WEBVTT 00:00:07.799 --> 00:00:15.559 so for this video I'm going to be taking the highest performing of the highest 00:00:11.880 --> 00:00:18.640 performing The Most Extreme of the EXT 00:00:15.559 --> 00:00:21.920 no not okay uh actually this video is 00:00:18.640 --> 00:00:24.599 going to be all about onboard Graphics 00:00:21.920 --> 00:00:31.400 because AMD has just released their new A8 3850 as well as their A6 30 something 00:00:28.720 --> 00:00:36.120 something uh well any anyway the A8 3850 is the one that I'm having a look at 00:00:32.719 --> 00:00:38.879 it's their new Apu and AMD claims has 00:00:36.120 --> 00:00:42.280 onboard discret level direct X11 Graphics so we're going to take that and 00:00:40.600 --> 00:00:49.399 we're going to put it to the test against Intel's core I3 00:00:45.480 --> 00:00:51.520 2105 so this uses Intel HD 3000 graphics 00:00:49.399 --> 00:00:55.359 which is the higher end of the graphics on the Sandy Bridge processor so this 00:00:53.719 --> 00:01:00.800 has the same onboard Graphics as the 2600k and the 2500k but it is a dual 00:00:58.719 --> 00:01:07.240 core lower-end chip that actually is priced similarly to the A8 3850 now for 00:01:05.000 --> 00:01:12.320 a third point of comparison I'm also going to be taking a Radeon 00:01:09.920 --> 00:01:15.600 6670 and I'm going to be plugging it in in dual Graphics mode this is the wrong 00:01:14.240 --> 00:01:21.119 motherboard but it's around here somewhere in dual Graphics mode with the 00:01:17.840 --> 00:01:23.960 A8 3850 to see how much of a performance 00:01:21.119 --> 00:01:28.040 boost we can get by adding a discrete card in dual Graphics mode which used to 00:01:25.759 --> 00:01:32.520 be called hybrid Crossfire so I'm going to be running at 1280 x 720 with a 00:01:30.280 --> 00:01:35.600 variety of games the first one is Dirt 3 if you come in and have a closer look 00:01:33.920 --> 00:01:41.040 you'll probably see that it looks like crap because it's running at 720p with 00:01:38.000 --> 00:01:43.960 everything set to low but the reality of 00:01:41.040 --> 00:01:48.680 it is I want to get reproducible results where I can actually show scaling 00:01:46.079 --> 00:01:52.479 between all of these different Graphics Solutions so I have to I have to have 00:01:50.759 --> 00:01:55.079 something reasonably playable or especially on games where I actually 00:01:53.920 --> 00:02:00.240 have to play through it's going to be very difficult to do consistent uh 00:01:57.240 --> 00:02:02.280 benchmarking runs so stay tuned for more 00:02:00.240 --> 00:02:08.000 and I will be showing you these three solutions up against each other boot wot 00:02:06.039 --> 00:02:11.400 Battlefield Bad Company 2 is a game that looks like it was released like check 00:02:10.080 --> 00:02:14.760 check out check that out check out that Alias thing looks like it was released 00:02:13.280 --> 00:02:20.800 like 8 years ago when you play it on onboard Graphics see here's my fellow 00:02:17.800 --> 00:02:22.160 characters such crappy textures but 00:02:20.800 --> 00:02:27.480 that's what I have to do to Benchmark on board so here we go for crisis 2 00:02:24.239 --> 00:02:30.160 ironically the lowest Graphics setting 00:02:27.480 --> 00:02:36.160 is high because they subscribe to the good better best even better than that 00:02:32.800 --> 00:02:40.400 uh system of doing things obviously 00:02:36.160 --> 00:02:42.800 direct X11 is off with the cor I 00:02:40.400 --> 00:02:47.280 32105 because it does not support direct X11 so here we go crisis 2 is the next 00:02:45.400 --> 00:02:52.760 in my list of oh I haven't even written it down yet 00:02:48.879 --> 00:02:54.800 crisis 2 games to 00:02:52.760 --> 00:02:58.840 Benchmark okay my Benchmark runs finished in crisis 2 but I just wanted 00:02:56.480 --> 00:03:03.519 to show you guys how utterly rubbish it looks with uh 00:03:01.319 --> 00:03:07.400 onboard Graphics so here I'll zoom in as far as I can so you guys can kind of see 00:03:06.080 --> 00:03:11.599 what I'm talking about here there's basically no substantiality to any 00:03:10.120 --> 00:03:16.720 textures at all you see that blurriness that's not from my camera that's just from 00:03:14.000 --> 00:03:21.200 like um being zoomed in but like you just you can't make out anything it's 00:03:19.159 --> 00:03:26.720 kind of terar bad like look at look at these sandbags it just looks like big 00:03:24.200 --> 00:03:30.879 white blobs and performance isn't any good 00:03:28.920 --> 00:03:36.159 either for The Witcher 2 I will be using the 00:03:32.840 --> 00:03:38.640 lowest presets once again at 1280 x 720 00:03:36.159 --> 00:03:42.920 which is technically HD but by the time you set everything to low and are 00:03:40.720 --> 00:03:47.879 running on a 1080p screen so you're dealing with interpolation 720p doesn't 00:03:45.000 --> 00:03:52.000 look very HD at all we'll see how it goes here's the results so far with the 00:03:50.680 --> 00:03:59.319 I3 2105 that's 20.0 not 200 under crisis 2 00:03:55.560 --> 00:04:02.680 by the way Witcher 2 runs pretty terad 00:03:59.319 --> 00:04:04.680 on the uh 2105 as 00:04:02.680 --> 00:04:10.400 well it actually doesn't look bad through the camera but I assure you it's 00:04:06.879 --> 00:04:13.319 not good combat in particular is uh 00:04:10.400 --> 00:04:17.400 pretty choppy you can see the frame rate up 00:04:15.400 --> 00:04:21.639 there ah it's been a while since I've run Mafia 2 as a benchmark so here's the 00:04:19.239 --> 00:04:25.720 settings I'm using all low 720p good stuff and then I'm just going 00:04:23.720 --> 00:04:28.800 to use the uh oh I got to restart the game here I am in the middle of 00:04:27.120 --> 00:04:33.880 benchmarking my cat wants me to throw his toy for him so I will this is his 00:04:38.560 --> 00:04:42.400 toy so now he'll go get 00:04:43.639 --> 00:04:49.560 it sometimes he brings it back sometimes he knocks it down the 00:04:47.840 --> 00:04:55.919 stairs and chases it on his own here's the big finale in the Mafia 2 00:04:53.240 --> 00:04:59.160 Benchmark where's the explosion oh never mind I missed 00:04:57.000 --> 00:05:02.400 it okay well there you go there's the results it's been a while since I've 00:05:00.759 --> 00:05:05.680 benched Black Ops but I figure it's appropriate it's fairly mainstream game 00:05:04.120 --> 00:05:09.000 this is a very mainstream sort of platform onboard Graphics all that stuff 00:05:07.800 --> 00:05:14.680 check out this guy's shirt yeah yeah we're not even getting 00:05:11.680 --> 00:05:16.680 like fantastic frame rates even so but 00:05:14.680 --> 00:05:21.919 yeah we'll see how it compares to the Apu for to come so here I'm 00:05:19.560 --> 00:05:25.800 configurating configurating configuring the integrated Graphics uh forcing 00:05:23.680 --> 00:05:29.319 multim monitor I'm setting everything at the highest I've got 8 gigs of system 00:05:27.240 --> 00:05:34.080 memory so I have no problem with 2 gigs on the uh frame buffer so let's see how 00:05:32.160 --> 00:05:39.560 this goes so now that I'm in Windows with my 6670 I can see that Crossfire is 00:05:37.440 --> 00:05:46.039 enabled with the Apu if I go into the hardware you can see I got my 00:05:41.759 --> 00:05:49.800 6550d my hyper memory I got my oh where' 00:05:46.039 --> 00:05:52.199 it Go 6550d linked adapter built by oh 00:05:49.800 --> 00:05:54.479 okay that's funny it said 6670 just a minute 00:05:56.360 --> 00:06:00.360 ago components 00:06:01.240 --> 00:06:06.759 display 6600 series there we go okay 00:06:04.319 --> 00:06:09.800 good enough so let's run some games now I want to take a short break in my 00:06:08.039 --> 00:06:17.199 testing to show you guys something this is the Aspen canned Benchmark for Dirt 3 00:06:13.440 --> 00:06:19.639 I'm running it on the 669 6670 with the 00:06:17.199 --> 00:06:23.599 Apu running in crossfire X mode now check out these frame rates so I'll wait 00:06:21.639 --> 00:06:30.080 till the race is actually running but when I ran it at 720p low settings it 00:06:27.840 --> 00:06:33.360 destroyed the core I 321 I mean that makes sense we've got a dedicated 00:06:31.360 --> 00:06:37.280 graphics card in there and you but you can see here I am 00:06:35.759 --> 00:06:42.479 still still 00:06:39.039 --> 00:06:45.680 demolishing the 2105 and I've actually 00:06:42.479 --> 00:06:48.280 turned the game all the way up to 1080p 00:06:45.680 --> 00:06:52.000 all high details now so you can see that it's much sharper looking there's a very 00:06:50.280 --> 00:06:56.960 noticeable difference I mean it looks like a racing game from you know 8 years 00:06:54.759 --> 00:07:01.000 ago when you turn the uh when you turn the graphics way down whereas now it 00:06:58.560 --> 00:07:05.560 looks like a Mod racing game with uh pretty decent Graphics so that's uh 00:07:04.440 --> 00:07:09.560 that's the difference it makes I just wanted to show you guys that actually 00:07:07.560 --> 00:07:14.160 the benefit is not necessarily in terms of frame rates from adding these 00:07:11.240 --> 00:07:18.000 Technologies but actually also in terms of dramatically improved graphical 00:07:16.639 --> 00:07:23.400 quality you can see it's still going to outperform it look at that okay guys so 00:07:20.400 --> 00:07:26.800 I'm doing my final run with Dirt 3 with 00:07:23.400 --> 00:07:28.639 the A8 3850 so I'm ready to present my 00:07:26.800 --> 00:07:31.879 results pretty much so I'll talk about everything so far then I'll show you 00:07:30.120 --> 00:07:37.520 guys the results of this one so I've got my 6670 here I've got my corai 3 over 00:07:35.479 --> 00:07:43.400 here and let's have a look at the results so this is my onboard Graphics 00:07:39.759 --> 00:07:45.120 showdown at 1280 by 720 I'm using all 00:07:43.400 --> 00:07:49.800 modern games so yes you're going to see poor scores here but the reality of it 00:07:47.479 --> 00:07:54.400 is these are designed to be run with modern gaming graphics cards not with 00:07:52.039 --> 00:07:58.240 modern onboard so you know they all did their best however I would like you to 00:07:56.560 --> 00:08:02.960 notice something right away so first of all here's the core I 321 105 we have 00:08:00.879 --> 00:08:07.840 for the most part not very playable frame rates we're at that best around 00:08:05.120 --> 00:08:11.800 that 30 FPS range and then in Dirt 3 we managed to get well over it whereas with 00:08:10.280 --> 00:08:16.360 the A8 3850 you can see we're for the most part 00:08:14.759 --> 00:08:20.960 well into that playable range the exception being Mafia 2 so that is the 00:08:19.360 --> 00:08:27.960 difference in performance between the core I 32105 and the A8 3850 in games so 00:08:25.199 --> 00:08:32.240 this is at a very GPU bound resolution but we're looking at more than double 00:08:29.759 --> 00:08:37.120 the performance in Dirt 3 double the performance in Bad Company 2 almost 00:08:34.360 --> 00:08:44.159 double in crisis 2 Double in The Witcher 2 more than double in Mafia 2 and um 00:08:41.440 --> 00:08:50.360 double again in Call of Duty Black Ops so that is the performance Delta between 00:08:46.480 --> 00:08:55.080 the I3 2105 and the 83850 now this was a 00:08:50.360 --> 00:08:56.760 bit of um uh an error on my part but at 00:08:55.080 --> 00:09:01.240 such low graphical settings we're not seeing much of a benefit from this 00:08:58.880 --> 00:09:06.600 additional graphics card so what I wanted to show you guys was yes it does 00:09:04.000 --> 00:09:10.959 consistently beat the non- graphics carded version especially see where 00:09:08.160 --> 00:09:15.519 we're GPU bound it absolutely blows it away it adds about double again the 00:09:12.800 --> 00:09:19.240 performance of a core I 32105 where we're not GPU bound anymore and we're 00:09:17.320 --> 00:09:23.440 running into CPU bound scenarios it doesn't make much of a difference see 00:09:21.079 --> 00:09:29.480 this so here in crisis 2 it makes quite a big difference as well so what I did 00:09:25.320 --> 00:09:31.519 was I actually ran um grid grid 3 Dirt 3 00:09:29.480 --> 00:09:36.519 at 1080p all high with zero anti-aliasing and I got about 60 FPS so 00:09:34.680 --> 00:09:40.440 you can see that we only lose about half of the frame rate versus running it on 00:09:38.880 --> 00:09:46.839 all low 720p however when we take that same game 00:09:44.079 --> 00:09:53.160 and we run it with the A8 3850 which is what I just did we get 00:09:48.959 --> 00:09:56.200 29.7 so you can see that when we are GPU 00:09:53.160 --> 00:09:59.040 bound the crossfire configuration can 00:09:56.200 --> 00:10:03.839 perform up to double of the 3850 on itself which is in 00:10:00.920 --> 00:10:09.800 turn pretty consistently about double a COR 32105 so the reality of it is the 00:10:06.839 --> 00:10:15.640 CPU component is not as powerful as a 2105 in pure um especially dual threaded 00:10:13.959 --> 00:10:22.240 benchmarks it's just not even going to come close but the A8 GPU just or the 00:10:20.839 --> 00:10:28.079 Radeon HD 6550d just destroys the Intel HD 00:10:25.360 --> 00:10:33.720 graphics 3000 so thank you for checking out my video on AMD's latest A8 Apu 00:10:32.399 --> 00:10:38.880 don't forget to subscribe to lus techtips for more unboxings reviews and other computer videos