WEBVTT 00:00:06.879 --> 00:00:14.120 for the ultimate no compromises compact gaming PC look no further than the 00:00:11.280 --> 00:00:18.119 obsidian series 250d from Corsair click now to learn more it's as though 00:00:16.000 --> 00:00:23.039 Gigabyte is determined to have all of their cards look pretty much the same 00:00:20.640 --> 00:00:30.240 it's yet another day and yet another windforce graphics card from Gigabyte so 00:00:26.640 --> 00:00:31.759 this one is the GTX 780 Ti wi for it's 00:00:30.240 --> 00:00:40.800 clocked at 1,20 MHz base clock 1,85 MHz boost clock 00:00:36.399 --> 00:00:45.800 it's got 3 gigs of 7 GHz gddr5 memory on 00:00:40.800 --> 00:00:48.920 a 384-bit bus and unlike the 290 X wind 00:00:45.800 --> 00:00:51.719 force it is pretty much the same price 00:00:48.920 --> 00:00:56.160 as other OC cards it's actually only about a $20 at or over a reference 00:00:54.000 --> 00:01:00.600 cooled reference clogged card making it a very interesting option it's got a 00:00:58.160 --> 00:01:05.119 massive six heat pipe triple fan cooler with their triangle cool uh slanted 00:01:02.719 --> 00:01:09.200 airflow redirection block with a Max cooling capacity of 450 WTS although 00:01:07.560 --> 00:01:14.000 guys it should be noted that not all windforce cards use this exact cooler 00:01:12.080 --> 00:01:18.880 this particular one is very very strong 450 Watts like I said it has a six pin 00:01:16.759 --> 00:01:22.920 and an8 pin PCI Express connector on the top and then that's pretty much it up 00:01:20.560 --> 00:01:27.640 there PCI Express 3.06 connector on the bottom at the back we find not a whole 00:01:24.920 --> 00:01:32.600 lot except we did notice that this card uses a reference PCB so in spite of the 00:01:30.119 --> 00:01:37.280 clock speed and cooler differences it has the same PCB as a reference GTX 00:01:34.880 --> 00:01:41.360 780ti so that's a 6 plus two-phase power design it supports the usual NVIDIA 00:01:39.200 --> 00:01:46.479 technology so that's SLI up to four-way with those two connectors on the top GPU 00:01:43.439 --> 00:01:47.960 Boost 2.0 Cuda physx GeForce experience 00:01:46.479 --> 00:01:54.119 so that's gameplay recording and streaming as well as the allnew gsync 00:01:51.079 --> 00:01:56.320 which finally has some sweet monitors to 00:01:54.119 --> 00:02:00.960 go along with it inside the Box the usual stuff's included a couple adapters 00:01:58.920 --> 00:02:06.119 anti-static beds driver disc Etc but what we're really interested in is the 00:02:02.960 --> 00:02:09.560 performance of this card with GPU Boost 00:02:06.119 --> 00:02:11.720 2.0 performance is directly affected by 00:02:09.560 --> 00:02:16.959 the GPU temperatures and that should in theory result in better clock speeds 00:02:14.200 --> 00:02:22.720 both at stock and overclocked settings at stock this was definitely the case 00:02:19.760 --> 00:02:27.239 although likely due as much to the much higher base clock of the card as 00:02:24.160 --> 00:02:28.959 anything else but we overclock all of 00:02:27.239 --> 00:02:34.200 our cards before testing and what we discovered is that that while the card 00:02:30.800 --> 00:02:36.800 ran relatively cool and quiet we did not 00:02:34.200 --> 00:02:41.440 get a Lucky Chip and the card actually didn't overclock as well as our 00:02:38.680 --> 00:02:47.040 reference 780 Ti bear in mind that our reference 780 Ti is a bit of a beast we 00:02:43.720 --> 00:02:48.560 were lucky on that one but in general 00:02:47.040 --> 00:02:53.480 let this be a lesson to you in the Silicon Lottery there is no sure thing 00:02:50.720 --> 00:02:58.040 we have found that reference cards usually lose and non-reference cards 00:02:56.680 --> 00:03:02.800 usually win when it comes to overclocking performance but here's the 00:03:00.680 --> 00:03:06.319 proof that there are exceptions I'll hand this over to Luke to deliver the 00:03:04.640 --> 00:03:11.400 bad news with this card it seems like we hit a luck of the draw case our 00:03:08.720 --> 00:03:14.400 reference card seems to be quite fast and with this card if I tried to make it 00:03:12.920 --> 00:03:19.480 as fast as our reference card we actually had some problems with artifacting and Screen glitches and some 00:03:17.799 --> 00:03:23.120 pretty serious issues it wasn't going to run that fast I was hitting power and 00:03:21.480 --> 00:03:27.480 voltage limits like crazy and I just couldn't go any higher they may or may 00:03:25.720 --> 00:03:31.879 not be some voltage mods out there that you could try if you run into this on 00:03:28.799 --> 00:03:33.439 your own but not going to go there um it 00:03:31.879 --> 00:03:36.599 it performed actually quite well just not as fast as our reference card and 00:03:35.120 --> 00:03:43.360 the main thing that I liked about it considering I couldn't make it go any faster was actually how quiet and how 00:03:40.879 --> 00:03:47.560 low temperature it was it was really not that loud at all it was really hard to 00:03:45.319 --> 00:03:52.120 hear which is a good thing and it usually sat around like 70° which is 00:03:50.439 --> 00:03:55.920 actually really solid for a graphics card so that was totally fine like and 00:03:54.480 --> 00:03:59.239 dislike the video as always in the comments below are you interested in 00:03:57.720 --> 00:04:02.239 something like this a reference design with an after market cooler on it or are 00:04:01.000 --> 00:04:06.879 you someone who's going more for aftermarket PCB and aftermarket cooler 00:04:05.400 --> 00:04:11.239 say that on the forums below and you'll get more interaction there which is great and as always subscribe to lus 00:04:09.799 --> 00:04:15.959 Tech tips and if you're on the Forum check out the PC gaming section where the star SE and group is we have now 00:04:14.400 --> 00:04:19.799 become official and we want to hit those upper echelons we're currently the 16th 00:04:17.880 --> 00:04:26.839 in the world for largest star seien team we want to get up there a little bit more so see you guys 00:04:28.720 --> 00:04:31.720 there 00:04:44.759 --> 00:04:47.759 he