1 00:00:01,040 --> 00:00:08,880 I'll take notes. Uh, we left off at implementing an open and transparent 2 00:00:05,680 --> 00:00:09,679 policy for the hom what this 3 00:00:08,880 --> 00:00:15,280 what is it? I need it for work and freeze. Like me, 4 00:00:13,280 --> 00:00:19,680 you've probably looked at the numerous desktop replacements on the market and 5 00:00:17,039 --> 00:00:24,560 thought, if these are so popular, why don't I know anyone who has one? It's a 6 00:00:22,160 --> 00:00:29,840 valid question. Who are the people driving ASUS and MSI to create these 7 00:00:27,119 --> 00:00:34,719 monstrosities? And more importantly, why? It seems like such a raw deal. You 8 00:00:33,040 --> 00:00:39,760 get less performance and less upgradability than a desktop and at a 9 00:00:37,120 --> 00:00:44,640 higher price to boot. Though, on the flip side, with the latest chips from 10 00:00:42,000 --> 00:00:52,000 Intel, the gap has actually narrowed a lot. In front of me, well, not me, that 11 00:00:47,920 --> 00:00:56,000 guy, is a 24 core CPU that turbos to 5.6 12 00:00:52,000 --> 00:00:58,000 to 6 GHz, a 4K 144Hz HDR display, and a 13 00:00:56,000 --> 00:01:01,920 fully mechanical keyboard. I mean, with a system like that, you can do a whole 14 00:00:59,840 --> 00:01:06,479 lot more than just gaming. And we're going to dig into the creative and 15 00:01:04,000 --> 00:01:10,880 professional use cases that are driving people like that to destroy their spines 16 00:01:08,799 --> 00:01:14,000 hauling these things around. And of course, we're going to see how it 17 00:01:12,080 --> 00:01:18,080 handles games, too. because a lot of what we need to talk about has to do 18 00:01:15,600 --> 00:01:23,360 with the RTX 4090 mobile and the misleading marketing that NVIDIA is 19 00:01:19,840 --> 00:01:24,880 doing around this fast but misunderstood 20 00:01:23,360 --> 00:01:29,360 gaming GPU. There have always been laptops that aspired to desktop tier 21 00:01:27,520 --> 00:01:33,119 performance. Some even went as far as to use desktop chips. But it's only 22 00:01:31,520 --> 00:01:38,159 recently that the performance gap between them has become so small that 23 00:01:35,680 --> 00:01:43,520 Apple, for example, no longer even designs different CPUs for use in its 24 00:01:40,560 --> 00:01:48,159 mobile and desktop lineups. So, with that fire burning right under their 25 00:01:45,280 --> 00:01:53,520 butts, Intel has responded with this, the Core i9 13980HX. 26 00:01:51,439 --> 00:01:57,920 And while it might not be the most power efficient, chugging back up to 10 times 27 00:01:55,920 --> 00:02:01,360 the wattage of their lowest end UER chips, if the on paper specs are 28 00:01:59,840 --> 00:02:07,600 anything to go by, it should be considerably faster than your gaming 29 00:02:03,600 --> 00:02:09,360 desktop is. Yes, even yours, Steven. The 30 00:02:07,600 --> 00:02:14,720 result then is a device that is both portable, well technically portable, and 31 00:02:12,239 --> 00:02:19,520 powerful enough to get anything done anywhere, even if what you want to get 32 00:02:16,879 --> 00:02:25,920 done is gaming at max settings, which I do. For our testing, we used the latest 33 00:02:22,640 --> 00:02:27,840 MSI Titan GT77, mainly because we 34 00:02:25,920 --> 00:02:31,920 happened to have it also available with previous generation hardware for 35 00:02:29,280 --> 00:02:35,599 comparison, but any similarly configured machine with adequate cooling and the 36 00:02:33,680 --> 00:02:39,360 same power envelope should perform about like this. We're going to have both 37 00:02:37,360 --> 00:02:42,879 models and more linked down below if you want to play around with pricing of the 38 00:02:40,640 --> 00:02:47,680 various options. Getting straight into our 4K gaming results, it's clear that a 39 00:02:45,920 --> 00:02:52,160 lot of the generational performance improvements here are thanks to NVIDIA 40 00:02:50,000 --> 00:02:56,879 rather than Intel, who managed a very respectable bump in GPUbound titles 41 00:02:54,480 --> 00:03:01,519 compared to our older 3080 Ti equipped GT77. 42 00:02:58,560 --> 00:03:05,840 MSI is feeding the 4090, the maximum permitted 175 watts when you include 43 00:03:03,920 --> 00:03:13,280 dynamic boost, netting us a gain of between 30 to 35% in Hitman 3, F122, and 44 00:03:10,560 --> 00:03:18,239 Cyberpunk 2077, with Red Dead Redemption 2 seeing a 49% increase and Assassin's 45 00:03:16,159 --> 00:03:24,239 Creed Valhalla jumping by a whopping 60%. Temperatures stayed reasonable 46 00:03:20,879 --> 00:03:26,560 while gaming, too. 73 degrees is around 47 00:03:24,239 --> 00:03:32,720 five more than the 12th gen laptop, but our core clocks were also 400 MHz higher 48 00:03:29,440 --> 00:03:36,640 on average. So why then is everybody 49 00:03:32,720 --> 00:03:38,319 complaining about this thing? Okay, well 50 00:03:36,640 --> 00:03:44,480 look at what happens when we add the desktop 4090 to the charts. It is 51 00:03:41,040 --> 00:03:48,080 between 30 and 50% slower, making the 52 00:03:44,480 --> 00:03:49,840 naming well misleading at best. I mean, 53 00:03:48,080 --> 00:03:54,319 this gap in performance makes it look like we're testing two completely 54 00:03:51,680 --> 00:03:59,360 different tiers of products. And the pricing is like piling insult on top of 55 00:03:56,720 --> 00:04:04,159 injury, even when we factor in a comparable monitor and peripherals, 56 00:04:01,200 --> 00:04:09,200 we're spending an extra $300 for this degraded laptop experience. But with 57 00:04:07,200 --> 00:04:13,760 that said, if we come back to our original question, some gamers have very 58 00:04:12,080 --> 00:04:18,239 good reasons for accepting these compromises. We heard loud and clear 59 00:04:16,160 --> 00:04:22,960 that for folks who spend more time away from home than in it, say living in a 60 00:04:20,799 --> 00:04:26,400 barracks or a dorm room, working as a seaf farer or a long haul driver, or 61 00:04:25,040 --> 00:04:32,720 even if you're just a game-loving executive, a desktop might technically 62 00:04:29,440 --> 00:04:34,880 be a better PC, but being able to 63 00:04:32,720 --> 00:04:39,759 actually use your laptop every day makes it a far more valuable investment. I 64 00:04:37,600 --> 00:04:44,960 also have to wonder, too, is this difference even NVIDIA's fault? Maybe 65 00:04:42,160 --> 00:04:51,280 the 13980HX isn't all it's cracked up to be, and Team Green's poor GPU is simply 66 00:04:48,000 --> 00:04:53,520 being hamstrung by a crap mobile CPU. 67 00:04:51,280 --> 00:04:57,759 Unfortunately, tightly integrated mobile devices make it a lot harder for us to 68 00:04:55,759 --> 00:05:03,040 isolate variables. I mean, we can't very well plug a desktop 4090 into a laptop 69 00:05:00,160 --> 00:05:08,720 and then see how the CPU runs. But what we can do is try a couple of more CPU 70 00:05:06,000 --> 00:05:16,720 limited scenarios and CPU specific gaming benchmarks where 71 00:05:11,280 --> 00:05:19,440 no excuses NVIDIA that CPU is fast. Not 72 00:05:16,720 --> 00:05:24,160 as fast as her desktop cousin, but close enough that I don't feel completely 73 00:05:21,440 --> 00:05:28,560 misled by the similar naming scheme. And close enough that it opens up new use 74 00:05:26,160 --> 00:05:33,199 cases for these chunky boys that might not have made sense just a few short 75 00:05:30,479 --> 00:05:37,039 years ago. Let me set the stage. And this is actually based on a true story, 76 00:05:34,720 --> 00:05:42,080 which is cool. You are in the deepest, darkest bowels of a ship. Your job is to 77 00:05:40,000 --> 00:05:46,639 produce a 3D scan of the crew's workspace. The iPhone 12 Pro hasn't been 78 00:05:44,560 --> 00:05:51,440 invented yet, so the contractor in charge of running the 3D scanner pulls 79 00:05:48,160 --> 00:05:53,840 out one of these. I mean, it looks out 80 00:05:51,440 --> 00:05:59,039 of place, right? But I assure you that for plenty of industry professionals, 81 00:05:56,000 --> 00:06:01,280 gaming laptops like these ones are their 82 00:05:59,039 --> 00:06:06,080 daily driver work machines. And if you stop and think about it, it makes a ton 83 00:06:03,280 --> 00:06:12,080 of sense. While they could be served by boring black workstation laptops, the 84 00:06:09,280 --> 00:06:18,960 certified hardware tax, man, that makes the mobility tax look like a like a a 85 00:06:15,840 --> 00:06:21,199 small convenient tax, one that you don't 86 00:06:18,960 --> 00:06:26,800 mind paying. I mean, check out these portable workstations from Lenovo and 87 00:06:22,800 --> 00:06:29,199 Dell. These are last gen hardware. So 88 00:06:26,800 --> 00:06:34,639 smart IT and purchasing managers figured out a long time ago that a clever way to 89 00:06:31,759 --> 00:06:40,800 dodge this premium for what is on a silicon level the same bloody thing is 90 00:06:37,199 --> 00:06:42,800 to buy consumer hardware instead. And I 91 00:06:40,800 --> 00:06:47,919 mean yeah I guess you could game on it after work or before work or during work 92 00:06:45,840 --> 00:06:52,319 assuming you're stealthy about it like our new Stealth deskpads. A simple all 93 00:06:50,720 --> 00:06:57,039 black look and the same excellent microtexture cloth means uh you need to 94 00:06:54,800 --> 00:07:02,400 go to ltstore.com. Whatever the point is, looking at some of our CPU specific 95 00:06:59,599 --> 00:07:07,599 benchmarks, there is a lot for those IT managers to love here. I mean, yeah, you 96 00:07:05,120 --> 00:07:11,120 could do it faster on a desktop, but you could also do it from a hotel lobby in 97 00:07:09,360 --> 00:07:15,039 the Bahamas on this thing. In our Chromium compile and sevenzip 98 00:07:13,039 --> 00:07:20,479 compression and decompression tests, we saw a clear advantage to going 13th gen 99 00:07:18,080 --> 00:07:26,240 mobile with it performing closer to our desktop CPU than to a lastg laptop one. 100 00:07:24,000 --> 00:07:31,440 And while Yuncher and Handbreak AV1 encoding showed a greater difference to 101 00:07:28,560 --> 00:07:36,319 desktop, they do still show a clear generational improvement over 12th gen. 102 00:07:34,319 --> 00:07:40,080 Now, we suspect that faster mobile memory would help close this gap a 103 00:07:38,080 --> 00:07:44,560 little bit. But another factor is also that multi-core workloads tend to favor 104 00:07:42,560 --> 00:07:49,280 desktop chips that have much higher power budgets. Speaking of multi-core 105 00:07:46,880 --> 00:07:53,280 loads, that brings us nicely to the last big group of people who gravitate 106 00:07:50,960 --> 00:07:57,360 towards these machines. Creators, makers, and artists. These are the kinds 107 00:07:55,840 --> 00:08:01,599 of people who need to be able to pack their entire studio into a single 108 00:07:59,120 --> 00:08:04,639 messenger bag. And thanks to the rich I/IO that you'll typically find on a 109 00:08:03,120 --> 00:08:08,639 desktop replacement, including Thunderbolt, that's as simple as 110 00:08:06,639 --> 00:08:12,800 bundling up a dock, a mouse, and a portable monitor, rolling up to a cafe, 111 00:08:10,960 --> 00:08:16,720 and enjoying the exact same workflow that you would at the office. This is 112 00:08:14,960 --> 00:08:20,879 also a group that tends to benefit from the creature comforts that make their 113 00:08:18,319 --> 00:08:24,639 way into premium devices like these. Our MSI unit, for example, has a fully 114 00:08:22,960 --> 00:08:30,960 mechanical keyboard with lowprofile CherryMX switches with numpad, a factory 115 00:08:27,840 --> 00:08:32,959 calibrated VSA certified HDR display, 116 00:08:30,960 --> 00:08:36,479 and high-speed 2 and 1/2 gig wired networking for running projects off of a 117 00:08:34,560 --> 00:08:41,360 network share. Once again, for this group, Intel's 13th gen HX series 118 00:08:38,560 --> 00:08:46,160 flagship shines, managing a flak audio encode only a single second slower than 119 00:08:43,599 --> 00:08:50,959 its desktop counterpart and 7 seconds faster than its older, outdated brother. 120 00:08:49,200 --> 00:08:56,320 And the young pup making the old dog look bad continues in Cinebench R23 121 00:08:53,760 --> 00:09:01,600 where multi-core scores are up 10,000 over our 12th gen equipped machine and 122 00:08:58,399 --> 00:09:03,920 only behind the desktop by 3200. 123 00:09:01,600 --> 00:09:07,600 While Blender GPU testing reveals the same embarrassing difference between 124 00:09:05,680 --> 00:09:12,160 mobile and desktop 4090s that we observed when gaming. Blender CPU has 125 00:09:10,240 --> 00:09:16,240 our mobile chip landing right in between the desktop and our old laptop with 126 00:09:14,399 --> 00:09:20,240 essentially the same story in Proion, Premiere, and Photoshop. So for the 127 00:09:18,560 --> 00:09:24,240 folks who need these machines, whether it's for heavy scientific applications 128 00:09:21,920 --> 00:09:29,440 in the field, visual effects, for live events, aerial surveying, industrial 129 00:09:26,240 --> 00:09:31,839 automation, outside of NVIDIA's relative 130 00:09:29,440 --> 00:09:36,640 weakness this time around, 13th gen looks like a big win. However, soon 131 00:09:35,200 --> 00:09:41,200 we're going to be getting our hands on a new generation of AMD laptops that are 132 00:09:38,880 --> 00:09:44,880 equipped with Ryzen 7000. And if the efficiency we've seen on the desktop for 133 00:09:42,959 --> 00:09:48,399 those is anything to go on, it's going to be a very interesting fight. If you 134 00:09:47,200 --> 00:09:53,760 liked this look at the latest and greatest in mobile performance, maybe you'd also like to check out the 135 00:09:51,519 --> 00:09:59,200 complete opposite. This AliExpress gaming laptop that not only didn't have 136 00:09:56,160 --> 00:10:00,720 an i9, it had a Celeron. Did it even 137 00:09:59,200 --> 00:10:03,200 have dedicated graphics? I don't think so. Yeah.