WEBVTT 00:00:00.000 --> 00:00:04.400 You want to build a sick gaming PC on a reasonable budget, but the thing is, 00:00:04.400 --> 00:00:07.600 a thousand bucks doesn't stretch as far as it used to. 00:00:07.600 --> 00:00:10.080 Even a console can cost that much these days. 00:00:10.960 --> 00:00:13.360 Maybe the solution is time travel. 00:00:14.720 --> 00:00:20.720 No, not literally. I mean taking your part selection back a few generations. 00:00:20.720 --> 00:00:23.920 But how much performance are you leaving on the table? 00:00:24.560 --> 00:00:29.600 I mean 9600 is a much higher number than 5700, 00:00:29.600 --> 00:00:31.920 so it must be a much better choice, right? 00:00:32.720 --> 00:00:40.560 Right? No. And in fact, these two CPUs are kind of the perfect springboard for our little 00:00:40.560 --> 00:00:46.240 investigation slash experiment here. Should we just stop buying the latest hardware? 00:00:46.800 --> 00:00:52.720 To find out, we built two entire systems using two wildly different philosophies. 00:00:52.720 --> 00:00:56.240 This one can only use current generation parts, 00:00:56.240 --> 00:01:00.720 while this one can reach as far back into the past as it wants 00:01:00.720 --> 00:01:06.080 in search of the best bang for the buck. Now, as you would expect, pricing is always changing. 00:01:06.080 --> 00:01:09.360 So if we built these systems a week ago or a week from now, 00:01:09.360 --> 00:01:15.200 we might get very different computers. But we're still interested in identifying any trends, 00:01:15.200 --> 00:01:20.080 tips and tricks that we can see that might help you get the most out of your next build 00:01:20.080 --> 00:01:24.640 or even your current one. We're also interested in identifying our sponsor. 00:01:34.560 --> 00:01:39.040 AMD doesn't really do new Ryzen 3 or Ryzen 5 CPUs anymore, 00:01:39.040 --> 00:01:43.840 so our brand new PC was kind of stuck with the Ryzen 7 9600X. 00:01:44.400 --> 00:01:49.200 Not that that's a terrible thing. I mean, when you consider that it uses Zen 5 00:01:49.200 --> 00:01:52.720 versus the Zen 3 we're going to be using in this system, 00:01:52.720 --> 00:02:03.520 you'd think the answer would be obvious. But our older 5700X3D has two extra cores and is packed to the gills with 3D vCache. 00:02:03.520 --> 00:02:08.080 Now, this chip hasn't quite achieved 1080Ti levels of legendary, 00:02:08.080 --> 00:02:12.640 but the more than $150 savings versus its bigger brother frees up some 00:02:12.640 --> 00:02:15.680 very interesting budget for some big upgrades later on. 00:02:16.640 --> 00:02:21.040 That's benefit numero uno of the oldest gold philosophy, choice. 00:02:21.040 --> 00:02:26.720 Not only is there generally more selection, but you can get substantial discounts on new in-box hardware 00:02:26.720 --> 00:02:31.120 that retailers are hoping to clear from their shelves. That is, as long as you're ready to strike. 00:02:32.240 --> 00:02:38.240 There can be trade-offs though. For example, our 5700X3D here uses Socket AM4, 00:02:38.240 --> 00:02:42.960 which is a dead end for future upgrades and even for some current ones. 00:02:43.680 --> 00:02:47.520 Even if you paired this chip with the best compatible motherboard on the market, 00:02:47.520 --> 00:02:52.560 you'd still be stuck with PCIe Gen 4 and DDR4 memory. 00:02:52.560 --> 00:02:58.320 Although, it's worth noting those probably aren't major considerations at a $1,000 budget. 00:02:58.320 --> 00:03:03.920 Even our SSD, we'd have probably stuck with a Gen 4 boot drive even if we had extra money to spend. 00:03:03.920 --> 00:03:08.640 Gen 5 drives cost more, don't really benefit gaming, and they run hotter. 00:03:08.640 --> 00:03:14.320 At least they do today. But who knows what the future holds for devices that'll fit in these slots. 00:03:14.320 --> 00:03:19.200 And being compatible with future hardware is a tangible advantage for Team Shiny here. 00:03:19.840 --> 00:03:25.360 It can also be an advantage today. DDR5, all other things being equal, offers a small, 00:03:25.360 --> 00:03:32.640 but measurable benefit over DDR4. And our 16 gigs of T-Force Delta RGB fit perfectly in our budget. 00:03:32.640 --> 00:03:38.480 We would have rather gone with something a little bit faster, but that kind of trade-off is kind of our point. 00:03:38.480 --> 00:03:42.400 If we want something new, we don't always get a best-in-class part. 00:03:42.400 --> 00:03:48.720 It has gotten really hard to spec a well-rounded $1,000 machine with all new Gen parts. 00:03:48.720 --> 00:03:52.880 Meanwhile, in TARDIS LAN though, we've got that same 16 gig capacity, 00:03:52.880 --> 00:03:56.320 but this time right in the price-performance sweet spot, 00:03:56.320 --> 00:04:00.640 with the only downside being that when we want to upgrade a few years down the road, 00:04:00.640 --> 00:04:05.120 we're going to need to replace our CPU, RAM, and motherboard all in one fell swoop. 00:04:05.120 --> 00:04:10.640 Speaking of motherboard, this ASUS Prime B550MA is a pretty good value, 00:04:10.640 --> 00:04:13.040 when it's on sale for around $100. 00:04:14.320 --> 00:04:21.360 Meanwhile, this Gigabyte X870 Gaming Plus Wi-Fi 6 ran us $220. 00:04:21.360 --> 00:04:30.560 That is a huge difference. But then again, so is Wi-Fi 6, PCIe Gen 5 for both the GPU and SSD, 00:04:30.560 --> 00:04:36.640 not to mention an AM5 socket, which AMD has committed to support until at least 2027. 00:04:36.640 --> 00:04:41.680 That means new chips, new drivers, and, maybe most importantly, new security updates. 00:04:42.240 --> 00:04:48.480 AMD has recently said that they will not be patching Ryzen 1000, 2000, or 3000 series chips 00:04:48.480 --> 00:04:55.520 for the sync close vulnerability. Although they backtracked on the 3000 series after some community outrage, 00:04:55.520 --> 00:04:59.440 and seeing as the latest 3000 series came out in 2020, 00:04:59.440 --> 00:05:03.040 the same year that we started getting 5000 series processors, 00:05:03.040 --> 00:05:06.240 the end of support for our older platform, oh not that one, this one, 00:05:07.120 --> 00:05:12.080 is probably coming soon. Clearly that won't matter when it comes to our next part though, the case. 00:05:12.800 --> 00:05:16.880 On this side, we've got our random old case that was cheap. 00:05:17.680 --> 00:05:21.840 And over here, we've got the Phanteks Eclipse G360A. 00:05:21.840 --> 00:05:24.960 Now in terms of size and features, they're both fairly similar, 00:05:24.960 --> 00:05:31.680 but with cases, the devil's in the details. Good ventilation, cable management that's not going to cut your fingers in, 00:05:31.680 --> 00:05:36.960 some high quality included fans. Very nice Phanteks. Meanwhile over here, this is what fit in our budget. 00:05:36.960 --> 00:05:41.760 Heck yeah. Cool, well thankfully we probably won't need as much cooling for these components 00:05:41.760 --> 00:05:47.120 because we're not going to have it. There is cable management though, and it's got one fan included. 00:05:49.200 --> 00:05:54.720 I expect our Corsair CX650W to give us a better long-term experience than our 00:05:54.720 --> 00:06:00.720 $41 thermal take smart 600W. But it was the only thing that fit in this budget. 00:06:00.720 --> 00:06:05.600 And honestly speaking, even our better Corsair could end up near its limit. 00:06:05.600 --> 00:06:11.840 If you were to replicate these builds, we would strongly recommend a juicier power supply for both of these systems. 00:06:11.840 --> 00:06:15.600 And if you are by the way, we're going to have all the parts linked down below. 00:06:16.960 --> 00:06:20.400 You'll notice that what's in the system is not actually the 600W one. 00:06:21.440 --> 00:06:25.040 This is such a crappy power supply, it wouldn't even boot the system. 00:06:25.040 --> 00:06:28.800 Is this over budget now? It's a $20 more for 60 more watts. 00:06:28.800 --> 00:06:32.640 Seems to be doing the trick, but yeah. 00:06:32.640 --> 00:06:37.360 For our cooler, we settled on the thermal write phantom spirit. It's so nice, we bought it twice. 00:06:39.200 --> 00:06:43.120 Yep, this thing is so good, we used the same one for both systems. 00:06:43.120 --> 00:06:48.080 Which actually brings us to the big difference, our GPUs. 00:06:48.800 --> 00:06:52.080 Until the NVIDIA RTX 5000 series comes out. 00:06:52.080 --> 00:06:55.920 And I guess also maybe eventually has an option that doesn't cost 00:06:55.920 --> 00:07:01.840 more than our entire budget. We're stuck with the Zotac TwinEdge RTX 4060. 00:07:01.840 --> 00:07:06.400 To be clear, it's not a terrible card. It's got decent 1080p gaming performance, 00:07:06.400 --> 00:07:11.440 with the usual NVIDIA features that turn it into a far more compelling package than its 00:07:11.440 --> 00:07:13.920 wimpy 8GB of VRAM would normally allow. 00:07:14.560 --> 00:07:17.840 I can see why it's popular compared to its AMD equivalent. 00:07:17.840 --> 00:07:25.040 But here's the thing. This XFX Speedster Radeon 7800XT is not an equivalent card. 00:07:25.040 --> 00:07:30.800 We were able to spend an extra $200 on it thanks to our frugality elsewhere. 00:07:30.800 --> 00:07:35.920 So even though it's hard to compare spec for spec across Team Green and Team Red, 00:07:35.920 --> 00:07:39.520 we can probably expect a huge performance delta. 00:07:39.520 --> 00:07:42.800 Not to mention the double PCIe bandwidth. 00:07:42.800 --> 00:07:46.160 Of course, in theory, that shouldn't impact the real world too much. 00:07:46.160 --> 00:07:50.000 But hey, why am I talking about theory? It's time to put these systems to the test. 00:07:50.960 --> 00:07:56.640 This... Representing Team Brand New will be David while I will be taking our old machine. 00:07:56.640 --> 00:08:00.320 Not because I know the outcome and I like winning. 00:08:00.320 --> 00:08:04.000 Because you're old. But you made that one too easy. 00:08:04.000 --> 00:08:07.760 Let's start with a Cinebench Drag Race. Ready and go! 00:08:13.600 --> 00:08:16.240 I'm tired. I'm going to sit down. I guess I'm the old one. 00:08:17.200 --> 00:08:19.280 My knee's cracked. 00:08:21.280 --> 00:08:22.400 I'm totally out of head! 00:08:24.560 --> 00:08:29.440 He must have been trying to distract from the fact that the all-new system is pulling ahead. 00:08:29.440 --> 00:08:35.520 You know what? The difference isn't that much. Especially when you consider that that is two generations newer. 00:08:35.520 --> 00:08:39.440 But then I'm not really sure what I expected because I have two more cores. 00:08:39.440 --> 00:08:42.560 Let's move on to gaming. I love what you've done with the backgrounds, by the way. 00:08:43.120 --> 00:08:45.040 Abe Simpson is my familiar. 00:08:46.240 --> 00:08:50.320 It follows me around everywhere. No one else can see him. I would think 1440p. 00:08:50.320 --> 00:08:53.360 Yeah, that's what I've been testing everything. 120. 00:08:53.360 --> 00:08:56.800 That's going to be pretty rough. 4060. Not my problem. 00:08:57.360 --> 00:09:00.080 I would also be curious to run it at a lower resolution to see... 00:09:01.280 --> 00:09:07.040 If your faster cores are a huge benefit or are they even faster? 00:09:07.040 --> 00:09:08.080 I got X3D. 00:09:10.160 --> 00:09:15.360 Your game just crashed. Booted me out. We were having some trouble with Expo on this system. 00:09:15.360 --> 00:09:20.080 That's the thing. It's technically overclocking. We might need to go back to Jdex settings. 00:09:21.600 --> 00:09:25.120 That's not going to help your performance numbers, but it's the luck of the draw. 00:09:26.320 --> 00:09:29.760 Hey, your performance is looking pretty good. Yeah. 00:09:29.760 --> 00:09:32.240 350. 350 FPS. I'll take it. 00:09:33.920 --> 00:09:37.520 Oh, I thought you had them. Oh, I missed. I like flinched. 00:09:37.520 --> 00:09:42.320 I'm a flincher. Well, you haven't dropped below 200 and I seem to live there. 00:09:43.040 --> 00:09:47.200 It's almost like having a way better GPU makes a pretty big difference for gaming. 00:09:47.200 --> 00:09:53.040 Who'd have known? Let's test my theory that I could switch to very high and still have better FPS than you. 00:09:54.080 --> 00:09:58.080 That's very high? Yeah. Oh, hey, I'm winning. 00:10:00.240 --> 00:10:08.000 Barely, though. Barely. This is an entire notch up in in-game settings and about the same level of performance. 00:10:08.880 --> 00:10:12.160 And I'm noticing my 1% lows dropping to below 100 too. 00:10:13.120 --> 00:10:17.600 Actually, yeah. My 1% lows are significantly better than yours. 00:10:18.240 --> 00:10:24.960 Let's do Final Fantasy XVI next. After about a one-minute unskippable cutscene, we can jump right to the pretty epic icon fight. 00:10:25.680 --> 00:10:31.040 That'll give us a pretty good indicator of what our performance is going to be. And this is a very on-topic game right now, 00:10:31.040 --> 00:10:35.920 since this is one of the big PlayStation 5 exclusives that has just come to PC. 00:10:35.920 --> 00:10:40.800 Right on the heels of launch of the PS5 Pro. And so if you guys want us to make a PS5 Pro killer, 00:10:40.800 --> 00:10:45.920 give this video a thumbs up, and we will make that happen. We wouldn't even need to save that much money compared to this budget, 00:10:45.920 --> 00:10:47.360 with how expensive that thing is. 00:10:48.960 --> 00:10:53.440 Oh, no. Oh, wow. 00:10:53.440 --> 00:10:57.440 Oh, no. Not even 30. 00:10:57.440 --> 00:11:01.120 That's not close, dude. That was chugging. 00:11:01.120 --> 00:11:09.360 Yeah, that was below 20. I mean, NVIDIA's got a lot of cool technology to make up for having an underpowered GPU, 00:11:09.360 --> 00:11:13.280 like Framegen. But wouldn't you rather just have more frames? 00:11:13.280 --> 00:11:15.760 This is definitely a game that's designed around upscaling, 00:11:16.560 --> 00:11:20.320 which is a bit unfortunate, but this is pretty hopeless, 00:11:20.320 --> 00:11:22.080 considering this is already 4, 240p. 00:11:23.440 --> 00:11:26.880 But even I'm having frame drops in the video. 00:11:27.600 --> 00:11:30.880 I also couldn't help noticing my GPUs running 20 degrees cooler than yours. 00:11:30.880 --> 00:11:35.280 Holy crap. There's something to be said for investing in a case with way better airflow. 00:11:35.920 --> 00:11:42.320 I mean, I don't know how much it would matter. My GPU is pinned at 100% just playing this freaking video. 00:11:42.320 --> 00:11:50.480 This may be an engine, but capped. But because yours can barely handle it, it's running at 100%, whereas mine's not. 00:11:50.480 --> 00:11:53.920 That makes a lot of sense. Chugging, dude. It's chugga-lugging. 00:11:54.480 --> 00:11:58.080 Look at that. 15 FPS. 00:11:58.080 --> 00:12:03.280 This is a thousand dollar PC. Like this is a pretty hefty investment. 00:12:03.280 --> 00:12:06.800 And you haven't just blown money on dumb stuff either. 00:12:06.800 --> 00:12:09.920 You've spent too much on your motherboard by getting a top tier chipset. 00:12:09.920 --> 00:12:15.760 Second to top tier. That's true. But other than that, you haven't really done anything stupid. 00:12:15.760 --> 00:12:18.960 No. I legitimately tried to make a good thousand dollar build. 00:12:19.920 --> 00:12:23.440 It's hard. You just tried a lot harder over here. No, it's just easier. 00:12:23.440 --> 00:12:28.480 There's so many more choices. And if you guys think, oh, you guys just cherry picked 00:12:28.480 --> 00:12:31.440 a couple of games that run way better on Linus's machine, 00:12:32.080 --> 00:12:36.880 that's not the case at all. David already pre-benchmarked a handful of other games. 00:12:36.880 --> 00:12:40.560 And even if we were able to get Expo running on this machine, 00:12:40.560 --> 00:12:43.680 it's not going to make up the kind of difference we're observing here. 00:12:43.680 --> 00:12:47.200 It's night and day. I'm tempted to turn on Framegen. 00:12:47.200 --> 00:12:50.320 Sure. And see if you can even make up the ground using that. 00:12:50.880 --> 00:12:54.320 Okay. Well, that gets you there. 00:12:54.400 --> 00:12:58.480 It's noticeably blurrier. You can tell that it's lower resolution. 00:12:58.480 --> 00:13:00.720 But it's not bad. How does it feel? It's pretty good. 00:13:01.920 --> 00:13:05.440 It's pretty snappy. I'm not upset. We're making a pretty good case for a DLSS. 00:13:07.360 --> 00:13:11.040 This went from a pretty unplayable mess to it's not bad. 00:13:11.040 --> 00:13:13.520 Okay. What about if we're looking off into the distance? 00:13:14.240 --> 00:13:19.360 Okay. Like how's the water look? Yours definitely has a little more detailed resolved. 00:13:20.080 --> 00:13:23.920 This is kind of, it almost feels less contrasty, even though these are the same monitors. 00:13:23.920 --> 00:13:26.720 Yep. And it's the same in the textures on the character model. 00:13:27.600 --> 00:13:30.400 It's clearly a little bit sharper for me. 00:13:31.440 --> 00:13:34.960 Oh yeah. You can really see that in the chain mail. Yeah. And to your point. 00:13:34.960 --> 00:13:38.560 Oh, ew. Oh, ew. Oh, not good. 00:13:39.120 --> 00:13:45.360 So it can be pretty distracting sometimes. And I haven't turned on any of my upscaling. 00:13:45.360 --> 00:13:49.120 If you just want to see FPS number go up, I got that too. 00:13:49.120 --> 00:13:54.240 I really did not get a lot of extra frames. And I just got a sharpness filter applied to everything. 00:13:55.280 --> 00:14:01.040 Dang it, FSR. Not nearly as good. Hey, and I'm getting my 30 FPS cutscenes now. 00:14:01.040 --> 00:14:04.320 I did it. Incredible. 00:14:04.320 --> 00:14:10.800 DLSS and FrameGen really is the compelling argument to make for the modern gen system. 00:14:10.800 --> 00:14:15.040 And it's worth noting that if you strongly do prefer one brand of GPU over the other, 00:14:15.040 --> 00:14:18.960 it's not like we couldn't have put an NVIDIA card in our older gen system. 00:14:18.960 --> 00:14:21.680 We could have gone with 3070 or something like that. 00:14:22.480 --> 00:14:30.560 You can even get some really great deals on 3080s. But we decided for the sake of this video that we weren't going to look at used cards. 00:14:30.560 --> 00:14:37.360 We were just going to stick to stuff that you could buy brand new. All that's left now is to answer the question of which of these systems would you buy? 00:14:37.360 --> 00:14:42.160 Let us know in the comments, the oldie but goodie or the shiny new toy. 00:14:42.160 --> 00:14:47.040 And also to talk about what we've learned. First of all, timing is everything. 00:14:47.040 --> 00:14:52.480 The AM5 platform is coming up on three years old with the AM4 platform crossing the eight-year 00:14:52.480 --> 00:14:58.640 mark just recently. So it's no wonder that there's lots of deals to be had here and not so many over here. 00:14:58.640 --> 00:15:04.880 But you can't take that for granted. Once NVIDIA cleared out their 3000 series charts, they cleared them out. 00:15:04.880 --> 00:15:08.640 And while you still can find them brand new from time to time, it's pretty rare. 00:15:08.640 --> 00:15:13.440 Another thing we learned is that generational gains don't seem to trump the savings of buying 00:15:13.440 --> 00:15:16.960 older hardware. At least it's not a guarantee. 00:15:16.960 --> 00:15:22.960 There were a lot of challenges in picking the parts for these. There's going to be a Floatplane exclusive closer look where David's going to be discussing 00:15:22.960 --> 00:15:27.440 how allowing ourselves to look at used hardware might have affected our choices. 00:15:27.440 --> 00:15:31.920 Either way, it's pretty clear that you should not lock yourself into the newest releases. 00:15:31.920 --> 00:15:35.280 PC hardware is not evolving at the same pace that it once was. 00:15:35.280 --> 00:15:39.920 And you just don't need the latest gear to have an excellent experience. 00:15:39.920 --> 00:15:47.040 What you do need is the segue to our sponsor. If you guys enjoyed this video and you like budget builds, why not check out our complete 00:15:47.040 --> 00:15:52.160 edit of Scrapyard Wars 9 where Luke and I competed to build the best PC for $800. 00:15:53.040 --> 00:15:56.960 No offense, David, but we crushed you. 00:15:56.960 --> 00:15:59.520 It's you. I know it's you. It's a whole different game.