GeoTrail wrote:It crashes on my computer.
After the CPU test is done, the result is displayed really quickly then the program ends with an error from Vista saying the program has stopped working. Same result when running it in the PB IDE and as a compiled console app.
Comparing an XP VM with a vista VM, the XP Box gets better CPU stats and a better multiplier for threads (on the same underlying hardware), it also gets better 2d gfx (but that doesn't mean much on a VM). For disk checks though, the vista one gives very odd numbers millions of times higher than my PC. I suspect there is some new caching there and the whole thing is happening in memory. I might add that flush file buffers code in there as a separate uncached disk check and see.
I compared the results to some other benchmark programs and I think I need to work on the disk stats a bit more and add a 3D GFX. It seems that many older Graphics cards run 2d graphics slightly better for some reason so I'm not really catching the data I want.
small steps...
Paul Dwyer
“In nature, it’s not the strongest nor the most intelligent who survives. It’s the most adaptable to change” - Charles Darwin
“If you can't explain it to a six-year old you really don't understand it yourself.” - Albert Einstein
=====================================================================
| PureBench32 v1.0 |
| by Paul Dwyer |
| and the PB community |
=====================================================================
CPU Tests:
Single thread score: 15269
Multi thread iterations: 39187
Multi thread perf increase: 2.57x
Disk:
Cached Disk Writes: 122485172
Disk Reads: 105627035
PureBasic 5.73 | SpiderBasic 2.30 | Windows 10 Pro (x64) | Linux Mint 20.1 (x64)
Old bugs good, new bugs bad! Updates are evil: might fix old bugs and introduce no new ones.
Vista is definately doing something funky with the disk buffers! (which is a good feature I guess provided the power on the system or disk doesn't die suddenly)
I'll try and add a non cached disk check.
What do you mean about only the last posted exe worked? do you mean that you couldn't compile, only run it? (even with debug off and threadsafe on?)
Thanks for the test results
Paul Dwyer
“In nature, it’s not the strongest nor the most intelligent who survives. It’s the most adaptable to change” - Charles Darwin
“If you can't explain it to a six-year old you really don't understand it yourself.” - Albert Einstein
PureBasic 5.73 | SpiderBasic 2.30 | Windows 10 Pro (x64) | Linux Mint 20.1 (x64)
Old bugs good, new bugs bad! Updates are evil: might fix old bugs and introduce no new ones.
The default location for the app is c:\ to write the file but vista doesn't let it get created so it just failes on each loop which is a lot faster than reading and writing to the disk. If I use /p and set a path I can write to then the performance goes to realistic levels.
I'll need to check that the app can write to the specified location and show an error
Paul Dwyer
“In nature, it’s not the strongest nor the most intelligent who survives. It’s the most adaptable to change” - Charles Darwin
“If you can't explain it to a six-year old you really don't understand it yourself.” - Albert Einstein
=====================================================================
| PureBench32 v1.0 |
| by Paul Dwyer |
| and the PB community |
=====================================================================
CPU Tests:
Single thread score: 24328
Multi thread iterations: 54032
Multi thread perf increase: 2.22x
Disk:
Cached Disk Writes: 101281526
Non Cached Disk Writes: 85631425
Disk Reads: 87732823
Grphics:
2D Graphics Score: 29768
I Stepped On A Cornflake!!! Now I'm A Cereal Killer!
You are using vista are you? Those disk scores imply that the file couldn't be created. I need to put the check in for that still. you may need the /p option to run it in your docs directory till I fix it
What's your CPU? Core 2 duo or AMD dual? What's the Ghz?
Cheers for the test run
EDIT: I've updated the main code and the download file, if the Disk check file can't be created the checks return 0 right away
Paul Dwyer
“In nature, it’s not the strongest nor the most intelligent who survives. It’s the most adaptable to change” - Charles Darwin
“If you can't explain it to a six-year old you really don't understand it yourself.” - Albert Einstein
pdwyer wrote:=====================================================================
| PureBench32 v1.0 |
| by Paul Dwyer |
| and the PB community |
=====================================================================
CPU Tests:
Single thread score: 29092
Multi thread score: 103754
Multi thread perf increase: 3.57x
Disk:
Non Cached Disk Writes: 1178
Cached Disk Writes: 1132
Disk Reads: 66436
The CPU is a Intel Centrino dual core 2 at 2 GHz.
Here are the results from the updated exe:
=====================================================================
| PureBench32 v1.0 |
| by Paul Dwyer |
| and the PB community |
=====================================================================
CPU Tests:
Single thread score: 24452
Multi thread score: 52684
Multi thread perf increase: 2.15x
Disk:
Non Cached Disk Writes: 0
Cached Disk Writes: 0
Disk Reads: 0
Grphics:
2D Graphics Score: 29333
I Stepped On A Cornflake!!! Now I'm A Cereal Killer!
=====================================================================
| PureBench32 v1.0 |
| by Paul Dwyer |
| and the PB community |
=====================================================================
CPU Tests:
Single thread score: 13913
Multi thread score: 36729
Multi thread perf increase: 2.64x
Disk:
Non Cached Disk Writes: 414
Cached Disk Writes: 228
Disk Reads: 22371
Grphics:
2D Graphics Score: 108832
It's starting to look like flushing buffers and playing with disk caches isn't really meaningful, I think I'll get rid of that little extra.
CPU status look good though and 2d graphics do to, but not in the sense of "I have a better card" older cards do seem to be doing this better but the numbers are not syncing with CPU scores at all which was my worry.
I think I will try adding the PB OpenGL cube with an FPS count, I can't tell if it will be meaningful or not so it might be worth adding to find out.
The other thing I will add next is exclusion switches so you don't have to run all tests, you can just do disk or cpu or whatever.
Any other ideas? additions, removals, improvements?
Paul Dwyer
“In nature, it’s not the strongest nor the most intelligent who survives. It’s the most adaptable to change” - Charles Darwin
“If you can't explain it to a six-year old you really don't understand it yourself.” - Albert Einstein
