm always expecting to you know........... just open it, and it just works 100% like real dos, but it never does. I might eventually figure out how to get this working.... I guess I'm a moron or something but this seems so clunky.

EDIT: yep, have to exit the dosbox program, then re-open it, then re-mount it again to see changes to the folder contents, so annoying. I finally got quake shareware installed and I'll get to testing it after I go eat.

EDIT #2: Finally figured it out, here's results.

EDIT #3: Re-ran it to get each one 3 times, and updated post here with screenshots and results.

LGA-771 quad core 12MB cpu @ 3.75 ghz in 775 motherboard. Dual channel DDR2-938

Demo 1
=====
Run1 - 30.8 FPS
Run2 - 30.7 FPS
Run3 - 30.7 FPS
771-quad-3.75-Ghz-Quake-Test-1.jpg

Demo 2
======
Run1 - 36.3 FPS
Run2 - 36.2 FPS
Run3 - 36.1 FPS
771-quad-3.75-Ghz-Quake-Test-2.jpg

Demo 3
======
Run1 - 34.2 FPS
Run2 - 39.3 FPS
Run3 - 33.9 FPS
771-quad-3.75-Ghz-Quake.jpg

There you go, that's my contribution then to the thread here. 😎

Last edited by kithylin on 2015-05-17, 19:35. Edited 1 time in total.

Reply 33 of 196, by kithylin

User metadata
Rank l33t
Rank
l33t
joacim wrote:

Yeah. Demos run 3 times each. 😀

Which CPU is that exactly?

Working on that, had to go re-run stuff. Working on updating above post now.

EDIT: Above post updated.

Reply 34 of 196, by Tertz

User metadata
Rank Oldbie
Rank
Oldbie
smeezekitty wrote:

New run

Thanks for new run.
You mentioned 2.9 GHz, but as I've understood you intended to run the CPU at standard frequency 2.8. I've written standard frequency 2.8 as results are same as at slower CPU, so I conceded that frequency 2.9 could be incorrectly shown somewhere. You may check the frequency by CPU-Z and fix me if it's 2.9 indeed.

kithylin wrote:

I never used dosbox instead of real dos PC's

For those who never used DOSbox:

1) create new dosbox-0.74.conf as described
2) change settings in dosbox-0.74.conf as described
3) Then. In dosbox-0.74.conf, in its section [autoexec] place 2 strings:
mount c c:\games\
c:
[c:\games\ - is your folder for dosbox games (create a new one). it will be associated in dosbox with disk C]
4) create in c:\games\ a folder "qinst". Extract quake's install files from quake106.zip to c:\games\qinst\
5) run dosbox. Write there 2 strings:
cd qinst
install
[This will install Quake in a new folder c:\games\quake_sw\]
It will run the game after the installation. If you'll exit dosbox and will want to run quake again then:
run dosbox. Write there:
cd quake_sw
quake

The End.

I think I have to close dosbox and re-open it and re-mount it every single time I change contents of the folders with windows (outside of dosbox)

Folders mounting is ordinary done with .conf. To update files list (changed outside of dosbox) you may press CTRL-F4 or re-run dosbox.

I might eventually figure out how to get this working....

It needs some experience with configuration as most emulators. At least you know DOS commands, what is Sound Blaster etc, while many people today don't.

A remark. Your 1st results were higher for demo1 and demo2 on 3 fps. I suppose you diverged from the recommended procedure at 2nd measurement (had simultaneous application like AIDA, lost game's window focus, or else), so I'll use your 1st more correct results.

Thanks for your quest and its results.
Also I need information about your OS and its bits (32/64).

Your CPU was mentioned as QX9770. AIDA shows Xeon. Looks as you changed CPU, or I mistaked when read text at night.

Last edited by Tertz on 2015-05-17, 20:10. Edited 1 time in total.

DOSBox CPU Benchmark
Yamaha YMF7x4 Guide

Reply 35 of 196, by sunaiac

User metadata
Rank Oldbie
Rank
Oldbie

Intel Core i7 980X 4.13 GHz (Win8.1-64): 43, 48, 43

R9 3900X/X470 Taichi/32GB 3600CL15/5700XT AE/Marantz PM7005
i7 980X/R9 290X/X-Fi titanium | FX-57/X1950XTX/Audigy 2ZS
Athlon 1000T Slot A/GeForce 3/AWE64G | K5 PR 200/ET6000/AWE32
Ppro 200 1M/Voodoo 3 2000/AWE 32 | iDX4 100/S3 864 VLB/SB16

Reply 36 of 196, by kithylin

User metadata
Rank l33t
Rank
l33t
Tertz wrote:

A remark. Your 1st results were higher for demo1 and demo2 on 3 fps. I suppose you diverged from the recommended procedure at 2nd measurement (had simultaneous application like AIDA, lost game's window focus, or else), so I'll use your 1st more correct results.

The earlier tests (The first one I had listed) was with tons of background crap running (Chrome, fraps, AIDA64, download managers, steam, chat clients, trillian, mirc, etc.) The later tests (and the finals I posted in my post up there) are with all background processes closed, even the intel tray icon for managing my raid-0 arrays on that machine, and even some background processes I closed.. (java updater, acronis monitor, etc).

Tertz wrote:

Thanks for your quest and its results.
Also I need information about your OS and its bits (32/64).

This machine was run on Windows 7 64-bit. I don't know if it matters but the system has 6GB of ram (3x3GB) run in "Flex Mode", where this system lets me run as Dual-Channel while only having 3 sticks of ram.

Tertz wrote:

Your CPU was mentioned as QX9770. AIDA shows Xeon. Looks as you changed CPU, or I mistaked when read text at night.

I'm not sure where you got that from as I have never in my life owned a QX9770 on any computer. I've never (to the best of my knowledge) ever claimed to own one either. You might be confusing me with Skyscraper (Skyscraper) that I -THINK- might own one of those chips. Myself however, I went the cheapass route (Because I have hardly any income) and I've waited a few years and opted for the xeons and used those in my 775 system, then overclocked and went that route instead. The QX9770's are still super expensive today even used on ebay years later. And I mean $240 for the cheapest one right now. In comparison I paid $17.75 after shipping for my L5408 chip in feb 2015.

Reply 37 of 196, by Tertz

User metadata
Rank Oldbie
Rank
Oldbie
kithylin wrote:

The earlier tests (The first one I had listed) was with tons of background crap running (Chrome, fraps, AIDA64, download managers, steam, chat clients, trillian, mirc, etc.)

GeorgeMan had similar miracle above, - loss of ~10% with cleaner memory. For example, if AIDA had no opened window previously, but had it during demos run at 2nd measurement - this could lead to the difference, as I saw opened similar frequency monitor CPU-Z lead to less results. Another possibility - the game's window had no focus and hence game got less resources. Stable (demo1 then demo2) higher fps can't be farther from correct ones, - I don't know a mechanism for it.
Another thing. E8600 from same as your Xeon 2008 year has at 4 GHz 37 fps on demo1. I doubt Xeon is worse CPU at same frequency. On 3.8 GHz E8600 would get 35 fps what is close to 34 on your 1st results, but not to 31 on 2nd results.

I'm not sure where you got that from

From one post wich I've read incorrectly.

DOSBox CPU Benchmark
Yamaha YMF7x4 Guide

Reply 38 of 196, by kithylin

User metadata
Rank l33t
Rank
l33t
Tertz wrote:

GeorgeMan had similar miracle above, - loss of ~10% with cleaner memory. For example, if AIDA had no opened window previously, but had it during demos run at 2nd measurement - this could lead to the difference, as I saw opened similar frequency monitor CPU-Z lead to less results. Another possibility - the game's window had no focus and hence game got less resources. Stable (demo1 then demo2) higher fps can't be farther from correct ones, - I don't know a mechanism for it.
Another thing. E8600 from same as your Xeon 2008 year has at 4 GHz 37 fps on demo1. I doubt Xeon is worse CPU at same frequency. On 3.8 GHz E8600 would get 35 fps what is close to 34 on your 1st results, but not to 31 on 2nd results.

Just so you know, I disable all of that "fluxating frequency" bulls--- on both of my intel systems and they always run at one constant-and-fixed clock speed at all times, no matter what load is on them. So the displayed frequency in my screenshots, is what speed it is at, -at all times-, even if the dosbox window is in focus or not. I did however make sure to click on the window and make it in-focus for my later tests, the 3-each with screenshots.

Reply 39 of 196, by Tertz

User metadata
Rank Oldbie
Rank
Oldbie
kithylin wrote:

So the displayed frequency in my screenshots, is what speed it is at, -at all times-, even if the dosbox window is in focus or not. I did however make sure to click on the window and make it in-focus for my later tests, the 3-each with screenshots.

Focus on the game's windows has to be not only for frequency, but what priority gets dosbox in OS and hence resources. But according to what you say focus was appropriate.
Another thing left still, - if you had AIDA frequency monitor window opened during demos run - it could lead to less fps by taking resources, as CPU-Z did.

Your 3rd picture for demo3 has correct results for demo1, demo2 and demo3. If these results were gotten on 2nd run of demonstrations then you may remember what you've done to get them. Maybe nothing, just some background process has finished.

DOSBox CPU Benchmark
Yamaha YMF7x4 Guide