• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Borderlands Gameplay Performance and IQ @ [H]

FrgMstr

Just Plain Mean
Staff member
2FA
Joined
May 18, 1997
Messages
58,488
Borderlands Gameplay Performance and IQ - Borderlands is a colorful co-op shooter in more ways than one, coming to retail after a lengthy development cycle including at least one redesign. It uses a familiar graphics engine, but with some tweaks and upgrades which have brought with it some unexpected performance challenges. We play Borderlands with 9 of today's best video cards.

Borderlands was launched by Gearbox Software and 2K Games in late October for the PC, Xbox 360, and PlayStation 3. Originally conceived as a gritty and realistic (before "gritty and realistic" became a code-phrase for brown and grey) post-apocalyptic shooter, Borderlands was re-designed as a colorful and cartoony off-world shooter. The PC version of the game has been generally well-received, garnering mostly positive press and a Metacritic score of 82.
 
was surprised to see the 5770 and 5750 suffer so much game runs perfect smooth on my 4870 512MB 1920 x 1200 with max effects turned on.
 
Nice review.

Just curious, any word on crossfire working/compatibility?

Both mine and Mark's testing seem to point to poorer performance with AMD video cards compared to NV video cards currently in this game. Is it simply a matter of drivers or is it more, who knows. We did point out this is a UE3 game and TWIMTBP, so take that as you will.
 
It wasn't around at the time, I'm sure.

As for performance differences, in some cases of some games, there's a noticeable performance difference depending on what mobo/cpu platform you're running your card of choice on. Saw a thread on Rage3D about that, specifically some rare Intel issue where it was unable to get as much out of (Nv?) GPUs as the AMD equivalent, even though the AMD platform has the slower RAM and weaker CPU, but it was a while ago and might have received a bios fix.
 
TWIMTBP..... It's an interesting delima. As a consumer am I really better off buying the "faster" card if all the big name games are going to be TWIMTBP and actually faster on the "slower" card. Whether I like the marketing strategy or not, think it is fair or not, think it is right or not, the fact doesn't seem to change that Nvidia's cards are better optimized for games and are currently giving better "clock for clock" results. It is a scary future if ATI doesn't do something to better their developer relations and driver optimizations.
 
This game clearly favors Nvidia cards. The 8800GTX actually beats the 4870 while the 4850 is surpassed by even the 9600GT.

As if PC gaming wasn't bad enough already - you essentially have to buy two complete gaming systems, one for "Nvidia games" and one for "Radeon games". Or at least have both video cards available so you can swap depending on who paid the developers' bills...
 
This game clearly favors Nvidia cards. The 8800GTX actually beats the 4870 while the 4850 is surpassed by even the 9600GT.

As if PC gaming wasn't bad enough already - you essentially have to buy two complete gaming systems, one for "Nvidia games" and one for "Radeon games". Or at least have both video cards available so you can swap depending on who paid the developers' bills...

but at the same time, nVidia cards experience some abnormal stutter in certain area.

there is few place that will cause it, one is nearby the female that you meet before going to the junk yard. the other one is in the junk yard...

there is still a few, I can't remember the exact places..
 
but at the same time, nVidia cards experience some abnormal stutter in certain area.

there is few place that will cause it, one is nearby the female that you meet before going to the junk yard. the other one is in the junk yard...

there is still a few, I can't remember the exact places..

Bugs abound.
 
meh its new modified unreal engine so nothign suprises me first batman now this. now sad thing is that unreal engine was the one where ati ruled before
 
You most certainly CAN use AA on ATI cards. Just set the MSAA= line to 2, 4 or 8 in the WillowEngine.ini files. I run at 1280x960 with 4xMS Adaptive AA on my 4850.

The results are the exact same as you see with Nvidia cards: edges are softened, but the black outlines are untouched.

One additional benefit you people did not mention: unlike a lot of recent DX9 games, Borderlands does not run transparency-AA filter on their foliage textures. But with MSAA=4 in the ini file, I can get the textures filtered by turning on Adaptive AA in CCC.
 
You most certainly CAN use AA on ATI cards. Just set the MSAA= line to 2, 4 or 8 in the WillowEngine.ini files.

I tried the INI setting, and it didn't do anything on NV cards or AMD cards. Granted, that was when I first started testing, and a patch has rolled out since then. It could also be a driver-specific problem. It would be wonderful if we were able to test every different configuration combination, but we would never be able to actually get anything done.
 
Last edited:
You most certainly CAN use AA on ATI cards. Just set the MSAA= line to 2, 4 or 8 in the WillowEngine.ini files. I run at 1280x960 with 4xMS Adaptive AA on my 4850.

The results are the exact same as you see with Nvidia cards: edges are softened, but the black outlines are untouched.

One additional benefit you people did not mention: unlike a lot of recent DX9 games, Borderlands does not run transparency-AA filter on their foliage textures. But with MSAA=4 in the ini file, I can get the textures filtered by turning on Adaptive AA in CCC.

At any rate, gamers shouldn't have to edit INI files to set such simple things as AA, it should be officially supported through either in-game options or by forcing from driver control panel.
 
At any rate, gamers shouldn't have to edit INI files to set such simple things as AA, it should be officially supported through either in-game options or by forcing from driver control panel.

But that seems to be getting less and less common... officially supported AA I mean.

Nice review though, I always love seeing the side-by-side pictures of different settings. For reference as well, I'm playing on my mobile 8800m GTS @ 1440x900, all settings on/high and it plays beautifully. I love this game.
 
But that seems to be getting less and less common... officially supported AA I mean.

Nice review though, I always love seeing the side-by-side pictures of different settings. For reference as well, I'm playing on my mobile 8800m GTS @ 1440x900, all settings on/high and it plays beautifully. I love this game.

Not common on UE3 based games at least.

I too would like greater support of AA in games.
 
Not common on UE3 based games at least.

I too would like greater support of AA in games.

It seems to have gotten better for a while.. but there was quite a few UE3 games that didn't support it, wasn't there? The R6:Vegas games come to mind.
 
It seems to have gotten better for a while.. but there was quite a few UE3 games that didn't support it, wasn't there? The R6:Vegas games come to mind.

That was primarily a DX9/DX10 issue with the deferring lighting model used by UE3. The proliferation of Vista, Windows 7, and DX10/11 hardware has largely alleviated that particular problem.

But that seems to be getting less and less common... officially supported AA I mean.

I think it is caused by the change in focus in the industry to focus primarily on console games, and compounded by the differing implementations of AA by AMD/ATI and NVIDIA.
 
I think it is caused by the change in focus in the industry to focus primarily on console games, and compounded by the differing implementations of AA by AMD/ATI and NVIDIA.

I'm sure that's the main problem. It's too bad, but it's getting also getting to the point where the PC users have to be happy if we even get the port of certain games.
 
*yawns*

Where is the Dragon Age Gameplay Performance and IQ @ [H]? That game gives very interesting results with slower quads; it blow the doors off faster dual cores.
 
That was primarily a DX9/DX10 issue with the deferring lighting model used by UE3. The proliferation of Vista, Windows 7, and DX10/11 hardware has largely alleviated that particular problem.

But the game has to support DX10 fully to allow MSAA in UE3 games. That is why we have this issue to being with.

The source of the problem is, Epic assumed that game developers would be happy to create a DX10 version of their engine, despite the fact that console tech is stuck at DX9. Epic left us PC gamers up a creek without a paddle because most developers are lazy, and don't want to put in the effort for a full DX10 port.
 
But the game has to support DX10 fully to allow MSAA in UE3 games. That is why we have this issue to being with.

The source of the problem is, Epic assumed that game developers would be happy to create a DX10 version of their engine, despite the fact that console tech is stuck at DX9. Epic left us PC gamers up a creek without a paddle because most developers are lazy, and don't want to put in the effort for a full DX10 port.

I'm pretty sure that DX10+ hardware on a DX10+ platform can force AA for deferred renderers running in DX9. It will vary in implementation from game to game, of course.
 
Almost seems like we are hitting a case with the various AA, the best solution would be the older super sample AA(which does wonders in a variety of games that I've played it with in the hd5870), except for the completely massive performance hit it causes.
 
Hmm... surprising in some parts, not surprising in others. I'll have to go back and re-do my benchmarking, I just reached that area Saturday night when I decided it was just time to beat the damn game already.
 
Shame this Article does not test a dual core CPU s, many games are still running.

From my own testing anything below a Core2 Duo at 2.8GHz will start to bottleneck cards like a GTX285 even at even 1920x1080 with both cores running around 100%.

Its another rare game that truly benefits from a quad core.
 
Just a note for those running older video cards looking to 'upgrade' w/20-22 inch display. I just benched the same area (multiple run throughs) on a 4870 and 720BE tri core and out performed all cards listed in the review at that resolution. Taking the lowest fps of each of my runs I hit a min of 37, my lowest max was 107, and using the same run though that got me my lowest fps point; that average run was 64.9. If you're stuck at 1680x1050 world of gaming like me and think you might need an upgrade to play Borderlands, don't worry... you're probably fine.
 
was surprised to see the 5770 and 5750 suffer so much game runs perfect smooth on my 4870 512MB 1920 x 1200 with max effects turned on.
Same here, I'm playing maxed out at 1080p with a GTX 260 Mobile card (rough equivalent of an 8800), even forced some AA, not that it made things look any better.
 
This kind of killed the hopes I was having for a 5770 Crossfire upgrade :(
 
With the rig in my sig Borderlands runs awesome at 1680x1050 all quality max (used .ini to max everything) except AA. 90% of the time I am over 80fps. Very rarely does it go below 50.
 
I don't quite understand the screwy results at 1920x1200. I play at 1920x1080 on a 4890 in an old system (Athlon 64 x2, 2gb DDR) and it is more than playable with everything maxed. I would hope that on 5770, or 5750 for that matter performance would be similar if not better:confused:
 
I don't quite understand the screwy results at 1920x1200. I play at 1920x1080 on a 4890 in an old system (Athlon 64 x2, 2gb DDR) and it is more than playable with everything maxed. I would hope that on 5770, or 5750 for that matter performance would be similar if not better:confused:

The 5770 is just shy of the performance of a 4870 (which itself is slower than the 4890), and the 5750 is just barely north of a 4770.

Plus, we know from other experiences that the HD 5000 series drivers have a lot of maturing to do.
 
Really? The 275 again? There have to be crazy amounts of us w/ 260's gaming on 20" inch displays.

Tried out the game on PS3 and did seem fun, but the vertical split screen and lack of clarity definitely took it's toll. So how is everyone liking this on PC? Is it worth sitting in a chair again? :D
 
Really? The 275 again? There have to be crazy amounts of us w/ 260's gaming on 20" inch displays.

Tried out the game on PS3 and did seem fun, but the vertical split screen and lack of clarity definitely took it's toll. So how is everyone liking this on PC? Is it worth sitting in a chair again? :D

I'll let you know when my copy arrives from Amazon. Jumped on the $19.99 special. Will be playing on 5970. =)
 
Not so much a comment on this article, but more on all your articles as a whole.

I was wondering if you guys could add a price column on the test set up page. I know the prices are always changing and can be different at different vendors, but if you picked one vendor and always picked the lowest priced card or the average card price, and have a note stating prices are based on the time of writing and may be different. Sometimes you mention the price of cards in the intro, or conclusion, but for the most part on the reviews the reader is left to find the price on his own.

I realize, especially on newer cards, that they have a "set" (msrp) price and usually hover around that, but I can't tell you what a 275 goes for atm, while I know the msrp of a 5870. It would help me as the reader better compare each card if I knew the price differences between each card, card A performs a little better but costs twice as much, while card C is a little slower then card b, but half its price. I realize this price extremes aren't likely, but being able to see pricing information next to performance information would help me better make a decision on which card is better, and which card is a better bang for my buck. I know in the heatsink articles you talk about price, and even compare the best bang for your buck in a chart, but fail to even list the price for a video card, unless you happen mention it in the conclusion.

Adding the price to the names of each card in the title bar on each chart would also be nice, to help put performance and price together, and help readers get a better idea on bang/buck.

I've added some pictures to give you an idea of what I am talking about. Note: I have 0 paint skills, and made up some numbers, but you get an idea of what it would look like.



 
Back
Top