Stat - Game Performance
 

I thought I would post some notes about game performance on my system for those still waiting on the game! Those who got it, please share your findings as well.

My system:
AMD Phenom x2 550 (3.1GHz, Dual-Core); 4GB RAM; ATI 5770 (1GB); Win7 x64

1680x1050 resolution, all details on full, AA setting at medium

Here is the FPS I get:

Alone at sea:
45-60 in external camera
55-60 on conning deck
average of 70 in sub interior (some closed-off areas like the torpedo room hit over 100)


At sea next to convoy:
5-10 FPS less than above figures. Very small FPS hit, however the time compression is a serious pain in the ass next to a lot of traffic (drags down the performance a lot).

At sea next to convoy with submarine flooding due to damage:
Same as above, except interior FPS now averages 35

In port:
30 in external camera
30 on conning deck
30-35 in sub interior

In bunker:
20-25 average (FPS hog, this)

Other remarks:
Loading times are LONG. Not as long as some later-day SHIII mods like GWX, but substantially longer (as much as 3-5 times) than SHIV.
TC, as noted above, is a big pain when a lot of contacts are nearby. 256 or 512 is a realistic maximum next to a port or large convoy (in SHIV I would get up to 1024-2048 in same situations). However in open seas with few contacts, TC performance is comparable to SHIV.

Overall, though, the game performs remarkably well. Aside from TC and loading times, the performance is very close to what I get in SHIV on my machine, with substantially more detail.

I concurr about loading times and TC chug. It's even harder to stop it a high rates than it was in SHIV (atleast for me).

Also, In past games you cold sweep your hydrophone quicker with 2x (or 4x) time compression. Not here.. It won't move any faster. And at times the needle seems to move a tad slower under TC.

Though i did find a work around to get the thing to sweep at normal speed under TC. Hold one of the hot-keys (Home/End), and spin the mouse wheel at the same time. It will move faster than it will using only one of the control methods.

is it actually using more than 1 cpu core?

and if you're going to post specs+framerates...i think the O/S you are using will be rather important for comparison too (including if its the 32/64bit version)

Oh right, edited on the OS...

AFAIK, the game does use both cores. System monitor shows both cores running about evenly.

I think some measure of multi-core support has been in here since SHIV though.

It was no rocket science to predict that loading times will be long. No fast internet connection can compete with direct loading from the hard drive on the player machine. Too bad for modders when it would start to make sense to do modding after the end of patches from UBI servers (if any).

Can my 2400 single core with 2 gb of ram and 2600 XT move the game ??

:D :hmmm: :doh: :wah: :dead:

havin minimum specs will surly suk no super mods without upgrade:down:

I have a Intel Duo core E8500, ati 4890, win XP. Running SH5 on a 27" widescreen Dell at 1920x1200. Max graphics option, but AA off. I have not checked the FPS, but have not noticed any slowdowns whatsoever, even in harbours with many ships or when watching explosions.

I don't think that's a server issue at all. The game doesn't load anything from the servers, unless you put save-games there (which are small). All the mission files are on the HD, and that's where it's loading those from. In fact I don't think it talks to the server at all when starting a new game.

I did not mean the actual loading of the game via internet, but rather an UBI server checking and trying to synchronize the version of the game and making sure that the game you bought is run only on one computer at the time.

Fair enough, although I actually wasn't talking about that... the game startup time isn't really different from SHIII/IV actually. The mission loads are the different story here.

But now that I looked at the campaign files, I think I understand why this is...

Your machine has a secret "Hardware Convoy Detector" installed!

Hehe, yeah. That's been the case since SHIII though. It's always a house rule for me to ignore this slow-down.

Not a valid test, even if you're probably right about it using multi-core. If the process were a monolithic single-threaded elephant, the OS would context-switch it to balance load on multi-cores and you'd see exactly what you'd describe (but it would run like a slug).

My guess is that the UBI servers try to synchronize both the version of the game and also each mission the player tries to load every time. Should be easy to check that by creating a small/tiny mission, unless the game always loads campaign files in the backgound as in SH3/SH4.