Gadgetory

- You guys have been wanting this one for quite a while, and here I have all these eight core behemoths, right here on the desk and on my head, so we've got the Ryzen 7 2700X at 4.2 gigahertz, going up against the 9900K at 5 gigahertz. And then we're going back to 2013, taking that Xeon, which is the 1680v2, and we're clocking that to 4.6 gigahertz. We're gonna be testing these 3 CPUs with the RTX 2080 Ti Aurus. Now I did set the fans speeds to 100%, up the power limits to 122%, and it's an even apples to apples to apples test. As for the motherboards, we've got the ASRock Taichi Ultimate across both the Ryzen and also the 9900K. As for the Xeon that's got the Rampage IV Extreme, X79 motherboard. And what are we waiting for? Let's do this. That was the 8700K box by the way, I was just gonna pretend it's a Xeon box but there's nothing in the box. (" I Wish That I Was a Mad Man" by Staffan Carlen) - Welcome back to Tech Yes City and right here is part one of three with the ultimate showdown. The whole purpose of this if you haven't guys seen already, I did a prelude where I talk about X79, is that we're looking at 2013 versus now, 2018. And we're gonna see how far teach has come in the last five years, at least for a single end power user or someone that plays games. In other words, I guess the majority of my audience and myself included. And so we've tested seven different games here at 1080p ultra and also 1440p ultra. This will be part one, we're looking at the gaming performance and then in part two, we're gonna take a look at the input lag, and then in part three, we're gonna be taking a look at the productivity. So let's have a look. First up, pulling up CS GO for you guys, 1080p we saw that 9900K coming out well ahead of the pack here. 558 average FPS, 1% low of 141. Contrast that to the 1680 v2, that was getting 430 average FPS, with a 1% low of 122. Then we look at the Ryzen 7 2700X, that was getting 358 FPS, so I think in CS GO this was probably the weakest showing for the Ryzen CPU. But moving along now to Dota 2, and we saw here, 211 FPS with Vulkan. And this time you guys have been telling me, look I wanna see the Vulkan API tested in Dota 2, So I decided to switch things up and test Vulkan across the board here, and 211 FPS, 98 1% low versus 191 and 78 on the Xeon. And then going to Ryzen 162 and 61. So the intel variants were pulling ahead of the Ryzen in this game, and moving over to 1440p showed a similar scenario, with 151 on the Ryzen versus 206 on the 9900K versus 194 on the Xeon 1680 v2. Moving over to Far Cry 5, the game that still loves the single core performance, as well as the higher clock speeds. So, it was a no surprise to see the 9900K coming out at 1080p with 154 average FPS and a minimum of 124 and then looking at the Xeon 1680 v2, it scored 116 with a minimum of 88, and the Ryzen 7 2700X at 4.2 gigahertz scored 107 with a minimum of 83. And then stepping over to 1440p saw the FPS drop on 9900K to 138, on the Xeon 113 and then on the Ryzen it scored 107 again. So pretty much plateauing out at that resolution. But one thing I will say with the RAM as well, memory, quick interlude here before we move on with more of the benchmarks, DDR4 memory, Sniper X G.Skill, at 3600 megahertz, CL18, and then looking at the Xeon that was using 1600 megahertz at CL8. So from the Xeon also use that in quad channels, since it can support that, so that brought the total bandwidth of the memory roughly that of the same of the 9900K and also the 2700X. Anyway, back to GTA 5 at 1080p we had some engine busting performance here on the 9900K, scoring 167 with a 1% low of 3. That's right, 3. So this test I decided to take off MSAA completely, and just let the CPU run to the hill, moving over to both the Ryzen and the Xeon, we scored very similar scores here, 133 with a 1% low of 111 versus 132 on Ryzen and 104. Then stepping over to 1440p, the 9900K actually did better scoring 171 average FPS, but again a 1% low of 4, so it was still a horrible experience. And then, the Xeon did 130 and the Ryzen did 126, so both the slower CPUs did do better in GTA 5 because they're not breaking the engine, and if you go 9900K and it's clocked at 5 gigahertz, you may wish to manually cap the frames to around 160 in this game in particular, I find after that really just starts going crazy and breaking the engine. Moving over now to the latest and greatest Battlefield V, we've got 177 average FPS at 1080p with the 9900K with a 1% low of 118 versus 168 on the Ryzen 7 2700X versus 165 on the Xeon. So the Xeon actually surprisingly got the highest 1% lows here. And this is actually a trend that I'm noticing through a lot of these benchmarks, was the 1% lows. And interestingly enough, the Xeon has 25 megabytes of level three cache on- did I just say cache? Cache on board the CPU. So stepping things up now to 1440p saw the 9900K 136, the Ryzen 7, 134 so really closing that gap, and then the Xeon scoring 128. And then the 1% lows were in tandem with the clock speeds in this particular benchmark from 5 gigahertz to 4.6 to 4.2. Moving across now the latest reiteration of Hitman at 1080p 242 average FPS 9900K, and then moving across to the Xeon we scored 201 FPS, and then the Ryzen 7 2700X 191. And the stepping things up to 1440p on the Ryzen saw 182 average FPS, the Intel scoring 233, and then the Xeon scoring 192. All the 1% lows were really well controlled, very smooth experience in this game on all three CPUs. And then moving over to Assassin's Creed Origins we're using the in built benchmark but I still used MSI Afterburner overlay where I can capture the 1% lows during this benchmark. And we saw were 1080p on the ultra high settings, so this is max the slider will go with the 2080 Ti we saw 112 average FPS at 1080p on the 9900K, 59 1% low, stepping across to the Xeon, that got 96 average FPS with a 1% low of 33. And then looking at the Ryzen 7 2700X 84 with 41. Stepping things up to 1440p we got 83 and 39 on the Ryzen, 100 and then 58 for the 1% lows on 9900K, and the for the 1680 v2, 9900K, 90 and 32 respectively. So this game pretty much scaled along those clock speed numbers, even though the IPC and the architectures are different, we had the 5 gigahertz coming out on top, 4.6 in second and 4.2 in third place. And there it is ladies and gentlemen, there are the gaming results with the 1680 v2, the 2013 Xeon at 4.6 gigahertz versus the Ryzen 7 2700X at 4.2 gigahertz and then the 9900K, the latest and greatest at 5 gigahertz. But the thing is, we'll look at the power consumption results and I'll probably show this in the other two part series as well, this is getting over 200 Watts usage when it's at 5 gigahertz. The idle power consumption is pretty good at just under 40 Watts, but the Xeon at 4.6 gigahertz, which is five years old, using around 180 Watts, and then the Ryzen's close to around 140 Watts. So when we look at what came out in the 2013 and we fast forward now to 2018, not a whole lot has changed. At least for someone like me, who plays games occasionally with a CPU like this or someone out there who's maybe looking for a bargain X79 motherboard, don't be deterred from getting the X79 platform. If you can get a really good deal on it, or if you're an enthusiast and you wanna travel back to the future and get something that you can enjoy. Because what we saw in today's video was a 9900K that's pushed to the brink, and when we look a that out of the box, that's getting 4.7 gigahertz, you can only really get that thing realistically an extra 300 megahertz higher. The Xeon's coming out of the box with a 3 gigahertz base clock speed. We're taking that to 4.6 gigahertz, that's over a 50% overclock. Then we look at the Ryzen 7 2700X, that's pretty much boosting to 4 gigahertz out of the box automatically, so we're getting an extra 200 megahertz on that. So there's something that we can see here, with these three CPUs. The one from 2013 had a lot of headroom, the two CPUs from both AMD and Intel in 2018, are being pushed higher to the brink. I guarantee you if you got this CPU, and you got a really bad power supply, and you got a really bad, or just a really entry level Z390 motherboard and you tried running it, even out of the box you might have problems, it might crash. You take that Xeon 1680 v2, you chuck it in with the crapest power supply money can buy, chuck it with the crapest X79, even a Huanan board, and I think that thing will boot and work at its base clock speeds. And that's a big difference between 2018 versus 2013 with these CPUs I think the envelope is being pushed to the brink and so we're not really seeing true gains. When we look at the 4.6 gigahertz CPU, it's beating the Ryzen in pretty much all those benchmarks, Battlefield V being the exception. And I believe maybe that has to do with some coding, using FMA3, which is an instruction set that's not on the Ivy Bridge-E CPUs, the 2013 release. AVX is still there and so a lot of the other titles are doing very well, Hitman included. That was beating out the Ryzen 7 2700X, but of course, the 9900K is still coming out on top, it's got 5 gigahertz clock speed, and it's got a slightly better IPC than the 1680 v2. But really, when it comes down to it, in five years I'm surprised, not a whole lot has changed. And that's the shocking truth of what's going on here. And as we said in the previous video with the prelude, if you haven't seen already I'll put the link up here, for what use the 1680 v2 for, I don't need all those extra like the DMI 3.0, I don't need the reduction in PCIe lanes from 40 to 20- (laughs) (bleep) I also don't need USB 3.1, I don't need ThunderBold 3, but what I do need is the best IPC, the fastest clock speeds so I can get the snappiest experience known to man. But in this case, I think that snappiest experience could've been five years ago. And you're gonna find out in part two of this video, where we're gonna be testing the input latency with 1000 FPS camera. And one thing I will say it is that we did try to keep this as apples to apples as possible, so we used the same monitor, the Asus XG35, we used the same SSDs. We did use a better power supply on the Ryzen and also the 9900K machines, the AX1500 and we did use 3600 megahertz G.Skill memory, and I slightly tuned that to tighten the timings up, just ever so slightly. The 1680 v2 we're using real, just run of the mill core set 1600 megahertz memory. And of course, I've tuned that down a little bit from CL10 to CL8, and that of course, is one of the many benefits of X79, even in 2018. Anyway, there it is. (bleep) Anyway, guys I hope you enjoyed part one of three of the epic X79 showdown versus the latest and greatest in 2018 from AMD, from Intel. And I'm pretty shocked at how well this CPU did. It beat out the Ryzen 7 2700X with its really good memory, and its 4.2 gigahertz overclock and of course the 9900K that's using more power, so it's really crazy how far tech kind of hasn't come in five years. But anyway, love reading your thoughts and opinions as always, drop a comment in the comments section below. And also Steve if you're watching this, don't have the viewers believe otherwise, unless of course, you wanna get X79 for yourself and that 1680 v2 and run it at 4.6 gigahertz. And I'll catch you in the next one very soon. Peace out for now. Bye. (upbeat music)