1 00:00:00,000 --> 00:00:05,000 Holy crap, it's the world's fastest processor again. 2 00:00:05,320 --> 00:00:10,640 You know, if you look at the fine print, it says they're reaching the cutting edge of CPU frequency 3 00:00:10,640 --> 00:00:15,440 to release the fastest processor ever. But if you start reading between the lines 4 00:00:15,440 --> 00:00:19,520 and specifically down to the footnotes, you might notice that by fastest CPU, 5 00:00:19,520 --> 00:00:22,840 the only thing they mean is the frequency. Sure, it is fast. 6 00:00:22,840 --> 00:00:27,840 We are talking a 4900KS with a whopping 6.2 gigahertz 7 00:00:28,840 --> 00:00:34,760 max turbo frequency, apparently. That is the fastest, that's the fastest they've ever done. 8 00:00:34,760 --> 00:00:38,000 I mean, wow, packaging-wise. They haven't changed anything in a while. 9 00:00:38,000 --> 00:00:41,640 I mean, this is in theory the last of a generation. 10 00:00:41,640 --> 00:00:45,640 The next Intel desktop CPUs are very likely going to be 11 00:00:45,640 --> 00:00:49,480 not monolithic anymore. And by that, I mean one single die. 12 00:00:49,480 --> 00:00:54,280 They're gonna be switching to a design somewhat similar to AMD with chiplets. 13 00:00:54,280 --> 00:00:59,680 But this is the new fastest processor, the Intel Core i9-14900KS. 14 00:00:59,680 --> 00:01:03,920 The video is done, I'm done unboxing it. You can see it right here. That is all you need to know. 15 00:01:03,920 --> 00:01:09,280 Jokes aside, it is a very similar CPU. You're talking the same core counts. 16 00:01:09,280 --> 00:01:13,160 You're talking the same architecture. The only thing that's really changed 17 00:01:13,160 --> 00:01:16,800 is the max turbo frequency and the power limits. 18 00:01:16,800 --> 00:01:20,340 You're looking at 6.2 gigahertz, like I mentioned earlier, 19 00:01:20,340 --> 00:01:26,200 which is up from 6 gigahertz on the 1400K. Was the 1300KS also 6 gigahertz, 20 00:01:26,200 --> 00:01:34,120 if I remember correctly? I'm getting a nod of approval, so yes. And the power limit, 320 watts on the PL1 and PL2 21 00:01:34,760 --> 00:01:38,280 and 150 watt base power, which is insane. 22 00:01:38,280 --> 00:01:43,680 I mean, I think about my 1300K at home. If I turn MCE on with like no limits, 23 00:01:43,680 --> 00:01:47,520 with a 360 mil rad and an EK water block, just cooling it. 24 00:01:47,520 --> 00:01:52,260 I'm stuck at like 340, 350 watts with it sitting at 90 degrees. 25 00:01:52,260 --> 00:01:54,920 I don't see how this is gonna be that much better. 26 00:01:55,920 --> 00:01:59,680 I'm sure it's gonna draw a ridiculous amount of power. I mean, it has to. 27 00:01:59,680 --> 00:02:05,360 They want you without even turning MCE on to be hitting 320 watts on this thing. 28 00:02:05,360 --> 00:02:09,320 I can't imagine this is gonna perform that much better. I'm getting some weird looks, 29 00:02:09,320 --> 00:02:14,040 but seriously, there's not that much more to say. I mean, other than just trying it out, 30 00:02:14,040 --> 00:02:17,040 I suspect what's gonna happen is we plug this in 31 00:02:17,040 --> 00:02:21,160 and we look at the lab's results and this is going to perform marginally better, 32 00:02:21,160 --> 00:02:24,760 maybe a couple percent at drastically worse power, 33 00:02:24,760 --> 00:02:29,360 which is kind of how it usually is. This is a scary precedent, is all I'm saying. 34 00:02:29,360 --> 00:02:32,800 I mean, you look at AMD, you get like a 7800X3D, 35 00:02:32,800 --> 00:02:35,840 that's a power efficient chip. Really good for gaming. 36 00:02:35,840 --> 00:02:38,920 If you're just gaming, I don't know about this, 37 00:02:39,920 --> 00:02:44,080 but hey, only a little bit prior judgment. Let's just see how it actually performs 38 00:02:44,080 --> 00:02:47,360 with this cooler on here. What is this, the 360 mil? 39 00:02:47,360 --> 00:02:55,520 What do they call this, the Ryujin 360? It's like the easiest ShortCircuit ever. 40 00:02:56,800 --> 00:03:00,240 Guess what, nothing changed. Our test bench is running, we're in Windows, 41 00:03:00,240 --> 00:03:04,240 I can't wait to see how obnoxiously power hungry this is. 42 00:03:04,240 --> 00:03:08,000 Do we have Cinebench? Let's run Cinebench, I don't see Cinebench. 43 00:03:08,000 --> 00:03:12,520 I mean, I can see it, turboing. I see 5.9 gigahertz, I don't see 6.2. 44 00:03:12,520 --> 00:03:16,560 Cause that's, in theory, that's just in the most optimal of circumstances, 45 00:03:16,560 --> 00:03:22,080 one core will do that. That's the idea, likely. So here, let's do Cinebench, one core. 46 00:03:22,080 --> 00:03:23,800 I've seen a lot of 5.9. 47 00:03:26,200 --> 00:03:29,440 Oh, I see 6.2, holy shit, right there, look at that. Oh, it went away. 48 00:03:29,440 --> 00:03:34,160 It was there for like a second. See, maximum, 6.2, hell yeah. 49 00:03:34,160 --> 00:03:37,680 Intel marketing confirmed, they can do it boys and girls. 50 00:03:39,080 --> 00:03:44,080 I mean, it was there for like a second, let's try again. Oh, look at that, 6.2, holy shit. 51 00:03:45,080 --> 00:03:50,080 That was first, it recorded 6.2 on one second increment. 52 00:03:54,000 --> 00:03:58,840 That's it. Wow, I really feel like I'm getting what I paid for. 53 00:03:58,840 --> 00:04:04,600 Now I'm sure the average clock speeds are probably a similar amount higher, 54 00:04:04,600 --> 00:04:08,360 but how much power are we drawing? Okay, this is only a single core test, 55 00:04:08,360 --> 00:04:12,320 and we're looking at 80 watts. Let's do multicore. 56 00:04:12,320 --> 00:04:15,760 Let's see the big scary number. It's probably gonna be big and scary. 57 00:04:15,760 --> 00:04:20,720 I'm scared and I haven't even done it yet. The AIO is like ripping now. 58 00:04:20,720 --> 00:04:25,720 Yeah, it hits 320 watts. 59 00:04:25,800 --> 00:04:32,040 The turbo power wasn't a lie. That's like out of the box, put it in a motherboard. 60 00:04:32,040 --> 00:04:38,680 It's doing 320 watts. I mean, with a 360 mil AIO, we're just vibing at like 85. 61 00:04:40,000 --> 00:04:43,840 I'm gonna restart this thing and turn all the limits off, but in the meantime, let's look at the Labs results 62 00:04:43,840 --> 00:04:48,560 because that's more proper testing than I just did right there. 63 00:04:48,560 --> 00:04:53,920 But I was right, it draws a lot of power. Let's see if it actually performs any better. 64 00:04:53,920 --> 00:04:59,680 Now, before I look at these Labs results, I'm gonna look at the price, 700 US dollars. 65 00:04:59,680 --> 00:05:06,400 Oh my God. A 1400K kind of the current retail suggested price is 560, 66 00:05:06,400 --> 00:05:10,860 which makes this a 140 US dollar upcharge. 67 00:05:10,900 --> 00:05:14,820 Oh my God. It's even more than a 7950X. 68 00:05:14,820 --> 00:05:20,580 Not that that is a well-priced CPU presently, but geez, let's look at how it performs. 69 00:05:20,580 --> 00:05:23,400 I bet you it's not gonna be $140 more. 70 00:05:24,380 --> 00:05:31,020 Hey, look, it's not. Oh my God, is it even better? Like Cyberpunk, exact same performance as the non-KS. 71 00:05:31,020 --> 00:05:34,940 F1 2023, three more FPS average. 72 00:05:34,940 --> 00:05:40,300 Red Dead, two more FPS average. The average of all of the games put together, 73 00:05:40,300 --> 00:05:45,300 one additional FPS. It's weird to think that like 200 megahertz more 74 00:05:45,300 --> 00:05:50,540 is not a big deal anymore. But when you think about it, when you're already talking six gigahertz, 75 00:05:50,540 --> 00:05:53,980 you know, bumping up 200 megahertz more is, 76 00:05:53,980 --> 00:05:57,440 it's a lot smaller percentage and it shows. 77 00:05:57,440 --> 00:06:02,900 I mean, God, look at that. Non-gaming tests, it's actually slower in some circumstances. 78 00:06:02,900 --> 00:06:07,600 Probably because it's thermal throttling would be my guess. It's really gonna depend on the cooler you have. 79 00:06:07,600 --> 00:06:11,380 Definitely make sure you're running this thing on like 360 mil AIO. 80 00:06:11,380 --> 00:06:14,380 I mean, can an NHD 15 even run this thing anymore? Probably not. 81 00:06:15,720 --> 00:06:19,660 Sort of can. Yeah, you basically just need a water cooler now. 82 00:06:19,660 --> 00:06:23,100 Just custom cool your computer. You're spending 700 US dollars on a CPU. 83 00:06:23,100 --> 00:06:26,900 I mean, why not, right? Let's see. 84 00:06:26,900 --> 00:06:30,700 Yeah, okay, with the Ryujin 3 360, 85 00:06:30,700 --> 00:06:33,860 Cinebench 2024, the average was 87 degrees 86 00:06:33,860 --> 00:06:36,900 on the 4900KS with a max of 100. 87 00:06:36,900 --> 00:06:43,020 If you turn on MCE with no limits, 96 degree average, 105 degree max, 88 00:06:43,020 --> 00:06:46,780 it's pretty much across the board, at least 10 degrees hotter. 89 00:06:46,780 --> 00:06:50,460 It's actually more than 10 degrees. 13 degrees hotter than a 1400K. 90 00:06:50,460 --> 00:06:54,820 I mean, that makes sense. We're talking like 250 watt to 320 watt power limit 91 00:06:54,820 --> 00:06:58,900 increase here. So yes, it's going to be hotter naturally. 92 00:06:58,900 --> 00:07:03,900 Oh my God, you guys measured 367 watts peak. 93 00:07:03,900 --> 00:07:07,820 Turns a Cinebench run with MCE on. 94 00:07:07,860 --> 00:07:13,020 Oh my God. That's literally more power than a Threadripper will draw 95 00:07:13,020 --> 00:07:16,220 without turning PBO on. That's insane. 96 00:07:16,220 --> 00:07:23,180 I mean, if you look across the board in terms of the amount of power you're consuming per frames, 97 00:07:23,180 --> 00:07:26,860 you are kind of delusional to go Intel 98 00:07:26,860 --> 00:07:30,980 if you live anywhere where power costs money. Is that safe to say? 99 00:07:31,900 --> 00:07:35,100 I don't know. Maybe if you live with your parents, you get free power, right? 100 00:07:35,100 --> 00:07:38,500 That's how that works. I don't. 101 00:07:40,500 --> 00:07:43,820 Oh God. It's somehow slower in Cinebench multicore, 102 00:07:43,820 --> 00:07:47,300 which is interesting. The single core of course is like a couple points higher. 103 00:07:47,300 --> 00:07:50,420 It literally four points across the board. 104 00:07:50,420 --> 00:07:55,420 I could see these variations in just buying multiple CPUs. 105 00:07:55,740 --> 00:07:59,940 We've seen our own variations, at least on AMD that were this large or larger even. 106 00:07:59,940 --> 00:08:03,300 But let's see how ridiculous it is in person. 107 00:08:03,300 --> 00:08:05,100 Remove all limits, no temp. 108 00:08:06,780 --> 00:08:11,580 Let's go. It reminds me of when they came out with the five gigahertz eight core. 109 00:08:11,580 --> 00:08:14,660 And everyone was like, oh my God, that's so much power. 110 00:08:14,660 --> 00:08:17,900 Oh, we only knew where we were headed. 111 00:08:17,900 --> 00:08:22,900 233, 358, 389, 389, oh my God, 388. 112 00:08:32,740 --> 00:08:38,740 It's at 102 degrees though. You're just at a point where regardless of how much, 113 00:08:38,740 --> 00:08:44,740 you know, how many radiators you have, the size of the die is not big enough for this much power. 114 00:08:44,740 --> 00:08:49,940 Unless you're de-lidding this thing in like direct dye cooling, or you're putting probably a copper IHS on it 115 00:08:49,940 --> 00:08:54,660 or something like that. We're getting down after a little bit of chilling here, 116 00:08:54,660 --> 00:08:59,660 370 watts-ish, and just casually idling at 102 degrees. 117 00:09:02,940 --> 00:09:06,540 It's actually beyond, wow, look at that. Oh God, you know what, it occurs to me. 118 00:09:06,540 --> 00:09:10,860 We have a 4090 here. What is a 4090 vibe at? Like 600 watts, 500 watts. 119 00:09:10,860 --> 00:09:15,940 This is almost a kilowatt of power draw. I guess if you take the power supply efficiency 120 00:09:15,940 --> 00:09:19,420 in there too, you're basically at a kilowatt of power draw 121 00:09:19,420 --> 00:09:23,500 if you were able or using an application that would load both your CPU and your GPU. 122 00:09:23,500 --> 00:09:29,420 What are we supposed to be saving the planet or something? It's not really faster, aside from like a percent or two. 123 00:09:29,420 --> 00:09:34,380 I don't even know what to say. I mean, props to you Intel. Congratulations, smack on the back. 124 00:09:34,380 --> 00:09:38,460 You made the world's fastest, technically speaking, 125 00:09:38,460 --> 00:09:43,060 frequency CPU. If only it was the fastest everywhere else. 126 00:09:43,060 --> 00:09:46,700 It is very expensive. It is very high on power draw. 127 00:09:46,700 --> 00:09:51,220 Unless you're just like the dude that wants to have the most ballinous CPU at the LAN party. 128 00:09:51,220 --> 00:09:54,980 It's a flex thing. It's like the 8086K, you know? 129 00:09:54,980 --> 00:09:59,020 It's purely just for ballers. If you hate money, you wanna light it on fire, do it. 130 00:09:59,020 --> 00:10:02,180 But otherwise, don't. Please. 131 00:10:02,180 --> 00:10:06,340 It's like I went on Amazon and sorted by most expensive. That's who this CPU is for. 132 00:10:06,340 --> 00:10:10,020 I'm a little sad. It's basically the last monolithic Intel CPU. 133 00:10:10,020 --> 00:10:14,180 I can't imagine they're gonna be releasing anything else. Like, it definitely not top tier. 134 00:10:14,180 --> 00:10:18,580 It's the end of an era. It's just, this is the most that this could do. 135 00:10:18,580 --> 00:10:23,940 This is the send off. It's like when a rocket ship launches and it explodes midair before it gets there. 136 00:10:23,940 --> 00:10:27,380 That's what this feels like, because it was hot. 137 00:10:27,380 --> 00:10:31,740 Good job, Intel. And good job for you for watching. Hit the like button, get subscribed. 138 00:10:31,740 --> 00:10:34,900 Maybe even check out the video that was also hot 139 00:10:34,900 --> 00:10:38,880 because we were in a hot tub. It was not hot heated. 140 00:10:40,220 --> 00:10:41,060 Bye.