AMD Ryzen 7 1800X vs. Intel Core i7 7700K Linux Gaming Performance

[indepe]

This maybe your opensource radv vulkan performance fix: Solution--reconfigure kernel: "make menuconfig"/Bus options/PCI support/PCI Express Port Bus support/PCI Express ASPM control/Default ASPM policy (set to "Performance")

Is it possible to translate this into english? Do I need to worry about running into the same problem? (Once I have a system able to run Vulkan without recompiling the kernel, that is, which will be soon.)

[Dark_Arc]

Hey Michael, any chance we could see a Xonotic benchmark for the Ryzen 1800X?

[charlie]

Phoronix: AMD Ryzen 7 1800X vs. Intel Core i7 7700K Linux Gaming Performance
http://www.phoronix.com/vr.php?view=24225

This maybe your opensource radv vulkan performance fix: Solution--reconfigure kernel: "make menuconfig"/Bus options/PCI support/PCI Express Port Bus support/PCI Express ASPM control/Default ASPM policy (set to "Performance")

[A Laggy Grunt]

If you're looking to find out how the distribution of threads across real and virtual cores affects the CPU's performance, you can mess with process affinity too. I used schedtool when trying to do that.

[pal666]

Not in gaming tests it doesn't. Ryzen also falls behind in memory bandwidth performance and has a lot less PCIe lanes compared to 6900k.

in gaming tests it has same speed for $600 less. it is utterly destroys in my book

[torsionbar28]

I think it is accurate, although it's nothing to do with AMD's manpower. Think about it, the heavy corporate Linux contributors are all server vendors. They don't give a crap about Linux on the desktop (aside from professional workstations which also use server chips), and they certainly don't give a crap about Linux desktop gaming. I'm talking about IBM, HP, Red Hat, SuSe, etc. Unfortunately for us desktop users, we may have to wait for the Zen server chips to get closer to market, before we see the big players spin up and contribute code.

To add to this point, another thing to consider is that Zen is a *major* architectural overhaul for AMD. When the Zen server chips hit the market, they have to be right the first time, and they have to be well-tuned on day 1. Getting these Ryzen desktop CPU's into consumers hands first, gives AMD some time to collect performance data and feedback and fine tune the microcode before the server chip release.

After all, the consumer market is far more tolerant to firmware updates than the enterprise space. See the firmware debacle that many consumer SSD's have, where the vendor puts out one after the other to get the thing working correctly. That simply doesn't fly in the datacenter. So just like intel does, AMD will likely have some microcode updates released in the next 6 months to a year that fine tune Ryzen. Just as in the past, the microcode updates will come in two ways - as a BIOS update for the mobo, and as an OS "patch" that applies the update at boot-time if the mobo BIOS wasn't updated.

[Holograph]

Michael needs benchmarking advice ASAP. Those games don't use 8 cores, so you need to go Hz vs the same Hz.

People don't change what software they use based on their CPU purchase.

Perfectly reasonable to benchmark any processor with any program that buyers of that processor might run.

And this is a consumer platform, so any consumer application that people would be likely to use is fair game.

Not that highly multithreaded applications and games shouldn't be tested, but it's ridiculous to imply that non-optimal workloads aren't okay to bench. Must every piece of hardware then only be benchmarked in applications it excels in? Then we wouldn't have standardized benchmark results.

[Michael]

Michael needs benchmarking advice ASAP. Those games don't use 8 cores, so you need to go Hz vs the same Hz.

My focus is on testing the out-of-the-box performance / what users will encounter. I didn't title this article a comparison of Intel/AMD CPU architecture effectiveness.

[artivision]

Michael needs benchmarking advice ASAP. Those games don't use 8 cores, so you need to go Hz vs the same Hz.

[Luke_Wolf]

That screenshot doesn't show the display resolution and graphics settings of the benchmarks. In benchmarks with 4K resolution, 1440p resolution, or even some of the benchmarks with 1080p resolution but the graphics settings on the highest levels, the difference between Ryzen and the recent Core i7s is tiny or even zero.

But that's because the GPU is doing all of the work. To compare the processors, the benchmarks should be 1920x1080 with low graphics settings or even lower resolutions. And in those, from what I've seen disabling SMT gives a 3-5% performance boost to Ryzen but not enough to catch Intel.

You can't compare processors with games running at low settings and resolution. Stop that myth. Games are far too complex of beasts to really say what exactly is causing performance differences when you do that which makes the results meaningless. You can get away with testing GPUs with games because you can hold everything else the same (Which you can't do with CPUs across vendors) and see how much load it'll take while still meeting target display frequencies, however you still can't take one or a small handful of games and objectively say GPU A is better than GPU B, there's countless games that show vendor preference one way or the other. Games are perfectly valid real world tests (When run as real world tests at proper resolution and settings), but do not.... absolutely do not delude yourself into thinking that you can make them valid synthetic benchmarks which is what doing that nonsense is trying to do.

I'm an AMD fanboy. I plan to get an R7 1700 or R5 later this year. But it looks like the Core i7-7700 is still a better buy for gaming enthusiasts. No matter how much I wish differently, that won't change the reality.

Currently higher clocked quad cores are better for gaming enthusiasts than octocores, doesn't matter if Intel or AMD octocores, the quad cores still win out. HOWEVER... that is the state of today, and AMD generally is beating out or at the same level as its octocore Intel counterpart. With DX12 and Vulkan in play it's entirely possible that that quad cores being better to octocores is going to flip in a year or two. Further keep in mind that AMD is launching it's quad core later this year, and if they clock it high enough (given its IPC is in the same ballpark) it should be close to the i7 7770K at a much lower price.