Jump to content

V 3.0 performance discussion


Recommended Posts

  • Replies 237
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

Posted Images

well speaking personally, I have a lowley HD6670 with 1 gig RAM coupled with an e8400 dual core. nothing steel beasts does troubles the core temps of either processor or graphics card (ie not huge amounts of processing is being done). however... give it a very busy map like the new hunfeld bundle and my framerate drops to sub 20 (or sub 10 in some cases!). I have eliminated the main RAM as being the problem so the only remaining bottleneck is the graphics card RAM.

Ok you still have double what my card is giving out (not to mention a better set up overall) but by a process of elimination you may come to the same conclusions as I have

Link to comment
Share on other sites

You can run GPU-Z to monitor GPU usage, and you'll probably find the same thing I have: it's not the GPU that's limiting the frame rate on that fulda scenario.

My CPU's an old Core i7-860, overclocked so that it runs between 3.4 and 4 GHz, depending on how many cores are busy, and the GPU's a GTX 760. If I run that fulda mission with the default detail settings, in windowed mode with just a bit of a gap on the right so I can monitor GPU-Z and the task manager, the frame rate will drop to around 20 or so FPS during the first part of that scenario. The GPU was loafing along at the minimum core speed of 135 MHz with only about 30-35% load while the CPU load from SBProPEcm.exe maxing out a core and then some (two threads were busy, the main SB one and one running nvd3dum.dll, nVidia's Direct3D driver.)

Link to comment
Share on other sites

  • Members
Wow, I'd be really surprised if my Radeon is the bottleneck! In the sticky, the GTX670M-i7 3720 combo is rated for Medium performance for levels 4 & 5 which means no framerate below 30fps.

Note that this is a scenario with a very light CPU load. Note also that the CPU performance does also influence the overall effectiveness of the graphics card. A GPU that is waiting for the CPU to finish some basic calculations cannot operate with its maximum computational brute force.

I'm not sure however how useful this theoretical debate is. I stick to my recommendation that you should wait until SB Pro PE 3.0 is out and then see for yourself if the performance can be made to be acceptable (by changing certain options) or not.

Which begs the question, will that benchmark eSim is using actually be a scenario that's going to be included in 3.0?

It's not really a "benchmark" in its closer sense (something that runs more or less automatically on rails through a certain sequence of scenes and then spits out a performance index). It's a rather primitive scenario of a mech platoon rolling into a village with some population filling the streets, and a sequence in which three screenshots are to be taken

  • from the external observer's position with no buildings nearby,
  • the commander's unbuttoned view at the village's edge, and
  • from the gunner's sight along the main street with the troops dismounted and pedestrians roaming the street.

From these three scenes the average frame rate is taken as well as the minimum frame rate. That's it.

Link to comment
Share on other sites

I'm not sure however how useful this theoretical debate is. I stick to my recommendation that you should wait until SB Pro PE 3.0 is out and then see for yourself if the performance can be made to be acceptable (by changing certain options) or not.

I think it's better to ignore this whole discussion and wait to try it out, although I understand that people are enthusiastic, the only sense I can make out of it is that the increased detail would likely require more hardware or more tradeoffs the player is going to make. Never mind the different hardware configurations, there are all kinds of different scenarios ranging from simple to complex with few to many actors, a whole range of visibility settings out to 18 km, different terrain detail, a few houses to large towns, sliders to adjust ground clutter-

quite simply, there is no objective standard for everyone to relate his subjective experience to. Since it's nigh impossible to come up with a baseline everyone would agree on for all occasions, the discussion is futile. A player can only decide for himself what he likes, make any adjustments according to his preferences. To that extent, he'll have to wait and see.

Link to comment
Share on other sites

Well “wait until the software actually comes out and then see what your performance is like for you” has been mentioned a few times but I guess nature abhors a vacuum so people will try and pre guess anyway.

Whether Scrapper_511 was wasted his money or not he wont know until after the software is released.

But its his money and if he wants to spend it now rather than wait is up to him.

Link to comment
Share on other sites

I think it's better to ignore this whole discussion...

Well that sucks. :wink2:

...I understand that people are enthusiastic, the only sense I can make out of it is that the increased detail would likely require more hardware or more tradeoffs...

Absolutely enthusiastic! Shadows under vehicles is something I have really longed for since SB1. That said, please understand my concern for my hardware's ability to run 3.0 nicely at Level 2...at least!

Another similar concern is for the new particle effects. Better pyrotechnics is the one other visual improvement that I've looked forward to.

Never mind the different hardware configurations, there are all kinds of different scenarios ranging from simple to complex...

The Performance sticky is there for good reason and it did reveal that my previous setup would've been unable to run well at just Level 2.

...there is no objective standard for everyone to relate his subjective experience to. Since it's nigh impossible to come up with a baseline everyone would agree on for all occasions, the discussion is futile.

I consider the standard as the "scenario" used for the Performance sticky. I also mentioned which scenario I've been using (and how) as a "benchmark". My comparisons and observations are strictly derived from these two "baselines".

A player can only decide for himself what he likes, make any adjustments according to his preferences. To that extent, he'll have to wait and see.

Yes, and I do apologize for my inability to shorten my winded posts. Not knowing for certain if my new CPU (and to a lesser extent, my new GPU) will be able to run 3.0 at my preferred visual settings has got me very anxious. It had better run well (with ground shadows and particle effects) because I won't be upgrading again any time soon!

Worst case scenario I'll have to limit myself to simpler scenarios, or as suggested elsewhere, just forego the fancy special effects and embrace 3.0 for its new engine and new vehicles/features. :smile:

Your reply is well read, Captain. I pretty much disected your reply, but please don't take it as a flame. I thank you for letting me clarify myself. :thumbup:

Link to comment
Share on other sites

Well “wait until the software actually comes out and then see what your performance is like for you” has been mentioned a few times but I guess nature abhors a vacuum so people will try and pre guess anyway.

Well that sucks too. :debile2:

Yeah, I couldn't wait. One thing that led me to upgrading now is the assumption that I will be upgrading the CPU at some point anyway. The Phenom X4 970 is an improvement over my older Athlon X2 and is just about the fastest CPU my motherboard will take. For $125 at Newegg I'm glad I bought it when I did because they're now out of stock.

Link to comment
Share on other sites

I consider the standard as the "scenario" used for the Performance sticky. I also mentioned which scenario I've been using (and how) as a "benchmark". My comparisons and observations are strictly derived from these two "baselines".

Right so have you made some objective comparisons?

Now that Ssnake has given a rough understanding of it being a mech platoon (so four vehicles) moving into a small town and the dismounts patrolling in a mixed population, how does that compare with your Fulda Gap scenario?

Do you have more than four vehicles in total (so Blue plus Red)?

Do you have artillery falling (because its not in the eSim baseline)?

How “busy“ is your scenario?

Link to comment
Share on other sites

Right so have you made some objective comparisons?

Now that Ssnake has given a rough understanding of it being a mech platoon (so four vehicles) moving into a small town and the dismounts patrolling in a mixed population, how does that compare with your Fulda Gap scenario?

Do you have more than four vehicles in total (so Blue plus Red)?

Do you have artillery falling (because its not in the eSim baseline)?

How “busy“ is your scenario?

I felt that the much more intensive Fulda Gap scenario would compensate for the lack of new shadows and particle effects. But more so I felt that a scenario that taxes my hardware would be a good starting point to compare performance before and after upgrades.

Link to comment
Share on other sites

Absolutely enthusiastic! Shadows under vehicles is something I have really longed for since SB1. That said, please understand my concern for my hardware's ability to run 3.0 nicely at Level 2...at least!

Sounds like you've already committed to new hardware, but I'm pointing out you still will not really know what it looks like based on these descriptions until you see it in motion yourself. I would wait and see- at the very least, if you determine that you would like to buy new components after you try it out first, prices may come down a little by then. It sounds as though your primary reason to upgrade is for SB 3.0, so what's the rush to buy new hardware before you've seen anything yet?

Link to comment
Share on other sites

  • Members

...and here we are again, discussing someone else's discretionary spending - well after the decision has already been made, and action been taken. What is this supposed to accomplish, aside from instilling doubt and making him feel bad about it?

Link to comment
Share on other sites

Okay...I have to admit that I'm a little concerned about my ability to run 3.0, now that I've stuck my head over in the performance threads. But just a little bit.

Currently I've got an AMD FX-8120 (3.1Ghz, 8 cores), 4gb of system RAM, and an NVIDIA GT610 with 1gb of video RAM. It's a brand new system, my old one sort of gave up the ghost. I was on a limited budget, so I bought the biggest processor I could afford and planned up upgrading RAM and video card if needed later.

Unfortunately I'm still on 32bit XP Pro (It's the most modern copy of Windows I own) so the OS doesn't know how to use more than 4gb of RAM, and I'm limited to using versions of DirectX that duplicate my video card RAM into system RAM, so my 1gb video card actually leaves me with only 3gb of RAM free for other stuff, if I understand correctly.

It runs the current version without any problems, but I haven't tried 'huge tracts of map', nor incredibly complex scenarios yet. If I need to do it to run 3.0, I can always go out buy a 64bit operating system, throw in some more RAM, and get a better video card.

For comparison sake, my previous system was a 3.0Ghz Pentium 4 (yes that's a single core) with 3gb of RAM and an HD5450 video card with either 512mb or 1gb of video RAM. I bought the 512mb model when I bought my upgrades, but I think they actually sent me a 1GB model.

That also ran the current version well, but I did make absolutely sure to eliminate all unnecessary background processes, get my drive defragged, turn down all graphics settings to minimum, and avoid huge maps and scenarios. I could still do pretty big scenarios, but I sometimes noticed slowdowns, particularly when using the low magnification in the TIS and overlooking lots of stuff. I had some problems in the mission editor and map editor as well - very large maps just wouldn't load.

Link to comment
Share on other sites

Well again my suggestion is "Don't Panic!" (at least not yet).

When it comes out borrow a license and try it out.

THEN panic IF you have to.

The way I see it, I've already sunk enough cash into Pro PE (IIRC I paid $100 for the dongle and license, then another $30 or so for a second license) that it doesn't make sense to NOT buy 3.0 just because I'm worried about hardware.

Besides, this PC is brand new, and I knew when I bought it that I was buying a PC that was short on RAM and video card because I spent all my cash on the processor. I was planning on upgrading RAM and video card as soon as I had the cash for it anyway.

Hopefully they will find some other ways to make it more compatible for older hardware. For instance, they might offer an option to revert to the older less intensive 3D models if they are available, etc.

Link to comment
Share on other sites

be aware that SB Pro PE is basically SB Pro dressed in civies. If the military customers that use SB Pro are anything like the MOD. I doubt they would be rushing around to spend hundreds of pounds/dollars/euros on new hardware just to get some shadows.

Link to comment
Share on other sites

What is this supposed to accomplish, aside from instilling doubt and making him feel bad about it?

There are well adjusted, mature adults here who have other things to do with their time than to make someone feel bad about their purchase- I know certainly that I'm not so cynical or petty as that. In fact, my comments originally were not aimed at anyone in particular until Scrapper responded to them.

I wish Scrapper the best and hope he enjoys. However, to reiterate because a sort of anxiety has a way of spreading like panic, there is no way to get a feel for what this will look like without hands on experience. I started following this discussion and then quickly understood it really won't give me much certainty. I may or may not find the shadows effect that much worthwhile compared to other settings- for example more vegetation detail, further visibility settings, I won't know at all until I've seen it in for myself. Maybe Ssnake tolerates better framerate performance than I do, I'm not Ssnake, even Ssnake's objective reporting that he can get so many FPS with an arbitrary setting as satisfactory to Ssnake may not mean that I will come to the same conclusion. It's too subjective.

Link to comment
Share on other sites

... instilling doubt and making him feel bad about it?

I'd think by now that most gaming/computer enthusiasts would have accepted that even newly released computer hardware is already obsolete before it makes it to your door. :D

Like everyone else said, I wouldn't worry about it until 3.0 comes out and you see how everything runs when playing "naturally" instead of just doing stress tests.

Link to comment
Share on other sites

...and here we are again, discussing someone else's discretionary spending - well after the decision has already been made, and action been taken. What is this supposed to accomplish, aside from instilling doubt and making him feel bad about it?

Thanks for looking out Ssnake. It is what it is. I hope 3.0 is coming along well enough. :thumbup:

Link to comment
Share on other sites

be aware that SB Pro PE is basically SB Pro dressed in civies.

That’s a pretty generalist statement based on ... ?

Different armies use the product for different training outcomes.

Some use it as a crew procedural trainer (see the Danish video recently published) where use of the 3D world is probably (again I’m not a Dane) a large proportion of the time.

For them graphics may well be important. They have also built facilities (buildings) around the computers so a computer refresh is probably not a big part of the equation and I guess if the had to upgrade/replace machines to achieve the same level of training they would.

The 3D world dropping from say 40 FPS to say 5 has a massive impact on gunnery drills, etc.

Other countries use it more for “post H hr” decision making (i.e. how your plan goes once you cross the LD). Those countries might use the 3D world for 10% of the training session and focus on F5 and the AAR.

For them how “pretty” the 3D world is, is not that important (shadows can be “off”, detail scaled back, etc. to support the equipment in a given location [dedicated classroom with desktops or portable laptop based suite]).

So I wouldn’t hang your hat on militaries requiring it to run on “old’ gear and that therefore your “old” gear will be fine too. ;)

Link to comment
Share on other sites

  • Members
Okay...I have to admit that I'm a little concerned about my ability to run 3.0, now that I've stuck my head over in the performance threads. ...

Currently I've got an AMD FX-8120 (3.1Ghz, 8 cores), 4gb of system RAM, and an NVIDIA GT610 with 1gb of video RAM.

I think it'll do fine.

1 GByte video ram should allow for shadow mapping level 3, maybe even 4. :)

Link to comment
Share on other sites

I think it'll do fine.

1 GByte video ram should allow for shadow mapping level 3, maybe even 4. :)

You're kidding, right?

I mean, yes I have 1Gb of video ram, but the card itself is just a cheap little GT610...

ETA: For the record, I am perfectly happy with the current level of graphical goodness.

I'm more excited by the additional vehicles in 3.0 than anything else...

2236083.jpg

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...