VOGONS

Common searches


DOSBox 0.74 CPU Benchmark

Topic actions

Reply 140 of 191, by RayeR

User metadata
Rank Member
Rank
Member

I installed XP with integrated SP3. I read somewhere that SP3 eats more memory than SP1 but shouldn't be significantly slower. It occupied about 50MB of 196MB RAM so no reason for big swapping, I didn't noticed any loud HDD work during test...

Gigabyte GA-P67-DS3-B3, Core i7-2600K @4,5GHz, 4GB DDR3, 128GB SSD, GF7900GT, SB Audigy + YMF724F + DreamBlaster combo

Reply 141 of 191, by Tertz

User metadata
Rank Oldbie
Rank
Oldbie
RayeR wrote:

I installed XP with integrated SP3. I read somewhere that SP3 eats more memory than SP1

If you are using the machine for Win games then there is a sense to try older OS (9x, 2000, original XP) in the hope to increase the performance.

DOSBox CPU Benchmark
Yamaha YMF7x4 Guide

Reply 142 of 191, by RayeR

User metadata
Rank Member
Rank
Member

Sure, and most of DOS games can run native as I have SB-AWE64 there 😀 This is my test machine where I have DOS, Win98SE, NT4, XP embedded, XP-Pro and ReactOS. I was curious how will XP perform on this old HW and I can say pretty well, GUI is responsive enough. But I cannot say this about ReactOS it's real pain of slowness, BTW some days ago I made this vid: https://www.youtube.com/watch?v=rjGmVs7rXOY

Gigabyte GA-P67-DS3-B3, Core i7-2600K @4,5GHz, 4GB DDR3, 128GB SSD, GF7900GT, SB Audigy + YMF724F + DreamBlaster combo

Reply 143 of 191, by Tertz

User metadata
Rank Oldbie
Rank
Oldbie
RayeR wrote:

Sure, and most of DOS games can run native as I have SB-AWE64 there 😀 This is my test machine where I have DOS, Win98SE, NT4, XP embedded, XP-Pro and ReactOS.

If you have Win98SE installed then may be checked, whether games and DB performance is higher there than in XP sp3 on PPro.

DOSBox CPU Benchmark
Yamaha YMF7x4 Guide

Reply 144 of 191, by zirkoni

User metadata
Rank Member
Rank
Member
zirkoni wrote:

Intel Core i5 4200M 2.5 GHz (laptop, dual-boot)
- Ubuntu 14.10 (64-bit): 19.2
- Windows 7 (64-bit): 44.4

I made this https://youtu.be/dcygzqsGVc4 comparison between 32 & 64-bit DOSBoxes on Linux. Without the screen recording the results of the 32 binary with Quake might have been closer to that Windows result. That 19.2 was the result of using the 64-bit DOSBox 0.74 that was installed from the Ubuntu repositories.

https://www.youtube.com/user/zirkoni88

Reply 147 of 191, by Tertz

User metadata
Rank Oldbie
Rank
Oldbie
MrEWhite wrote:

Getting a Kaby Lake 7700k soon (is that new enough? 😜).

I hope it still will be so to the time when we'll see your fps. 😀 Report RAM's frequency also.

DOSBox CPU Benchmark
Yamaha YMF7x4 Guide

Reply 150 of 191, by DosFreak

User metadata
Rank l33t++
Rank
l33t++

Need someone to test the compiled DOSBox for Windows 95 on a P3 or below:
Post 619398

Need to see if it works with WINDIB or Directx or if it works with neither and crashes on launch

Directx needs ddraw acceleration enabled and 16bit color on the desktop.

Thanks

DOSBox Compilation Guides
DosBox Feature Request Thread
PC Game Compatibility List
How To Ask Questions The Smart Way
Running DRM games offline

Reply 151 of 191, by MrEWhite

User metadata
Rank Oldbie
Rank
Oldbie

Finally got around to delidding my i7 7700k, so here are 5 GHz benchmarks for Quake.

i7 7700k @ 5 GHz, 16GB DDR4 RAM @ 3600 MHz, Windows 10 Pro 64-Bit

75.9, 85.7, 74.4 - 800x600
233.6, 237.6, 221.8 - 320x200

Reply 154 of 191, by Tertz

User metadata
Rank Oldbie
Rank
Oldbie
MrEWhite wrote:

If OP still is on here, I could do a 8086k @ 5 GHz run.

You may post new results in the theme. People will can read them, anyway. Later I may input them in the list.

The data for new kinds of CPUs is more interesting as of Ryzens, Atoms, CPU for smartphones, etc.

leileilol wrote:

Considering the recent recompiler changes this thread is all moot now (as i've warned before).

It's the overstatement. Without the concrete numbers the moot is to think the changes in the speed are so big to change the situation principally in 0.74-2. Also there may stay in the further usage versions close to 0.74 like ykhwong's one which gave close results. Besides the speed's changes may be in general speed's improvement alike +X% in average, what leaves the previous results as useful as they may allow to approximate the results for newer emulator's version.

DOSBox CPU Benchmark
Yamaha YMF7x4 Guide

Reply 155 of 191, by latalante

User metadata
Rank Newbie
Rank
Newbie

Intel Core2 T5300 1.73GHz, VIA Chrome P4M900
Dosbox 0.74-2, Windows XP 32-bit
Quake demo1 800x600 - 10.3
PC Benchmark - 30.0

Dosbox 0.74-2, Linux 4.19.0 32-bit
Quake demo1 800x600 - 12.1
PC Benchmark - 33.5

dosemu2 (git), Linux 4.19.0 32-bit
Quake demo1 800x600 - 72.7
Quake demo1 320x200 - 228.1 (scaled2x)
PC Benchmark - 353.8
Speed Test /L - 1784314 parrots, 661.6 XT

FreeDOS 1.2
Quake demo1 320x200 - 414.3 (nosound, an attempt to change the resolution, ends with crashes - invalid opcode, only native work 1280x800)
Speed Test /L - 11375000 parrots, 4217.6 XT

Edit:
I've recompiled DOSBox using g++ 4.8.4, it's the best combination for the dosbox code and this processor. At Quake demo1 800x600 the result is better by 1-1.2 FPS under Linux ~13.3.

Last edited by latalante on 2018-11-11, 20:47. Edited 1 time in total.

Reply 157 of 191, by Qbix

User metadata
Rank DOSBox Author
Rank
DOSBox Author

Is there a huge difference ? The algorithm that determines the max speed was changed a lot between 0.74 and 0.74-2 and the 64 recompiler got a lot of changes.
However on win32 the only change is the max cycles algorithm.

Water flows down the stream
How to ask questions the smart way!

Reply 159 of 191, by Tertz

User metadata
Rank Oldbie
Rank
Oldbie
Qbix wrote:

Is there a huge difference ?

The difference should not be "huge".
But for more correct CPU comparision it's better to use the same and it's not a problem to do as the original version is in the links.

The testing of 0.74 vs 0.74-2 would be better to do on the modern desktop CPUs in Windows 8.1 x64 or later OS.

DOSBox CPU Benchmark
Yamaha YMF7x4 Guide