WEBVTT 00:00:00.560 --> 00:00:08.160 every once in a while we come across something that is so bizarre that 00:00:05.680 --> 00:00:12.759 whether it's a good value or not we kind of just have to buy it and that's 00:00:10.240 --> 00:00:18.400 exactly what happened on taobao.com 00:00:14.320 --> 00:00:21.600 this my friends is a core i7 00:00:18.400 --> 00:00:25.600 4720 hq 00:00:21.600 --> 00:00:28.240 what a 4720 hq is a solder only 00:00:25.600 --> 00:00:31.359 processor for laptops i mean it's a few generations old but it actually 00:00:29.760 --> 00:00:37.520 performed pretty decent but this is what's weird about it it's a 4720hq 00:00:35.040 --> 00:00:42.960 and it looks like they have jerry rigged like a heat spreader and a 00:00:41.200 --> 00:00:48.320 substrate for it so that you could install it in a desktop motherboard 00:00:46.800 --> 00:00:54.079 so that translated to 123 canadian rubles is it 00:00:52.160 --> 00:00:58.719 a good deal well we got it 00:00:56.239 --> 00:01:03.280 so i guess we gonna find out today's video is brought to you by 00:01:00.239 --> 00:01:05.040 origin pc origin bc builds custom 00:01:03.280 --> 00:01:09.119 desktops and high performance laptops with free lifetime 24 7 tech support and 00:01:07.520 --> 00:01:14.960 they only use high-end components like samsung's 970 evo pro m.2 SSD visit 00:01:12.320 --> 00:01:25.200 origin pc today at the link below for a special promotion 00:01:25.200 --> 00:01:31.040 so step number one i guess is inspect the goods because when you order from 00:01:30.240 --> 00:01:34.720 like weird random chinese sites and like how 00:01:33.280 --> 00:01:38.880 long did this thing take to come just over a week dang that's not bad but when 00:01:37.040 --> 00:01:42.560 you order from these kinds of sites you kind of never know what kind of 00:01:40.640 --> 00:01:48.079 experience you're going to have so the first thing that stands out to me is 00:01:44.400 --> 00:01:51.119 that this heat spreader is clearly 00:01:48.079 --> 00:01:53.200 custom milled by whoever is 00:01:51.119 --> 00:01:56.399 finding these cpus wherever the heck they're finding them here i'm gonna flip 00:01:54.880 --> 00:02:03.280 them over and this is where stuff gets really crazy so this is a 4790k and this 00:01:59.759 --> 00:02:04.799 is a 4720hq so desktop chip mobile chip 00:02:03.280 --> 00:02:08.959 so you can actually that's great you can see the comparison between the thickness 00:02:06.560 --> 00:02:13.440 of the but this is nuts you can actually see the ball grid array soldering job 00:02:11.200 --> 00:02:16.480 between the original CPU and then this like 00:02:14.239 --> 00:02:21.840 aftermarket substrate here and then you can also see that not all 00:02:19.040 --> 00:02:25.840 the contact pads seem to be actually mapped to anything like some of them 00:02:23.280 --> 00:02:30.239 look blank like they don't have a trace running to them which is especially 00:02:27.520 --> 00:02:34.560 weird because the mobile versions of these fourth generation processors 00:02:32.239 --> 00:02:41.200 actually had more pins on the bottom of them than the desktop ones so how do you 00:02:37.120 --> 00:02:42.720 map more to less how do you find a board 00:02:41.200 --> 00:02:47.120 that it'll work in you guess 00:02:44.239 --> 00:02:53.599 now to be clear the manufacturer does list the compatible chipsets so anything 00:02:50.480 --> 00:02:55.040 z87 z97 and then the more mainstream 00:02:53.599 --> 00:02:58.959 versions of that couple of generations theoretically 00:02:57.519 --> 00:03:04.720 should work but CPU compatibility from board to 00:03:01.360 --> 00:03:08.159 board is not always as cut and dried as 00:03:04.720 --> 00:03:10.959 that so we just kind of grabbed the best 00:03:08.159 --> 00:03:17.360 bet z87 board we had lying around in oh boy we are running up against our first 00:03:13.599 --> 00:03:20.480 hiccup here so on a typical Intel lga 11 00:03:17.360 --> 00:03:22.560 5x CPU you've got these little tabs here 00:03:20.480 --> 00:03:26.239 these are for these to push the CPU down into the socket and make sure these pads 00:03:24.159 --> 00:03:29.840 are making good contact with the pins well it's too thick because you've 00:03:28.000 --> 00:03:33.040 effectively got like two substrates worth here so we gotta kind of my 00:03:31.760 --> 00:03:37.599 goodness it doesn't even come back up now we're gonna brick this thing for sure oh 00:03:35.519 --> 00:03:40.480 crap it's coming out of the socket well if it falls back in the socket that 00:03:38.959 --> 00:03:44.560 socket's probably dead so we're gonna go ahead and throw this little just washer 00:03:42.319 --> 00:03:47.760 in here use that as a spacer and then hopefully we'll be able to install it 00:03:45.920 --> 00:03:51.599 properly all right so that's it 00:03:48.959 --> 00:03:53.920 mobile CPU desktop motherboard you guys saw it here first 00:03:52.959 --> 00:03:59.519 that is unless you've seen this somewhere else in which case you saw it here second or third or whatever 00:03:57.040 --> 00:04:03.599 all right put my cooler back on i like their shinier heat spreader 00:04:01.840 --> 00:04:06.640 be very careful that you aren't mounting your CPU with too much pressure because 00:04:05.120 --> 00:04:10.720 i'm pretty sure that's what killed that pf sense project that i was working on 00:04:08.480 --> 00:04:15.280 like three times too much pressure is bad for the board so this is z87 which 00:04:13.760 --> 00:04:19.519 means we're going to be using ddr3 memory and we actually had really good 00:04:17.600 --> 00:04:22.960 luck with these particular g-skill rip jaws kits over the years 00:04:21.440 --> 00:04:26.880 of course we have the one board left from this generation that 00:04:24.720 --> 00:04:34.479 doesn't have stupid on-board power switches there we go hey 00:04:30.240 --> 00:04:37.040 it actually detects it core i7 4720 00:04:34.479 --> 00:04:44.320 i wasn't done with that oh there we go 4720 hq 16 gigs RAM our ssds working 00:04:41.280 --> 00:04:46.880 press f1 to run setup heck yeah wow that 00:04:44.320 --> 00:04:50.800 was painless 26 degree temperatures that looks fine voltage zero degrees that 00:04:49.759 --> 00:04:57.520 looks less funny okay so the thing about CPU 00:04:54.560 --> 00:05:01.680 compatibility is that sometimes you can get things to work that aren't 00:04:59.759 --> 00:05:05.120 officially supported in fact there are people who have gotten coffee like cpus 00:05:03.919 --> 00:05:12.240 to run in like z270 and z170 boards but 00:05:09.520 --> 00:05:16.320 you might run into some unpredictable behavior so 00:05:14.400 --> 00:05:19.919 let's go ahead and jump into advanced mode here and see what else we can dig 00:05:17.759 --> 00:05:24.720 up i'm having fun like why does the cache voltage read and the core voltage 00:05:22.639 --> 00:05:28.479 doesn't system agent voltage is fine too these are all CPU voltages of one like 00:05:26.960 --> 00:05:33.199 one or another here we go i mean everything is actually showing up 00:05:30.960 --> 00:05:36.240 like totally normally why do you not install chrome 00:05:34.479 --> 00:05:39.280 yeah no it just like is very inconvenient because then we have to use 00:05:37.680 --> 00:05:44.639 edge and then we have to just manually type wow this is very important news 00:05:41.840 --> 00:05:48.160 worst ever dress the worst ever 00:05:46.400 --> 00:05:52.960 so everything has been shockingly painless so far we fire up task manager 00:05:50.800 --> 00:06:00.000 all four cores all eight threads are all detected so wow both speed step power 00:05:56.240 --> 00:06:01.440 saving and turbo boost are working 00:06:00.000 --> 00:06:08.400 who was it like which chinese reseller was like 00:06:05.280 --> 00:06:10.960 you know what we have too many 47 20 00:06:08.400 --> 00:06:13.840 hq's whatever shall we do with these you know what let's spend a bunch of time 00:06:12.639 --> 00:06:18.880 getting them working in desktop motherboard do a lot of that 00:06:16.080 --> 00:06:23.680 so our next step because you would think the benefit of a mobile processor would 00:06:21.039 --> 00:06:30.800 be that it would be very power efficient and it would be very cool so we want to 00:06:26.720 --> 00:06:32.639 find out just how cool it is so our next 00:06:30.800 --> 00:06:36.000 step here is we're running a stress test so we're using Intel burn test 00:06:34.400 --> 00:06:41.120 and then we're using hardware monitor in order to monitor our CPU temperatures 00:06:38.000 --> 00:06:43.360 now a potential benefit of using a 00:06:41.120 --> 00:06:47.919 mobile processor would be that you'd have less heat output and less power 00:06:45.360 --> 00:06:52.160 consumption so theoretically compared to a 4770k which was the desktop equivalent 00:06:50.720 --> 00:06:55.919 at that time we should run much cooler so we're going 00:06:54.160 --> 00:07:00.400 to go ahead and we're going to run this guy for a little while and then 00:06:58.080 --> 00:07:04.960 see how it's looking but quite frankly i am feeling pretty good so far because 00:07:02.880 --> 00:07:08.720 we've had this running already for about a minute and our maximum oh it just 00:07:07.199 --> 00:07:13.599 jumped it was sitting at about 40 degrees still 00:07:11.120 --> 00:07:19.380 this is very respectable somewhere in the 60 to 65 degrees celsius range and 00:07:16.800 --> 00:07:21.199 that's right on the dies 00:07:21.199 --> 00:07:26.960 i told you it's all the way in i do know how to mount a sequence 00:07:24.479 --> 00:07:31.520 the first time that you're saying that so that's pretty impressive we're stable 00:07:29.199 --> 00:07:35.520 which was not a given 00:07:33.440 --> 00:07:41.039 and we're running about 20 degrees cooler than a 4770k but there's a clear 00:07:38.800 --> 00:07:46.000 trade-off here so our maximum turbo speed is only about 3.4 gigahertz 00:07:43.440 --> 00:07:50.400 compared to 3.9 on a 4770k plus that would be an 00:07:48.240 --> 00:07:55.120 overclockable chip so let's see what that translates to in 00:07:52.880 --> 00:07:59.840 terms of performance then so we decided to pair our CPU with a gtx 1060 because 00:07:58.000 --> 00:08:04.479 i mean the reality of it is if you're trying to spend a hundred dollars on a 00:08:01.520 --> 00:08:07.759 CPU off taobao.com you weren't buying an rtx 2080 ti that wouldn't be a good 00:08:06.479 --> 00:08:13.680 match anyway so we still think 1920 by 1080 we should 00:08:11.440 --> 00:08:18.319 be able to get darn near very high or even ultra details even in a aaa game 00:08:16.400 --> 00:08:22.560 like shadow of the tomb raider so let's just see that is as long as our CPU can 00:08:20.240 --> 00:08:26.960 keep up this is really impressive i mean other 00:08:24.560 --> 00:08:31.120 than some wicked tearing when we pan the camera around which is i mean we're not 00:08:28.720 --> 00:08:36.719 using g-sync or freezing so that kind of comes with the territory 00:08:33.519 --> 00:08:38.959 this is running really smooth so it did 00:08:36.719 --> 00:08:45.600 a great job with shadow of the tomb raider it actually ended up performing 00:08:41.519 --> 00:08:48.240 nearly identically to our 4770k which 00:08:45.600 --> 00:08:52.800 means we could even go higher than a gtx 1060 graphics card and still be getting 00:08:51.040 --> 00:08:56.640 the full experience bearing in mind of course that we had 00:08:54.399 --> 00:09:01.200 our details cranked meaning that we are less likely to be CPU bound let's move 00:08:59.120 --> 00:09:05.760 on to another title you're using a hard drive 00:09:03.680 --> 00:09:10.320 and impressively our gaming performance story doesn't 00:09:08.399 --> 00:09:15.760 really change with assassin's creed origins or gta 5 both of them were 00:09:13.360 --> 00:09:19.519 within 5 to 10 percent of the performance which 00:09:17.839 --> 00:09:24.800 as long as we're getting a pretty good price difference means that this could 00:09:22.000 --> 00:09:29.920 make a ton of sense for someone who wants to game on the cheap bringing us 00:09:27.120 --> 00:09:35.200 then to our last gaming test the one that separates the men from the boys 00:09:31.920 --> 00:09:37.839 when it comes to CPU gaming performance 00:09:35.200 --> 00:09:37.839 csgo 00:09:49.360 --> 00:09:53.920 oh crap so clearly it's usable and more 00:09:52.399 --> 00:10:00.800 than that it was actually within again 00:09:57.040 --> 00:10:02.399 10 20 frames per second of our 4770k 00:10:00.800 --> 00:10:06.640 which i don't know what is that like three percent or something stupid like that like 00:10:04.640 --> 00:10:10.320 not even much of a difference but i guess that's kind of to be expected 00:10:08.399 --> 00:10:14.240 because our low core count turbo is a little bit 00:10:12.480 --> 00:10:18.640 higher than our all-core turbo so we're running cinebench now and you can see 00:10:15.760 --> 00:10:23.360 we're maxing out at 3.4 gigahertz so i'm going to be curious to see how 00:10:20.399 --> 00:10:28.640 our score here stacks up to a 4770k because even gamers occasionally do 00:10:25.839 --> 00:10:31.839 something else on their computers 671. 00:10:30.240 --> 00:10:36.480 okay that's not impressive what's the 4770k 00:10:33.959 --> 00:10:40.800 746. so yeah that's definitely going to be a 00:10:38.560 --> 00:10:44.880 noticeable difference if you're actually doing any 3d rendering or if you're 00:10:42.720 --> 00:10:50.480 editing video or something like that but probably not the end of the world i 00:10:47.200 --> 00:10:51.279 guess the real question here becomes one 00:10:50.480 --> 00:10:56.560 of value so here's the deal 00:10:54.240 --> 00:11:01.160 we ran all our numbers through again this time with an 00:10:58.120 --> 00:11:05.839 8700k so remember this is more like a 00:11:01.160 --> 00:11:07.760 370 us dollar CPU and obviously there is 00:11:05.839 --> 00:11:12.480 a performance improvement but if you're looking at it from a pure bang 00:11:10.800 --> 00:11:17.920 for the buck perspective these 4th gen cpus hold up incredibly 00:11:15.760 --> 00:11:24.480 well in modern games so the question really becomes 00:11:19.440 --> 00:11:26.640 do you grab a used 4770 or 4790k on ebay 00:11:24.480 --> 00:11:30.320 like a legitimate processor that was actually designed to go on the 00:11:27.920 --> 00:11:34.880 motherboard or do you go the taobao route now one of these is going to cost 00:11:32.959 --> 00:11:40.160 you somewhere in the neighborhood of around 200 us dollars so you factor in 00:11:37.920 --> 00:11:46.000 another 50 bucks for your motherboard another 80 bucks assuming that you want 00:11:42.320 --> 00:11:48.240 16 gigs of ddr3 memory of some sort and 00:11:46.000 --> 00:11:53.760 you're looking at what's what's that add up to 50 plus 200 plus 80 stuff 330 00:11:52.240 --> 00:11:56.480 which means if you could save a considerable amount here you'd be 00:11:55.600 --> 00:12:01.600 sitting pretty peachy keen and 00:11:59.120 --> 00:12:07.360 on the surface of things anyway that is exactly what our taobao CPU 00:12:05.279 --> 00:12:13.519 manages to achieve because based on the pricing on the website you 00:12:10.320 --> 00:12:16.320 could save almost half the price of the 00:12:13.519 --> 00:12:20.880 CPU i mean of course you still need the motherboard and you still need the RAM 00:12:17.760 --> 00:12:23.360 and all that but saving even 80 00:12:20.880 --> 00:12:27.200 on a budget gaming rig is a lot more money to put into i mean you could you 00:12:25.519 --> 00:12:31.600 can do a lot of things with 80 bucks you could get a better graphics card uh you 00:12:29.440 --> 00:12:35.440 could get a solid state drive like it makes a huge difference you could get a 00:12:33.279 --> 00:12:39.600 ball and cooler for example the issue for us 00:12:36.639 --> 00:12:45.680 is that it just didn't end up actually working out that way so after we placed 00:12:42.480 --> 00:12:48.480 our order the taobao seller contacted us 00:12:45.680 --> 00:12:53.519 and wanted another what was it 35 or 45. 45 u.s dollars 00:12:50.800 --> 00:12:57.600 for taxes and customs clearance into canada a situation that you might run 00:12:55.600 --> 00:13:02.399 into so now we're sitting up within about you know 30 00:13:00.000 --> 00:13:06.480 then when the package arrived we ended up having to pay another 15 bucks at the 00:13:05.360 --> 00:13:12.800 door so the issue here is that by the time we 00:13:10.320 --> 00:13:16.560 actually get this thing to our doorstep it's kind of ghetto you can't use a 00:13:14.959 --> 00:13:21.200 stock cooler meaning you got to factor that into your price 00:13:18.320 --> 00:13:24.639 and you've paid almost as much for it as a normal CPU 00:13:23.440 --> 00:13:28.720 meaning that it's really cool and i'm glad we bought 00:13:26.800 --> 00:13:33.040 one for the tech museum here at Linus media group 00:13:30.320 --> 00:13:37.519 but i probably wouldn't recommend it t-forces excalibur special edition ddr4 00:13:35.680 --> 00:13:44.720 RAM features a unique totem pattern design ultra wide 120 degree panel lighting 00:13:41.120 --> 00:13:46.720 speeds of 3600 to 4 000 megahertz and 00:13:44.720 --> 00:13:51.680 support for a variety of RGB software like rsync RGB fusion and more it's 00:13:49.600 --> 00:13:55.360 compatible with xmp 2.0 one step overclocking and they've worked with 00:13:53.600 --> 00:13:59.360 motherboard manufacturers to get their sticks qvl approved check out the link 00:13:57.519 --> 00:14:03.040 below to learn more so thanks for watching guys if you 00:14:00.639 --> 00:14:07.120 disliked this video then you can hit that button but if you liked it because 00:14:04.880 --> 00:14:09.839 it's just cool then you can hit the like button get subscribed wait did i tell 00:14:08.880 --> 00:14:14.399 them to hit the like button if they disliked it whatever everyone hit the like button you can get 00:14:13.040 --> 00:14:19.199 subscribed you can check out where to buy the stuff we featured at the link in the video description or while you're 00:14:17.360 --> 00:14:22.480 down there you can check out our merch store which has cool shirts like this 00:14:20.399 --> 00:14:27.600 one and our community forum which is definitely worth a join 00:14:25.440 --> 00:14:30.560 that this shirt's not but they're like this 00:14:28.560 --> 00:14:34.000 in the sense that they are Linus tech tipsy 00:14:32.320 --> 00:14:38.320 no like really though like it's got this shitty mounting hardware and stuff and 00:14:36.240 --> 00:14:40.160 it's like in the way can we can we throw it away no no you just you no no no no 00:14:40.160 --> 00:14:47.240 no no no no no no no no no no no no no no no no no no no no no no no no no no no no you only use this okay okay can we throw it away anyway though why because 00:14:44.079 --> 00:14:47.240 it's stupid