1 00:00:00,640 --> 00:00:11,400 external graphics cards dead or dormant 2 00:00:06,919 --> 00:00:15,120 stillborn or still relevant well we've 3 00:00:11,400 --> 00:00:18,880 got two identically speced laptops and 4 00:00:15,120 --> 00:00:21,880 both of the top external Graphics docks 5 00:00:18,880 --> 00:00:24,800 that are ready to battle it out each 6 00:00:21,880 --> 00:00:32,719 armed with a top-of-the-line NVIDIA GTX 1080 TI but will The Identical specs 7 00:00:29,119 --> 00:00:38,410 yield identical results for 8 00:00:32,719 --> 00:00:43,350 once no no they 9 00:00:46,120 --> 00:00:53,559 won't the be quiet! purebase 600 features 10 00:00:49,719 --> 00:00:56,079 a versatile design noise insulation and 11 00:00:53,559 --> 00:01:01,559 optimization for different liquid cooled radiator configurations learn more at 12 00:00:58,879 --> 00:01:07,960 the link below first we need to explain why we're using 13 00:01:04,199 --> 00:01:11,799 such op graphics cards the goal here was 14 00:01:07,960 --> 00:01:14,680 to expose any weaknesses or efficiency 15 00:01:11,799 --> 00:01:20,960 differences between these two by eliminating the graphics card itself as 16 00:01:17,159 --> 00:01:23,439 a bottleneck and also to make sure that 17 00:01:20,960 --> 00:01:30,560 both of them can keep even the power hungriest cards safely fed each are 18 00:01:27,600 --> 00:01:35,840 equipped with plenty of power power the Razer core sports a custom 500 W power 19 00:01:33,560 --> 00:01:42,200 supply while the Alienware Graphics amplifier has a hunt key 460 W ATX power 20 00:01:39,600 --> 00:01:48,399 supply that was clearly made for a full system and repurposed for this the cor's 21 00:01:45,320 --> 00:01:50,240 aluminum chassis feels solid and opening 22 00:01:48,399 --> 00:01:55,759 it up to insert the graphics card only takes a couple of minutes whereas the 23 00:01:52,799 --> 00:02:02,200 graphics amplifier well everything about that setup frankly felt a little janky 24 00:01:58,560 --> 00:02:04,520 the build is all plastic and although I 25 00:02:02,200 --> 00:02:08,640 had the right idea for opening it I actually ended up having to consult the 26 00:02:06,320 --> 00:02:13,640 user manual because it felt like I was going to break it oh God when it got 27 00:02:11,360 --> 00:02:19,040 stuck halfway I mean there isn't even RGB Lighting on this one for crying out 28 00:02:16,000 --> 00:02:21,760 loud and if you thought the poor ease of 29 00:02:19,040 --> 00:02:29,440 use for the amplifier was over think again the connectors on the proprietary 30 00:02:25,160 --> 00:02:32,599 cable that carries a PCI Express 3.0 X4 31 00:02:29,440 --> 00:02:35,239 signal from the computer to the dock 32 00:02:32,599 --> 00:02:39,280 have these latches that can actually get caught if you're not careful when you're 33 00:02:36,840 --> 00:02:45,159 removing them not to mention that Alienware solution requires a restart of 34 00:02:42,280 --> 00:02:49,800 the laptop whenever you connect or disconnect it so that the internal 35 00:02:47,560 --> 00:02:55,760 dedicated graphics card can be activated or disabled razor on the other hand 36 00:02:52,840 --> 00:03:01,560 keeps up their strong performance in this area just plug in a Thunderbolt 3 37 00:02:59,360 --> 00:03:06,159 cable into the blade notebook of your choice they don't officially support 38 00:03:03,720 --> 00:03:11,280 thirdparty laptops but we'll discuss that more later install the NVIDIA 39 00:03:08,159 --> 00:03:15,080 drivers and you're good to go every part 40 00:03:11,280 --> 00:03:19,760 of the process felt smoother and more 41 00:03:15,080 --> 00:03:22,560 polished on the Razor side though if its 42 00:03:19,760 --> 00:03:28,560 performance can hold up I suspect many will forgive the Alienware Graphics 43 00:03:25,159 --> 00:03:32,680 amplifier for its lack of uh spit and 44 00:03:28,560 --> 00:03:34,680 shine given that retails for 200 bucks a 45 00:03:32,680 --> 00:03:40,840 whopping $300 less than the core though in 46 00:03:38,720 --> 00:03:47,120 razer's defense with the purchase of a blade laptop you do get $100 off so I 47 00:03:44,239 --> 00:03:51,120 guess it's only twice the cost of the graphics 48 00:03:48,080 --> 00:03:54,720 amplifier let's get to the test every 49 00:03:51,120 --> 00:03:57,599 game Benchmark was Run 10 times in a row 50 00:03:54,720 --> 00:04:02,519 to reward strong thermal performance and we recorded both the average and and the 51 00:04:00,239 --> 00:04:09,079 minimum frame rate that was maintained 97% of the time we used identical 52 00:04:05,959 --> 00:04:12,120 drivers and settings and ausa's 240 HZ 53 00:04:09,079 --> 00:04:15,040 gaming display for both systems 54 00:04:12,120 --> 00:04:21,000 everything between them was equal except for one thing gaming 55 00:04:18,479 --> 00:04:27,240 performance for Rise of the Tomb Raider and for honor the Alienware pulled out a 56 00:04:24,080 --> 00:04:29,000 bit ahead of the core while the core 57 00:04:27,240 --> 00:04:35,039 actually pulled slightly ahead in the all important minimum FPS metric but 58 00:04:32,720 --> 00:04:40,479 where things really got interesting was in Ghost Recon wildlands and DX mankind 59 00:04:38,360 --> 00:04:46,800 divided where the Alienware Graphics amplifier showed that it has the 60 00:04:42,479 --> 00:04:49,320 potential to absolutely spank the core 61 00:04:46,800 --> 00:04:55,639 delivering buttery visuals compared to the questionably playable performance on 62 00:04:52,560 --> 00:04:58,320 Razer solution and as a troubleshooting 63 00:04:55,639 --> 00:05:04,120 step then we tried comparing them using the built-in laptop display rather than 64 00:05:01,000 --> 00:05:07,039 an external monitor but this changed 65 00:05:04,120 --> 00:05:12,720 nothing for Alienware and then generated even slightly lower performance for 66 00:05:09,520 --> 00:05:15,320 Razer due to the extra display signal 67 00:05:12,720 --> 00:05:19,560 data being carried back to the blades display something that we actually 68 00:05:17,000 --> 00:05:26,319 tested more thoroughly in the Razer core Deep dive which you can check out here 69 00:05:22,520 --> 00:05:28,720 so what gives well we could only think 70 00:05:26,319 --> 00:05:34,639 of two differences between these solutions that might affect performance 71 00:05:31,440 --> 00:05:36,319 one thermals either of the CPU in the 72 00:05:34,639 --> 00:05:43,199 laptop or the graphics card in the enclosure and two the connection to the 73 00:05:40,639 --> 00:05:48,600 graphics card so to see if thermal throttling was the issue we booted up 74 00:05:45,520 --> 00:05:52,240 Ida 64 and monitored temperatures on our 75 00:05:48,600 --> 00:05:53,840 blade while gaming it got hot but it was 76 00:05:52,240 --> 00:06:00,400 nowhere near overheating then to rule out the laptop 77 00:05:57,280 --> 00:06:02,960 entirely we gave our 7600 100 HQ 78 00:06:00,400 --> 00:06:09,360 equipped Blade Pro a shot but it performed nearly identically within our 79 00:06:05,479 --> 00:06:11,680 margin of error no dice let's move on to 80 00:06:09,360 --> 00:06:18,160 that connection then the theoretical bandwidth of the two should be equal in 81 00:06:15,000 --> 00:06:21,080 GPU Z both have the same number of PCIe 82 00:06:18,160 --> 00:06:24,639 lanes and the same reported 40 gbit pers second 83 00:06:21,960 --> 00:06:30,440 bandwidth is there some way that we could pin this on Thunderbolt 84 00:06:27,319 --> 00:06:33,240 3 what if the gigabit ether controller 85 00:06:30,440 --> 00:06:39,000 on the core is sucking up resources we used disabling device 86 00:06:35,520 --> 00:06:41,479 manager and it was not very effective so 87 00:06:39,000 --> 00:06:46,080 it was time to go spelunking deep into the depths of the specs searching for a 88 00:06:44,160 --> 00:06:51,240 single difference between the two and finally we found one in the dark 89 00:06:48,560 --> 00:06:57,440 crevices of the device manager on the Alienware amplifier the GPU gets 90 00:06:54,160 --> 00:07:01,360 connected directly to the PCI controller 91 00:06:57,440 --> 00:07:04,840 on the CPU so the lanes that it gets are 92 00:07:01,360 --> 00:07:07,240 solely devoted to Graphics the razor 93 00:07:04,840 --> 00:07:12,879 core on the other hand gets connected through the motherboard chipset meaning 94 00:07:09,960 --> 00:07:17,319 that the graphics card has to share its PCI El lanes and perhaps more 95 00:07:14,680 --> 00:07:23,879 importantly the DMI interface with things like networking USB ports and 96 00:07:20,720 --> 00:07:26,479 probably the biggest bandwidth sucker 97 00:07:23,879 --> 00:07:32,039 storage but can we prove once and for all that the lanes are causing the 98 00:07:29,039 --> 00:07:34,440 bottleneck NE for a complete Apples to 99 00:07:32,039 --> 00:07:40,879 Apples comparison we needed to be able to connect the core to the Alienware 13 100 00:07:38,080 --> 00:07:45,800 something Dell doesn't want you to do in spite of the 13th's higher-end four-lane 101 00:07:43,840 --> 00:07:50,319 implementation of the Thunderbolt 3 standard good news though it turns out 102 00:07:48,599 --> 00:07:56,120 that if you disable enough Alienware Services it will begrudgingly let you 103 00:07:52,840 --> 00:07:58,680 use a Razer core and in games the 104 00:07:56,120 --> 00:08:03,680 Alienware 13 connected to the core recorded results within our margin of 105 00:08:00,840 --> 00:08:07,039 error proving then that the bottleneck wasn't in the laptop but in the 106 00:08:05,759 --> 00:08:13,479 connection itself so final pros and cons list then 107 00:08:10,840 --> 00:08:19,039 Razer has some big advantages they simple setup and hot swapability will 108 00:08:16,879 --> 00:08:23,080 definitely win over customers who value ease of use and the excellent 109 00:08:21,199 --> 00:08:27,960 compatibility enabled by using a standard Thunderbolt 3 Connection in 110 00:08:26,000 --> 00:08:32,560 spite of advertising the core as only working with their own blade and blade 111 00:08:29,800 --> 00:08:37,560 stealth makes them the default winner for many but if you're willing to look 112 00:08:35,599 --> 00:08:42,240 the other way on build quality a little bit you're willing to Shell out for a 113 00:08:40,039 --> 00:08:47,240 matching Alienware laptop and you want the best no compromises performance out 114 00:08:44,800 --> 00:08:55,480 of your graphics stock at a much lower price the amplifier comes out way 115 00:08:52,080 --> 00:08:58,440 ahead Fiverr is totally unique it's this 116 00:08:55,480 --> 00:09:03,560 huge digital marketplace where Freelancers startups and small 117 00:09:00,440 --> 00:09:06,760 businesses connect and as Fiverr says 118 00:09:03,560 --> 00:09:08,720 get stuff done whether you're in a 119 00:09:06,760 --> 00:09:13,079 startup or a small company and you need a logo or business plan or maybe you 120 00:09:10,760 --> 00:09:18,120 need an entire website developed for you Fiverr has the talent pool that can get 121 00:09:16,120 --> 00:09:23,240 it done at a price that fits your budget so check out some of the gigs that we' 122 00:09:21,040 --> 00:09:30,279 use through our landing page at the link below and the first 300 viewers who use 123 00:09:27,240 --> 00:09:33,240 code Linus will get 20% off Fiverr 124 00:09:30,279 --> 00:09:37,480 services so join the hash inders We Trust movement on Fiverr and start 125 00:09:35,519 --> 00:09:41,200 getting it done today so thanks for watching guys if you 126 00:09:39,800 --> 00:09:44,920 dislike this video you can hit that button but if you liked it hit like get 127 00:09:43,480 --> 00:09:50,040 subscribed maybe consider checking out where to buy the stuff we featured at the link in the video description also 128 00:09:48,600 --> 00:09:53,959 down there is our merch store where you can buy cool shirts like this one as 129 00:09:51,800 --> 00:09:58,680 well as a link to our community Forum where you can discuss all kinds of Great 130 00:09:56,320 --> 00:10:04,160 Tech things like could you run both of them in SL on the Alienware 13 to which 131 00:10:01,839 --> 00:10:07,160 the answer is unfortunately no we did try