The Definitive AMD Ryzen 7 Real-World Gaming Guide @ [H]

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
55,534
The Definitive AMD Ryzen 7 Real-World Gaming Guide - With our AMD Ryzen 7 overclocked to 4GHz we find out if this is a competitive real-world gaming CPU or not. We compare it with two overclocked Intel 7700K and 2600K systems across six different video card configurations at 4K, 1440p, and 1080p to find out which CPU provides the best gameplay experience using playable game settings.
 
Very, very impressive considering where the clock speeds fall.

The biggest surprise was how well the dusty 'ole Sandy Bridge hangs with the newborns. Due to this, I really have no reason to upgrade my IB for the sake of gaming until possibly Threadripper or Coffee Lake...and that's a big maybe.

Thanks for another great review, [H]!
 
Awesome article! It seems the crossfire penalty with the 1700x was often greater at 1440p than it is at 1080p. I did not expect that.

I would have to respectively disagree that clock speed matters more than IPC, at least from evidence here. Ryzen and SB have similar IPC, yet ryzen usually hangs with SB despite being 500 mhz slower. There is also the example of how 1440p crossfire often took a bigger penalty than the 1080p in CF with Ryzen. If it was a mhz issue, the opposite would be true.

Again, fascinating article.
 
Impressive! Brent is really good at matching up the runs - extraordinary.

The conclusions and thoughts are right on as well. Twitch gamers, high frequency monitors mostly gaming workloads 7700K, a more mixed workload with gaming Ryzen can be very compelling and a good solution. Great having competition back indeed!
 
So, the main takeaway is that you are not supposed to run mGPU with Ryzen?

Because frankly that's what was actually the most surprising thing for me... even if it should not have been in hindsight.
 
Nice review, although I feel it's a bit premature to call it "definitive" and "once and for all" since as you state it's only representative for now.
Also it could be interesting to see how much difference a Ryzen 5 with two cores (and $100) less compares at the same clock speed, since IPC is the same (or marginally better) than that for Ryzen 7.
 
I'm wondering what the results would be on systems that have been running for several months and have a bunch of things installed and running. My guess is that the Ryzen would be faster due to having the extra cores running things in the background.
 
Very nice work Brent & Kyle.

We can all look forward to the new CPUs from both AMD and Intel. Can't wait to see what 10+ cores bring us.
 
first off i want to say i hate both of you, i ended up staying awake an extra hour reading this article because i accidentally refreshed the front page instead of clicking my gmail bookmark to check my email before going to bed..

thats out of the way now, i really enjoyed the article and it answered a lot of questions people had comparing all 3 processors including whether or not the nvidia drivers were holding back performance on ryzen chips.. thanks for all the hard work you did with the testing Brent.


I'm wondering what the results would be on systems that have been running for several months and have a bunch of things installed and running. My guess is that the Ryzen would be faster due to having the extra cores running things in the background.

to many variables but i'm sure we'll find out eventually from some random users on the forum, lol.

Nice review, although I feel it's a bit premature to call it "definitive" and "once and for all" since as you state it's only representative for now.
Also it could be interesting to see how much difference a Ryzen 5 with two cores (and $100) less compares at the same clock speed, since IPC is the same (or marginally better) than that for Ryzen 7.

as the article says we live in the now and not what might happen 3 years down the road and quite frankly there's nothing to show game developers including better thread awareness in their games any time soon. even this whole i9 and threadripper/epyc stuff while cool and interesting isn't going to change that. by the time this article loses it's relevance as a comparison we'll more than likely have next gen ryzen and a kaby lake replacement available. but it pretty much answered the question everyone had, GPU limited all 3 processors are pretty much identical, CPU limited the 7700k is better which is what we all knew and this article definitively proves that.

i'd say the only thing that interests me further is whether or not SLI shows the same issues cfx showed on ryzen but that's a whole different thing.
 
Last edited:
I'm wondering what the results would be on systems that have been running for several months and have a bunch of things installed and running. My guess is that the Ryzen would be faster due to having the extra cores running things in the background.
Thanks, I was wondering who would be the first one to mention that. I started to talk about it directly in the article, but decided that talking up BS rumors was not part of what we wanted to cover.

i really enjoyed the article and it answered a lot of questions people had comparing all 3 processors including whether or not the nvidia drivers were holding back performance on ryzen chips.. thanks for all the hard work you did with the testing Brent.
Uh.....not sure, and we will not be testing that. We run these system very "clean and mean" to assure us the fewest variables possible impacting performance.
 
Great review boys. I can only imagine how long this took, considering all the different configurations you used.
This definitely answers a lot of questions. Now AMD just needs to get the clock speed up in these things.
 
Great review boys. I can only imagine how long this took, considering all the different configurations you used.
This definitely answers a lot of questions. Now AMD just needs to get the clock speed up in these things.
I would guess it represents around 400 in-game runthroughs. This is the "biggest" one of these we have ever done.
 
Great job!

Ryzen might be able to eke out a couple more percent now that the firmware is starting to mature. DDR4 3200+ is getting more realistic at the least with the new AGESA code
 
Very nice article. The conclusions weren't unexpected, but you provided a lot of data to back up the point that the Ryzen series is currently the strongest general use processor value for anyone that uses their workstation for more than purely gaming but still expects excellent gaming performance when they do.
 
Amazing article. One that I think is going to reverberate around the pc gaming world.

My main takeaway is the amazing power and longevity of the sandybridge chips.
 
Amazing article. One that I think is going to reverberate around the pc gaming world.

My main takeaway is the amazing power and longevity of the sandybridge chips.
I still have a 2500k sitting in a box. I can't find a replacement motherboard to replace the ASUS that died in the line of duty, unless I want to pay a ridiculous price for a used one on EBay. After reading this article I may need to start the search again.
 
I really applaud AMD for what they did with Ryzen. They brought a very competitive cpu to the market with fantastic pricing. They have brought back the cpu war again and us the consumer is going to reap the benefits. I cant wait to see how things turn out in the next few years. Ryzen is a huge win!
 
Great job!

Ryzen might be able to eke out a couple more percent now that the firmware is starting to mature. DDR4 3200+ is getting more realistic at the least with the new AGESA code
We waited around as long as we could for AMD to get its AGESA code "fixed" before starting testing. That said, I doubt and extra 267MHz on memory is going to amount to much in our testing. I have done some comparisons here where our testing is fully CPU limited, and I have not seen a solid "couple percent" in that, much less where we get to gaming being greatly GPU-limited.
 
Great read. Kyle, there is one interesting Ryzen myth still persisting that I wouldn't mind you guys taking a look at - and it is counter-intuitive to the "push the boundaries of FPS" crowd.

No one is arguing the 7700 pushes higher FPS than Ryzen - however there have been some vocal people (not sure if it has been a majority or minority) that has been saying the 7700 also has lower dips too, and the occasional "stutter" when gaming.

Is there a way to test the average FPS floor for the 7700 vs Ryzen? Or even FPS variance? It's more along the line of what you said when you said if you didn't know what processor was what in a blind test, you couldn't tell. Some people are claiming the Ryzen processor doesn't get "stutters" because the FPS never dips as low as the 7700, even though the 7700 can also hit higher highs than Ryzen.

Think this may be something worth investigating? Just know that nugget is still being perpetuated out there.
 
Thanks for the article that calms down the itch of upgrading from 6700k @ 4.6 that I'm pushing right now. No need to even look at anything else at the moment.
 
Great read. Kyle, there is one interesting Ryzen myth still persisting that I wouldn't mind you guys taking a look at - and it is counter-intuitive to the "push the boundaries of FPS" crowd.

No one is arguing the 7700 pushes higher FPS than Ryzen - however there have been some vocal people (not sure if it has been a majority or minority) that has been saying the 7700 also has lower dips too, and the occasional "stutter" when gaming.

Is there a way to test the average FPS floor for the 7700 vs Ryzen? Or even FPS variance? It's more along the line of what you said when you said if you didn't know what processor was what in a blind test, you couldn't tell. Some people are claiming the Ryzen processor doesn't get "stutters" because the FPS never dips as low as the 7700, even though the 7700 can also hit higher highs than Ryzen.

Think this may be something worth investigating? Just know that nugget is still being perpetuated out there.
I think if you look at the graphs, that information is already there if you require it.
 
What a crazy amount of work. Kudos.

No major surprises, but it does kill the speculation that Ryzen performs better with AMD cards (and or crossfire).

Other than that, it looks much the same as before. For the very high framerate twitch gamers, building a rig mainly for gaming, Intel with higher IPC and clockspeed is their choice.

The vast majority of us would never be able to tell the difference while gaming, so the Ryzen is a fine choice gaming and other workloads.
 
Thanks for doing all the hard work, came out about where I expected it. Was a bit surprised about crossfire and the penalty it took, perhaps a driver issue.
 
Wow. Every stone unturned. I feel dirty, like I got to see something I wasn't supposed to and I like it. Awesome review.

Intel should be ashamed. Seeing those 2 lines hug each other across that many generations of 'improvements' is shameful.

Hopefully Intel will send you a review pack with all those new i9's to evaluate and torture, although you mentioned you would probably have to source your own. I can wish though!!
 
Wow. Every stone unturned. I feel dirty, like I got to see something I wasn't supposed to and I like it. Awesome review.

Intel should be ashamed. Seeing those 2 lines hug each other across that many generations of 'improvements' is shameful.

Hopefully Intel will send you a review pack with all those new i9's to evaluate and torture, although you mentioned you would probably have to source your own. I can wish though!!

And AMD should be not? Catching up on Sandy in terms of gaming after all these years is not an achievement either.
 
And AMD should be not? Catching up on Sandy in terms of gaming after all these years is not an achievement either.
If you consider the R&D budgets available, and the personnel Intel has working for them, and the legions of fanboys deriding them at every turn.... you'd realize what a huge achievement it actually is.

Intel is now offering 12, 16, and 18 core i9 processors. This would never have happened without Ryzen.
 
And AMD should be not? Catching up on Sandy in terms of gaming after all these years is not an achievement either.


i think it says more about the current state of games then it does about the processors.. once you put workstation/multimedia usage into the picture theres no contest between sandy bridge and kaby lake/ryzen.
 
Kyle, thanks for the well-done article.

There are a couple of typos in the conclusion. I think you meant to say "And quite frankly we did expect that, since none of today's top games are tremendously thread-aware".
Predict vs predicate.

The Core Question

The big question here is, "How does the AMD Ryzen 7 8C/16T processor stack up against Intel's newest 7700K and 6 year old 2600K 4C/8T processors for the enthusiast?"

We experienced no performance improvements or gameplay differences because Ryzen 7 had twice the cores and threads. It seems to have done nothing, in terms of gaming performance in ten games, six video card configurations, and three different resolutions of gaming. And quite frankly we did not expect that at all since none of today's top games are tremendously thread-aware.

What did make the most difference was MHz and IPC. According to our findings it is still true that MHz rules when it comes to gaming performance. Better IPC, and faster MHz makes the most difference when it comes to improving your gaming experiences across resolutions and video cards. The Intel i7 7700K dominated performance in many instances because of its 5GHz clock speed and better single-core IPC. As it stands, that is what affects games the most, currently.

Will it always be that way? Maybe not. Maybe in the future CPUs with more cores/threads will be superior because games will be using multi-core better. We cannot predicate that of course, nor can we live in a fantasized future of which we know not the outcome. What we can do is live in the now, and tell you how it is right here, in the now. In the now, games prefer MHz and IPC, that’s just the way it is. Only time will tell if that trend continues, or busts.
 
After I read all your different tests, and how they were all "clean installs" I assumed you guys have a LOT of patience.

It's a great read and the numbers really aren't that different, and the gaming @ 1080p or 1440 aren't even HUGE drops, at least, it doesn't render them unplayable. The gaming FPS you give up makes up in, as you've both said, other uses for the CPU. If you're someone who plays games every so often but creates content most of the time, it seems the Ryzen is the chip for you!

Do you think there'll be more improvements to the current generation?
Or do you think we'll have to wait for AMD's next generation?
 
I think this article just proves to us that Games developers are still years away from having games which can take advantage of high core count machines.

Kinda sad that the best option for gaming is a sub 400$ cpu, when much more expensive cpus exist. We will make it there one day, hopefully soon seeing as Intel and AMD are releasing HEDT platforms with more than 8 cores. I already own a 8 core/16 thread cpu so I's ready :)
 
Time to upgrade the 2500k.... to a 2600k, 2700k, 3770k, or whatever I can find that the MOBO will take. Seems like I won't be missing out much vs a totally new rig.

Especially since it'll be GPU 1st to be replaced.
 
I think this article just proves to us that Games developers are still years away from having games which can take advantage of high core count machines.

Kinda sad that the best option for gaming is a sub 400$ cpu, when much more expensive cpus exist. We will make it there one day, hopefully soon seeing as Intel and AMD are releasing HEDT platforms with more than 8 cores. I already own a 8 core/16 thread cpu so I's ready :)

And why is that sad? Gaming is an entertainment activity that should be available for the masses, unlike content creation and data processing which is a professional activity. If your game requires $1000 cpu - you're in trouble. Sub $400 is already too much for a gaming machine cpu and should be even lower.
 
I kind of figured all the naysayers were wrong about gaming on Ryzen. The vast majority of gamers wouldn't be able to tell the difference at 1440p and up in a blind taste test.

Not related to gaming, but perhaps you can speculate:

Given what we saw with crossifre performance, do you think Ryzen would be a solid choice for a workstation dedicated to multi gpu deep learning research?

In other words, is the fuckus with crossfire just a byproduct of game drivers, or do you think something else hobbles the system to make it unsuitable for running more than one gpu, for any task?
 
  • Like
Reactions: n=1
like this
I'm not sure I'd spend the money Chris. You'd have to pay in excess of $100 to just do a CPU swap that would get you 5-10 fps at low res gaming.
 
And why is that sad? Gaming is an entertainment activity that should be available for the masses, unlike content creation and data processing which is a professional activity. If your game requires $1000 cpu - you're in trouble. Sub $400 is already too much for a gaming machine cpu and should be even lower.

2500K and Z77 can be had for sub 200, my son games quite nicely on such a system.
 
It really is sad that the 2600k still holds its own, we all knew it would, but really hoped it wouldn't. Only can hope the trend doesn't hold for another 6 years.

The only other take away I really see for this is that GPU is more important than CPU for gaming, which really shouldn't be a surprise to anyone
 
I'm still sitting on the sidelines for my next build....but this is an awesome amount of work here.

My AMD homer side REALLY wants a Ryzen build but as literally all I do on a home computer is gaming I'm torn. I don't do anything that would take advantage of the extra threads if gaming isn't.

How far off are we from 4/8 being a limitation? My assumption is that Ryzen will "age well" similar to how low-IPC Bulldozer chips can sneak up on i3s that used to crush them in gaming.

I would still also LOVE to see a single CCX Ryzen 4/8 chip that doesn't have to deal with the "Infinity Fabric"...but part of me is wondering if there isn't much to be gained there hence why they have't released it.
 
Back
Top