Ryzen Versus Core i7 In 11 Popular Games

[captaincharisma]

gotta love all the AMD fanbois on here complaining that this website is bias

 

[COLGeek]

Now, if a tin-foil hat is needed, I am sure we can find one around here somewhere.

Would you mind? I want a fedora like Indy wears!!

Excellent! I would like one of those as well....just in case.

 

[Onus]

I mostly agree with Mitch and Sakkura, and do not agree with freak777power. While an Intel chip outperforms these new Ryzen offerings in most games, across the board, there is no objective measure by which Ryzen "sucks." Particularly as Sakkura points out, on the 60Hz monitors still extremely common, there will be little or no visible difference in actual use. Furthermore, as is appropriate for benchmarking, these are "clean" systems. I'd like to see what happens if streaming during gaming is tested, or fully loaded systems with lots of other processes (AV, firewall, driver updaters, Steam and/or Origin, vapor (cloud) services, etc) are running while playing.

 

[mitch074]

These Ryzens are very good all rounders. I looked at a YT review by Bitwit that showed that on 7 games tested the 1700 had an average of 118fps vs 127 from the 7700k (1080 testing). Other review sites showed the margin also within 10%, and the gap considerably closed at 1440/4k. Really not bad for an 8core chip, considering the current teething issues that keeps it a little back.
Here's hoping that upcoming DX12/Vulkan titles with better threading and core usage will run even better.

Vulkan and DX12 do not allow better use of multiple cores, at least not directly; instead they reduce the load on the display driver's thread, making better use of a single core's power. Where DX12 and Vulkan can help is if a graphics engine can efficiently make use of multiple rendering queues - but even then, that's trivial compared with improving a game's multithreading capability: threaded physics, threaded IA etc. using as many cores as there are instead of making use of 4 and stopping there...

 

[Onus]

Would be interesting to see, how High Performance plan affects idle/gaming power consumption.

Agreed. This seems a notable omission.

Now, if a tin-foil hat is needed, I am sure we can find one around here somewhere.

Would you mind? I want a fedora like Indy wears!!

Excellent! I would like one of those as well....just in case.

Sorry, mine's a tricorn, and I usually don't have it with me.

 

[Quora]

Title of thread is Ryzen vs core i7 in 11 popular games. They then choose 11 games I don't know of, and asked around and not a single person plays these games. How are they popular? Why not choose games people actually play? POE? Planet Coaster? World of Tanks or Warships? Hell, even WoW if you're into that sort of thing. It's a lot more popular than these games, most of which I've never even heard of before and never seen on Steam for sale even once.

 

[elbert]

Looks good for AMD Ryzen actually as the i5 represents the really test. SMT is disabled on both so its really the only quasi fair compare. Once games developers line out the SMT for Ryzen then it would be a fair compare.
Also I seen tests of Ryzen with 4 cores disabled that showed almost no loss in performance so Kaby lakes time is coming when they have to compete against a much cheaper CPU nipping at their heels. The Ryzen R5 X1500 4c/8t is sure to be a gaming marvel for the price. Its clock may even match the 7700K due to its not a cut down chip like the 6 core version.

 

[unclemao]

Near the end you put battlefield 4 instead of battlefield 1 when your comparing performance in the conclusion

 

[Guest]

People forgot that 6900k is not even being overclocked in this test and when you buy that CPU you set that thing to run at 4.2Ghz with no sweat (everyone does it). Despite having lower clock speed than Ryzen out of box, it still kicks its ass in gaming. That tells me enough. Now, you cannot really set Ryzen to run at 4.2Ghz not without enormous amount of voltage, lot of heat and bad ass cooling and even then there is a big If. x1800 is not even that cheap...$500 so when you add all up to me it does suck.

Maybe, i would have different opinion if AMD was a bit honest. I did pointed out at crippled Intel setup AMD did in their testing i got so many dislikes about. Now, there you go....

 

[mitch074]

I mostly agree with Mitch and Sakkura, and do not agree with freak777power. While an Intel chip outperforms these new Ryzen offerings in most games, across the board, there is no objective measure by which Ryzen "sucks." Particularly as Sakkura points out, on the 60Hz monitors still extremely common, there will be little or no visible difference in actual use. Furthermore, as is appropriate for benchmarking, these are "clean" systems. I'd like to see what happens if streaming during gaming is tested, or fully loaded systems with lots of other processes (AV, firewall, driver updaters, Steam and/or Origin, vapor (cloud) services, etc) are running while playing.

Current games make use of 2 heavy threads and 2 lighter ones - they are sized for i3 processors (and 'U' versions too). As such, i5 and up have more CPU power available for background tasks than they'd ever need. You might tax this kind of system more if you made a live stream of a gaming session using CPU-only video compression, but most people would make use of a hardware video compressor anyway (those are integrated in most GPUs nowadays, and not significantly taxing their gaming performances).

Now, AFAIK, Shadow of Mordor and DE:MD are the only games in that list that have been ported to both Mac and Linux on top of Windows and a console; this just might indicate that these games are better geared towards portability, and thus can retain the multicore capabilities they needed from consoles: XB1 and PS4 provide 6 or 7 inefficient, low-clocked cores to games (a couple are reserved for system use), that fast PC porters may simply hack together to run on 2 fast cores and 2 slow cores; better ports may just try to use as many cores as can be found.

 

[Guest]

At the end when 'game optimization' happens for AMD processor you will have newer and faster Intel CPUs meaning AMD Ryzen will never get performance edge. I put a big question mark on 'game optimization', i don't believe it will do a miracle.
I am just being honest here. If AMD Ryzen wanted to win they had to deliver processor which will be 20-30% faster than Kaby Lake in gaming but they delivered opposite. AMD also did cut the cost for their x1800 $500 CPU by having fewer PCIe lines, dual memory channel and what not. If there were to put more PCIe lines etc it would end up costing close to $1000. That stuff is not cheap.

 

[redgarl]

Please, you guys... again... cmon. Where are the 4k benches? You are still not representing a real world scenario. These chips are going to be used at 1440p - 2160p. Nobody in their right minds care about 1080p for a 500$ CPU, and that goes for a 1150$ CPU as well.

The 7700k is 40$ cheaper than a 1700x, however doesn't match Ryzen in anything except older games at 1080p and barely at 1440p. At 4k, there is no differences in gaming because the resolution is GPU bound anyway. You will only see the difference with 1080 TI in SLI, and even there, not in every games, and only in older titles.

That's real world use. Once again, without 4k benches, people cannot make a clear picture. If you were showing people that whatever CPU you are using at 4k, it doesn't make a difference with any GPU, than people would realize that Ryzen make way more sense even for a gamer.

Without presenting this info, you are manipulating data and make this an non-objective article.

 

[Thelatestnewbgamer]

so a first generation ryzen 14mm cpu can go toe to toe with a 4th generation intel 14mm...seems like amd is a tremendous value for a third of the price you get the same intel gaming performance at 1440p and beyond..ryzen even beats a 4ghz intel refined kabylake clocked cpu with less cores..too...if people think ryzen is not a tremendous value for gaming a workstation and encoding blu-rays they better take a damn BC and get a reality check

 

[dstarr3]

In short, these CPUs are competing with Xeon CPUs, which they compete very well. But if all you do is game, the value is either equal or worse than Intel.

 

[Joe Black]

As I was perusing the benchmarks one thing became clearer and clearer... You never know which Intel CPU is going to come out tops. And its clear that not one of them is the top dog, or the underdog across all games. Which is confusing considering the pricing. As for the AMD CPUs... It's a nice linear progression: 1700 < 1700x < 1800x almost each and every time.

No bs. All round decent CPUs for a decent price. We already know they are well rounded for non-gaming workloads as well so that helps a lot.

Right?..

 

[cwolf78]

"Theories abound as to why Ryzen processors are struggling in gaming metrics"

I big one I don't see you mention is there is a bug with the Windows 10 thread scheduler and it does not correctly identify physical cores from "virtual" ones with Ryzen's SMT implentation. Nor does it correctly identify how much L1 and L2 cache it can utilize. AMD and MS are actively working on a fix, but until then I don't see the point of releasing more benchmarks. Windows 7 does not have this bug, so testing on that platform would be a far better measure of the chip's actual performance. (I don't have a dog in this fight, but I hate to see an unfair comparison regardless of the outcome.)

 

[redgarl]

Intel does not pay for our reviews. Thank you.

nice defense XD

Not a defense. Rather, it is a fact. Have a good day.

So basically, you are just saying you are incompetents...?

 

[COLGeek]

Intel does not pay for our reviews. Thank you.

nice defense XD

Not a defense. Rather, it is a fact. Have a good day.

So basically, you are just saying you are incompetents...?

In what regard? Please explain.

 

[cwolf78]

"Theories abound as to why Ryzen processors are struggling in gaming metrics"

Well there's more than a theory, there's an actual fact. The Windows 10 thread scheduler does not work properly with Ryzen. It identifies all 16 threads as a physical cores complete with their own dedicated L1 and L2 cache. This is obviously incorrect and adversely effects performance which is why many reviewers disable SMT on the processor during gaming benchmarks. Microsoft and AMD are actively working on a fix. Until then, I don't see the point of releasing more benchmarks for a crippled processor. I think you should either wait for the upcoming patch or test using Windows 7 which does not have this bug and utilizes Ryzen properly provided the correct drivers are installed. (Note, I do not have a dog in this fight, but I hate to see skewed results from an unfair comparison regardless of the outcome.)

(I just looked at this again and AMD apparently doesn't think there is a problem but Microsoft has a statement indicating that there is something wrong and are working on it. Might be worth investigating.)

 

[firefoxx04]

When intel wins, the frames are already plenty high and the margin is really enough to discount the fact that AMD has 8c16t for when you do real workloads.

Gaming at 100+fps is nice and all but once I am over 60fps I really stop caring. AMD has a winner, even if they by 5-20fps in a few games at LOW resolution. The 8 core chips are much more well rounded than the i7 and way better than the i5, and that is without considering the PRICE.

 

[mac_angel]

is Metro Last Night Redux DX12?

 

[ravewulf]

"Many of us use our CPUs for several years, and as we swap out for faster graphics cards, the bottleneck will start swinging back to host processing. In many ways, today’s 4K is tomorrow's QHD."

There's a fault in that thinking. Games improve through better software optimization and taking advantage of more threads, so low resolution gaming is not necessarily a good predictor for how future games will perform.

AdoredTV on YouTube did a comparison between low resolution reviews of the 2500K and FX 8350 from 2012 to present and found that the FX 8350 went from being behind to being ahead as game software improved.

Could you look into it and do your own comparison of how software improvements have affected games? This could potentially include driver improvements in addition to the game software itself. It would be really nice to have some hard data on this instead of assuming that low resolution gaming is a good indicator that scales up to future games.

 

[Brian_R170]

I guess this article is meant to be the defense for the current "Best Gaming CPUs for the Money". You can't argue with it for anyone building a system whose primary purpose is gaming. No excuses, the fact remains that Ryzen 8C/16T didn't cause a disruption in recommended gaming CPUs that benefit gaming consumers. So, now we just have to wait to answer the question of how will the 4C/8T Ryzen change the recommendations and how far up the performance ladder will it compete?

 

[mitch074]

is Metro Last Night Redux DX12?

No - it's a DX10 game with DX11 compatibility. It predates DX12 by quite a lot and while the main developer is very happy about DX12 (and probably Vulkan by now), he did mention that game engines need to be written for those newer APIs to really make use of them. As a matter of fact, DX12/Vulkan bring very little performance improvements in current games, except those that got a non-trivial rewrite to support them better (Doom comes to mind with async compute, but even then id's developers mentioned that there's still untapped performance to be gained - async was a low hanging fruit for shadow computations).

As Croteam detailed when developing Vulkan support for The Talos Principle, it's perfectly possible to port a current OpenGL 4.5/DX11 engine to Vulkan/DX12, and with some work to get a noticeable boost, but real performance improvements would only come with a ground-up rewrite (to make better use of resources management, multiple queues, etc.)

 

[filippi]

Title of thread is Ryzen vs core i7 in 11 popular games. They then choose 11 games I don't know of, and asked around and not a single person plays these games. How are they popular? Why not choose games people actually play? POE? Planet Coaster? World of Tanks or Warships? Hell, even WoW if you're into that sort of thing. It's a lot more popular than these games, most of which I've never even heard of before and never seen on Steam for sale even once.

So instead of Battlefield 1, GTA V or Middle-earth: Shadow of Mordor they should test Planet Coaster and POE?
Man now i feel kinda oblivious of the gaming scene:
Never heard of Planet Coaster.
And i had to google poe. I though it was the writer...