some little problems with LOFC2 in Win7

[EtherealN]

As an aside, some interesting information that can be seen in the HD Tune graph:

 

The 600GB Velociraptor looks very very good there - it's on the top straight through the graph. But: it's a 600GB drive, and it costs (at my local retailer) 2 000 SEK. (333 US dollars). The Hitachi Deskstar is 3TB and costs the same.

 

Speed span for the Velociraptor is ~155 to 100 MB/second, Deskstar span is ~145 to ~75. But say what we want is 600GB of storage. Which will be fastest? Span for the first 600GB of the Deskstar is 20%. This makes the speed span for the 600GB on the deskstar 145 to 140. So with the Velociraptor we get median ~135MB/s, with the Deskstar we get median ~142MB/s. As far as seek times go, Velociraptor seeker head traverses a 2.5 inch platter. Deskstar only needs to traverse 20% of it's 3.5 inch platter - 0,7 inches.

 

Now, this is all extremely synthetic, so I cannot say that the Deskstar will be faster in both transfer and seek than the Velociraptor when using 600GB. I can say that it will have a higher median transfer rate for sequential transfers, and I feel reasonably certain to say that the seeker head will "seek" faster on the Deskstar as well (the difference in area traversed is large enough to accomodate for slower seeker actuation). But it does get more complex in that a more dense drive (the deskstar has more data on a given area) makes it harder for the seeker head to really "lock on" to the track it wants to be at. (LOTS of variability here: some drives have denser platters, making it harder to "lock on", other drives just have extra platters with extra seeker heads.)

 

But what I want to highlight is that through limiting seeker movement, a sort of "normal" drive just beat a "performance" drive in synthetic numbers - and it still has 2.4gigabytes left if you were to end up needing it. (But this would cause the seeker head to move across a wider span, sacrificing some of your performance.)

 

And all of this will affect your computer a lot more than one might originally think. I know I had a huge "AHAH!" moment when I started digging into this stuff, coming from a place where I basically just thought of hard drives as spindle speed and storage size...

Edited April 29, 2011 by EtherealN

[EtherealN]

No worries Caponi, I'll try to fix a guide for you until you get back. :)

[asparagin]

 

On my drive, I have asked my utility (Ultimate Defrag) to ensure that all files in C:\Games\DCS Warthog are in sequence.

W.

 

Did you select the DCS folder for "high performance" and used "frag protect" for that?

 

I can confirm that after defragging With UD, and defining the DCS FC2 folder for high performance (moves it to the outside layer) and using frag protect I get faster loading

_____________Before ________After

Preload:______20 sec________15 sec

Load:________40 sec________15 sec

Total:________60 sec________30 sec

 

Ps: can't rep you for info Ethereal, because "I have to spread some rep around".. but you don't need it anyway:P

Edited April 29, 2011 by asparagin

[EtherealN]

Oooh, very happy to see those results. I've only very recently started using this utility (I previously used Defraggler and used some arcane tricks to make similar things happen), so this is a good datapoint that it wasn't just me, so to speak. (I've been playing around with it, but in all "science" you want repetition from independent sources to verify your data :) )

 

One thing to keep in mind there:

You have two specific types: High Performance and Archive. Say you set C:\Games\DCSBS as High Performance, and then to vacate the high-speed sectors you set a general "Archive" rule of 40 days. You migh then end up in the unlucky position that there's some file in the DCSBS folder that just happened to not have been used - it would get moved to the slow part. For this, you can (under archive settings) set an exception for that folder. This means the program will "respect" the High Speed setting for that folder, but will not "respect" the day limit on Archiving for that folder.

 

This can be useful if there's a game you haven't played for a while, but you know you will play it some time in the future and you don't want to re-do the operation for that. In this case, add it's folders to "High-Prformance" and add exception to it's folder in "Archive".

 

Frag Protect I haven't really explored much as far as performance postfactum goes - it's more a measure to increase defragmentation performance. Unfortunately the program itself is a bit aged (you can see that in the manual) but since NTFS is still NTFS it still works.

 

Again though, thanks for sharing those numbers. :)

 

EDIT: Now, even though that program is payware, when you look at cutting 60 seconds total to 30... That's fairly nice for just a simple software run operation, isn't it. :)

Edited April 29, 2011 by EtherealN

[manfrez01]

A little help for users which have WD Caviar Green series.

I had many problems with this series and the low performance of the disk.

Just in case

 

http://foro.noticias3d.com/vbulletin/showthread.php?t=350892

[asparagin]

Yes, it worked well for FC2, for DCS A-10C, didn't notice a big improvement in the loading times, maybe that part was already in good shape.

[caponi]

good morning ...

 

i have used UD with folowing options:

 

consolidate, high-performancve for iracing, ka-50, fc2 ...

run UD and see that the most of data lay in high performance area ...

 

result is:

iRacing ... i didn't try, because i have no comparison values, but i feel loading is good

ka-50 ... loading bahaviour is as before but 5 sec faster

fc2 loading time is faster (5 sec) after a 30sec delay time :(

 

fc2 has the same behaviour ...

simulator.exe get 5% cpu time 180k memory and has to wait for round about 30 sec ...

after this 30 sec the application runs like a pimped race car :thumbup:

 

conclusion: all loading times are better, but the main problem (it is not a problem, but it is anoying) still exists

 

regards

Cap

[caponi]

i have found some unusal things, but i have no explanation ...

 

have a look on newInputStart.txt in my temp directory

quote:

_____________________________________

2011.05.02 18:52:11:922

Try to create device. Device type is: 274. Device name is: Maus

2011.05.02 18:52:11:922

diDevice->GetCapabilities

2011.05.02 18:52:11:922

Mouse created[Maus].

2011.05.02 18:52:11:922

Try to create device. Device type is: 1043. Device name is: Tastatur

2011.05.02 18:52:11:922

diDevice->GetCapabilities

2011.05.02 18:52:11:922

Keyboard created[Tastatur].

2011.05.02 18:52:11:930

Try to create device. Device type is: 65553. Device name is: Saitek Magic Hotkey Device

2011.05.02 18:52:11:930

diDevice->GetCapabilities

2011.05.02 18:52:11:946

Try to create device. Device type is: 65553. Device name is: SCISSORS Keyboard

2011.05.02 18:52:11:946

diDevice->GetCapabilities

2011.05.02 18:52:11:961

Try to create device. Device type is: 65553. Device name is: SCISSORS Keyboard

2011.05.02 18:52:11:961

diDevice->GetCapabilities

2011.05.02 18:52:11:961

TrackIR dll location key is not presented!

2011.05.02 18:52:11:961

Cannot load headtracker.dll!

2011.05.02 18:52:24:988 <<<<<<< 15 sec ! ... in XP 10 sec.

Device unacquire Maus

2011.05.02 18:52:24:988

Device unacquire Tastatur

2011.05.02 18:52:30:052 <<<<<<< 6 sec ... not present in XP

Try to create device. Device type is: 274. Device name is: Maus

2011.05.02 18:52:30:052

diDevice->GetCapabilities

2011.05.02 18:52:30:053

Mouse created[Maus].

2011.05.02 18:52:30:053

Try to create device. Device type is: 1043. Device name is: Tastatur

2011.05.02 18:52:30:053

_____________________________________

 

could i deactivate TrackIr with some special lua options ?

 

i would try to load FC2 without missung TIR

[Chaquan]

You can do is press the start menu and in the "Resource Monitor" in the search field type. Start the program. Now, start the emulator in windowed mode. Expand CPU section and in the list to find simulator.exe, the mark....

[caponi]

thx, i did it and couldn't find any solution ... :(