WEBVTT 00:00:07.720 --> 00:00:12.719 so guys I'm doing my Battlefield 3 benchmarking this weekend right now I've 00:00:11.000 --> 00:00:19.160 got the 6970 on the test bench so I'm using my 00:00:16.039 --> 00:00:21.680 core i7 2600k test bench I'm running the 00:00:19.160 --> 00:00:30.320 CPU at stock speed for this one uh the cards that I'm benching are the 590 580 00:00:24.199 --> 00:00:32.759 570 560 TI 550 6990 6970 6950 and the 00:00:30.320 --> 00:00:36.480 6870 so the settings I'm using are the high preset in Battlefield 3 just 00:00:35.120 --> 00:00:42.079 because I wanted something that was going to be not quite so highend that I 00:00:39.160 --> 00:00:47.399 might run into frame bu buffer issues I wanted to get a good idea of how the 00:00:44.200 --> 00:00:50.879 game performs on all these gpus relative 00:00:47.399 --> 00:00:53.719 to each other so high settings and I'm 00:00:50.879 --> 00:00:59.559 running at 1080p so my Benchmark sequence is on the level o I think it's 00:00:57.039 --> 00:01:04.680 called comrades something like that whatever this level's called and then as 00:01:02.079 --> 00:01:10.080 soon as I get out of the car here I'm going to start and he's going to shoot 00:01:08.520 --> 00:01:14.759 the guy and then I'm going to get out of the car so I'll start I'll do my run 00:01:11.960 --> 00:01:20.079 through and then I'm going to record 2 minutes of fraps gam 00:01:17.159 --> 00:01:25.680 play that I play the same way every time the repeatability of this particular 00:01:22.040 --> 00:01:28.520 Benchmark is uh can be done to within 00:01:25.680 --> 00:01:33.159 about 0.5 of a frame per second so that's what I would consider the 00:01:31.320 --> 00:01:38.720 thank you that's what I would consider to be the margin of error about5 to 1 00:01:36.119 --> 00:01:43.040 frame per second so I'm going to follow him I'm going to have to apologize in 00:01:41.040 --> 00:01:49.880 advance for my gameplay here my mouse sensitivity is off and 00:01:45.880 --> 00:01:49.880 uh oh cat go 00:01:51.759 --> 00:01:57.880 away okay hopefully that doesn't interfere you know what I'll probably 00:01:55.240 --> 00:02:01.719 have to redo this run off camera but I just want to show you guys the general 00:01:59.439 --> 00:02:05.399 run thing through whenever my cats don't interfere things go a little bit more 00:02:03.200 --> 00:02:08.720 smoothly so wait for the car to go and then I'm going to go pillar to pillar 00:02:07.200 --> 00:02:16.680 and try and see if there's any guys to shoot and just keep moving up until I 00:02:12.959 --> 00:02:19.120 gain this position here at which point I 00:02:16.680 --> 00:02:24.120 will fight these guys well rather shoot them while they're not looking and then 00:02:21.720 --> 00:02:27.879 I will fight the guys that come up the stairs duck back here a little bit kill 00:02:26.959 --> 00:02:32.120 this guy wow like you can tell I'm playing on 00:02:30.239 --> 00:02:36.599 easy right like one shot to the guy's hip and he goes down oh that's not 00:02:34.319 --> 00:02:42.519 usually the way I go run over here look over here come back here kill this dude 00:02:40.480 --> 00:02:47.640 now the autosave is kicked in this game was really hard to find a good place to 00:02:45.959 --> 00:02:51.239 uh to Benchmark for 2 minutes straight because what I really like to have is a 00:02:49.159 --> 00:02:55.959 checkpoint so that I don't have to sit through a cinematic sequence in between 00:02:53.239 --> 00:03:02.200 gpus every time but I wasn't able to find a single 2minute long uninterrupted 00:02:59.400 --> 00:03:04.920 segment M of play absolutely ridiculous and I killed those guys too quickly so 00:03:03.799 --> 00:03:11.159 I'm just going to kind of stand here and shoot a little bit because 00:03:08.000 --> 00:03:15.560 uh the fraps counter shouldn't reappear 00:03:11.159 --> 00:03:17.959 until I get to about right here there it 00:03:15.560 --> 00:03:20.840 goes just like that so so yeah it was very frustrating and I do have to sit 00:03:19.280 --> 00:03:24.480 through that car sequence every time a lot of the earlier levels especially are 00:03:22.720 --> 00:03:28.239 like in the dark and like there's nothing going on but I thought this was 00:03:25.799 --> 00:03:33.480 a pretty good gameplay sequence so uh without further ado I will present to 00:03:30.680 --> 00:03:37.640 you guys my my graphs once I'm finished rerunning this one and then I think I 00:03:35.040 --> 00:03:41.159 still have the 6870 to go so one thing that I did forget to mention when I did 00:03:39.159 --> 00:03:45.920 my initial video talking about the uh benchmarking procedures that I am using 00:03:43.280 --> 00:03:50.959 the latest driver so that is the Catalyst 00:03:46.959 --> 00:03:53.239 11.1 or rather 11.10 drivers from AMD 00:03:50.959 --> 00:03:58.760 and the GeForce 28562 drivers from NVIDIA so these are 00:03:56.319 --> 00:04:01.680 the blah blah blah blah blah you can read more about it this is is a 00:03:59.879 --> 00:04:06.319 recommended driver for Battlefield 3 and then I think on the ATI rather AMD 00:04:04.200 --> 00:04:11.200 release notes Battlefield 3 provides great Crossfire scaling and resolves 00:04:09.120 --> 00:04:17.440 issues when running in crossfire mode now here come the results and where's my 00:04:14.040 --> 00:04:19.759 Excel it's here there's my graph so 00:04:17.440 --> 00:04:25.240 Battlefield 3 FPS more is better I'm going to go ahead and give AMD major 00:04:23.240 --> 00:04:31.160 props because yes their Crossfire scaling is outstanding the 00:04:27.639 --> 00:04:35.360 6990 performs on average at 00:04:31.160 --> 00:04:38.880 114.1 FPS compared to the 6970 at 64.2 00:04:35.360 --> 00:04:40.960 now my 6990 was running without the OC 00:04:38.880 --> 00:04:48.560 switch enabled on it which means that that is near perfect Crossfire scaling 00:04:44.160 --> 00:04:50.560 simply outstanding uh rarely rarely seen 00:04:48.560 --> 00:04:56.680 that in any kind of a real world scenario from NVIDIA or from AMD uh so 00:04:53.759 --> 00:05:01.240 the 590 does scale but not as well as the 6990 and then here are the rest of 00:04:59.000 --> 00:05:06.639 the cards so for these ones bear in mind that these sort of equivalent GeForce 00:05:03.680 --> 00:05:09.720 and Radeon cards um equivalent in terms of performance usually the Radeon cards 00:05:08.400 --> 00:05:13.520 tend to be a little bit cheaper so I think AMD's done a great job but I look 00:05:11.400 --> 00:05:17.560 forward to seeing what NVIDIA can do to optimize their cards because generally 00:05:15.280 --> 00:05:21.840 speaking the way I have it arranged is the way I'm typically expecting to see 00:05:19.160 --> 00:05:25.600 the performance lineup and uh that is not what I observed with Battlefield 3 00:05:23.880 --> 00:05:31.520 we're seeing a serious equivalent see between the 580 and the 6970 the 570 and 00:05:28.120 --> 00:05:35.520 the 6950 the 560 TI and the 00:05:31.520 --> 00:05:38.880 6870 I'd say anything above a 550 is 00:05:35.520 --> 00:05:41.440 suitable for 1080p with all high presets 00:05:38.880 --> 00:05:45.800 uh but the 550 TI did slow down during firefights I know I don't have minimum 00:05:43.160 --> 00:05:49.000 FPS on this particular graph but this was the only one where the minimum FPS 00:05:47.600 --> 00:05:52.680 really bothered me all of the rest of them provided a smooth gameplay 00:05:51.120 --> 00:05:56.639 experience and thank you for checking out my Battlefield 3 Performance video 00:05:55.160 --> 00:05:59.759 see there's my notes where I have my benchmarks written down don't forget to 00:05:58.440 --> 00:06:06.080 subsscribe to lus techip from more unboxings reviews and other computer 00:06:03.080 --> 00:06:06.080 videos