Core i7 8700K vs. Ryzen 7 2700X With Rise of The Tomb Raider On Linux

[ramrod]

Putting like a 1080Ti there isn't there to benchmark 1080Ti but to show differences between benchmarked CPUs. What you see in this benchmark are results capped by weak GPU for which it doesn't matter if it's a mid or high end CPU. With RX580 it could turn out even that Core i3 is as good as i7 8700k and so on...

It still sounds useless to me. I don't care how a CPU performs when paired with a too expensive GPU, using settings that look awful. It's a scenario that will never apply to me. If you're going to do a CPU bench do something that might actually apply to the majority of people interested in buying the CPU.
If a RX580 performs the same with both an i3 and an i7 then that's great. If you're buying an RX580 or less theres no longer a reason to buy an i7 if gaming is your concern. Otherwise if you care about more than games look at other productivity benches.

[riklaunim]

It still sounds useless to me. I don't care how a CPU performs when paired with a too expensive GPU, using settings that look awful. It's a scenario that will never apply to me. If you're going to do a CPU bench do something that might actually apply to the majority of people interested in buying the CPU.
If a RX580 performs the same with both an i3 and an i7 then that's great. If you're buying an RX580 or less theres no longer a reason to buy an i7 if gaming is your concern.

That's good and valid point but it's not the title of the benchmark. Current title is misleading and/or click-bait.

[dungeon]

Nope, even when using Nvidia who have the fastest Linux Vulkan driver it's still 15-20% slower than the windows dx11 and dx12 version. Not a great advertisement for Vulkan...

It might not be that VK makes all the difference, could be something else. See what anandtech found recently what HPET do in Windows, for RoTR especially - 76% difference @ 1080p and especially on Intel CPU On AMD CPU happens also but not to this extent, up to 15% there... but on Intel it is a mess



https://www.anandtech.com/show/12678...yzen-results/4

[theriddick]

vega64 > gtx 1080

Not quite, most the time the 1080 will prevail. I'm sure the Vega64 will eventually beat the 1080, one day. But no ITX and high watt usage kinda broke the deal for me. If it was a 210-250 watt card like the vega56, I'd probably have got one. (and if it had ITX)

Source: http://www.madshrimps.be/articles/ar...eo-Card-Tests/

[carewolf]

There is a reason why many CPU comparisons is done with the most powerful GPUs and/or games running on resolutions even smaller than 1080p.

Because the people writing the comparisons have no idea what they are doing, so they are testing something tied to CPU but completely pointless?

[pal666]

Let's face it we do have good drivers now on Linux (both AMD and Nvidia).

well, good drivers on linux is a prerequisite for windows being slower, isn't it?

[ramrod]

That's good and valid point but it's not the title of the benchmark. Current title is misleading and/or click-bait.

I didn't realise we were discussing the title. I thought we were discussing your complaint

There is a reason why many CPU comparisons is done with the most powerful GPUs and/or games running on resolutions even smaller than 1080p.

That said the title dosen't look misleading/click-baity to me...

[riklaunim]

Because the people writing the comparisons have no idea what they are doing, so they are testing something tied to CPU but completely pointless?

If you want to compare CPUs you have to put a lot of load on them. Small resolution and very strong GPU is used to be sure that GPU isn't a limiting factor. It would be odd to publish benchmarks that 8 core CPU is as good as 2 or 4 core... It's not something you would do to play but it's used to showcase CPU efficiency with given game. Normal setup (high resolution, high settings) benchmarks are also done but they aren't represented as CPU tests if the results are obviously not capped by CPU.

[cbxbiker61]

There is a reason why many CPU comparisons is done with the most powerful GPUs and/or games running on resolutions even smaller than 1080p.

It's been demonstrated that low-res/1080P comparisons do NOT give a proper representation of what is going to happen at higher res. But reviewers (and the people that watch the reviews) are slow to learn...

With low res absolute single core performance may be the bottleneck of the CPU....at higher res the bottleneck of the CPU may move to the CPU's capability to handle lots of data (which the Ryzen is pretty good at).

In a nutshell...if you're going to game at 1440P/4K, use the benchmark results at 1440P/4K and ignore those idiots that think low-res results somehow mean somthing to you.

[carewolf]

If you want to compare CPUs you have to put a lot of load on them. Small resolution and very strong GPU is used to be sure that GPU isn't a limiting factor. It would be odd to publish benchmarks that 8 core CPU is as good as 2 or 4 core... It's not something you would do to play but it's used to showcase CPU efficiency with given game. Normal setup (high resolution, high settings) benchmarks are also done but they aren't represented as CPU tests if the results are obviously not capped by CPU.

The problem is you are still testing something pointless. The fact is the CPU is fast enough to power the GPU, and thus any CPU power above that is completely pointless to that game, so they are testing something pointless. To test CPU power they should instead be testing something that isn't merely bound to CPU but where the extra CPU power actually does something usefull. The whole thing smell of trying to sell more powerful CPUs to gamers that doesn't need them.