Does Ryzen Perform Better with AMD GPUs?

[Julio Franco]

Since our initial review we've been looking at Ryzen from a few different angles. But there's a rumor going around that Ryzen's gaming performance is better than we think... if you use a Radeon GPU. Curious to see if there is any truth to the story, we put together a test designed to eliminate GPU bottlenecks and see what happens.

Read the full article here.

https://www.techspot.com/article/1374-amd-ryzen-with-amd-gpu/

 

[ScubaRhys]

I don't really understand why you downclocked the Intel chip. It makes it an unbalanced comparison if you ask me. I would have put them both at stock or both overclocked to a level an EU might achieve. Downclocking one and overclocking the other is odd to me: If this was the other way round and you downclocked the AMD chip the fanboys would be out for your blood!

 

[Steve]

I don't really understand why you downclocked the Intel chip. It makes it an unbalanced comparison if you ask me. I would have put them both at stock or both overclocked to a level an EU might achieve. Downclocking one and overclocking the other is odd to me: If this was the other way round and you downclocked the AMD chip the fanboys would be out for your blood!

"We configured both the 7700K and 1800X to run at 4.0GHz but you can ignore Kaby Lake's downclock because I'm not trying to show which processor is faster. "

It's not a CPU comparison!!!

 

[VersPrime]

Somewhat random thought I had: It could be interesting to note both CPU and GPU utilization during each run to monitor for bottleneck scenarios and possibly shed light on that Ghost Recon performance. Certainly not something I'd expect you to repeat all that work for, but the added data could be useful in targeted situations when something doesn't seem right?

 

[ScubaRhys]

"We configured both the 7700K and 1800X to run at 4.0GHz but you can ignore Kaby Lake's downclock because I'm not trying to show which processor is faster. "

It's not a CPU comparison!!!

Sure, it's not a CPU comparison. I do think it was an interesting test and that the results lead to an accurate conclusion. I'm just curious as to what reasoning there is to match the clock speeds of the chips when comparing driver performance, I could see why you might match the single core performance or rather the gaming performance to eliminate any performance differences between the CPUs allowing us to concentrate on driver differences. But is a 7700ks gaming performance the same as Ryzens at 4ghz?

I'm not complaining, actually I would find it very interesting to see how a Ryzen CPU with 4 cores turned off and SMT enabled compares to a 7700k at the same clocks.

 

[anselhelm]

Hi. I think you misunderstood what people are saying. It's not a rumour either: it's an assertion that Nvidia's DX12 driver is poor and isn't as effective as AMD's DX12 driver. For Intel CPUs with up to 4, higher-clocked cores, this is less of an issue, but with AMD Ryzen CPUs with more, slower-clocked cores, the issue becomes apparent. I'm trying to remember where I saw a test comparison and analysis of this. It might have been AdoredTV. The bloke was Scottish, if this helps.

 

[Steve]

Hi. I think you misunderstood what people are saying. It's not a rumour either: it's an assertion that Nvidia's DX12 driver is poor and isn't as effective as AMD's DX12 driver. For Intel CPUs with up to 4, higher-clocked cores, this is less of an issue, but with AMD Ryzen CPUs with more, slower-clocked cores, the issue becomes apparent. I'm trying to remember where I saw a test comparison and analysis of this. It might have been AdoredTV. The bloke was Scottish, if this helps.

Did you read the article? Did you look at the results? Yes, there is an issue on Ryzen with Nvidia's DX12 performance, we also found that Ryzen worked better in quite a few games using Nvidia on DX11 though.

 

[dirtyferret]

Steve
While everyone appreciates your efforts in showing how Ryzen 1800x performs with both Nvidia and AMD video cards in various games, we believe you failed bigly. Please re-run all the tests to show in fact Ryzen is being handicapped by every bit of software and/or hardware in the PC community (if possible please prove or confirm the conspiracy against AMD) and that the Ryzen CPU is in fact better then Intel's CPU offering even if the test has nothing to do with CPU performance! If at any point your test shows Intel/Nvidia superiority against AMD, re-run the test until it shows otherwise and make no mention of such results unless in reference to the AMD conspiracy. Thank you.

- The entire AMD fan boy community

 

[Daithi]

Looking forward to everyone disagreeing with hard numbers again, because some obscure no name blogger did some esoteric tests that confirmed their biases and therefore everyone else must have done something wrong.

 

[cliffordcooley]

If at any point your test shows Intel/Nvidia superiority against AMD, re-run the test until it shows otherwise

- The entire AMD fan boy community

lol

 

[ScubaRhys]

Steve
While everyone appreciates your efforts in showing how Ryzen 1800x performs with both Nvidia and AMD video cards in various games, we believe you failed bigly. Please re-run all the tests to show in fact Ryzen is being handicapped by every bit of software and/or hardware in the PC community (if possible please prove or confirm the conspiracy against AMD) and that the Ryzen CPU is fact better then Intel's CPU offering even if the test has nothing to do with CPU performance! If at any point your test shows Intel/Nvidia superiority against AMD, re-run the test until it shows otherwise and make no mention of such results unless in reference to the AMD conspiracy. Thank you.

- The entire AMD fan boy community

Haha!

What I don't understand is that a lot of people seem to be making a lot of effort to try and paint the picture that any reason AMD performs worse is because of another company's doing. Why does this matter? At the end of the day it's the performance the end user is getting if they buy that product, if you arent happy with the performance I don't see how it is of any comfort that its not the manufacturers fault!

But then that's what makes a fanboy a fanboy I guess!

 

[pan1c]

I would like to see a more in-depth memory clock benchmark, I've seen lots of youtube reviews which display a dramatic increase of performance.

I am still thinking what I should buy, 7700K or Ryzen 1700 -.- ...

 

[Kenrick]

Thanks Steve. Surely the AMD fanboi community won't approve this article. You have to find all text "gtx1060" in this article and replace them with "gtx1080 superduperTI". In the eyes of the AMD fanbois, the RX480 is equivalent to GTX1080.


Ryzen 5 should be what the community is talking about.

 

[Adi6293]

I would like to see a more in-depth memory clock benchmark, I've seen lots of youtube reviews which display a dramatic increase of performance.

I am still thinking what I should buy, 7700K or Ryzen 1700 -.- ...

Well both CPU's will play any game at max settings so unless you have 100+Hz monitor I would buy Ryzen 1700 and OC it to 4.0Ghz, it should last longer

 

[mosu]

@ anselhelm This is the link you were looking for. link:[https://www.youtube.com/watch?v=0tfTZjugDeg]. It's about Nvidia poor or maybe incorrect implementation of DX12 drivers.

 

[ScubaRhys]

Well both CPU's will play any game at max settings so unless you have 100+Hz monitor I would buy Ryzen 1700 and OC it to 4.0Ghz, it should last longer

Should it last longer? That might not be the case. Games are clearly moving to multithreaded architecture but how quickly? Ryzen has a lower IPC than Kaby Lake which will be exposed more on the future as graphics cards get stronger. Currently Ryzen looks like it matches Kaby Lake with the current graphics cards at 1440p or 4K because the graphics card is the limiting factor. As graphics cards improve we will see the limiting factor switch back to the CPU and we will start to see Intels stronger IPC showing up again and delivering stronger frame rates like we currently do when testing at 1080p.

 

[Spect]

So for the normal everyday dude looking to build a new PC for web browsing, Kodi streaming, and some light gaming (not looking for "liquid cooled alien technology" but just to be able to play any game for the next couple years at a playable frame rate) what would you do? I've noticed that in a hour-to-hour comparison I spend twice the amount of time streaming movies & tv shows than anything else. I don't even use a monitor anymore, I keep my PC hooked up to a 55' 4k tv. But my computer is very very old now and I've put off upgrading for many years now. There's just so many charts, graphs, reviews & opinions that I get lost trying to research a new system. Where's the best bang for your dollar at this point? Would you even buy now? Does it make sense to wait and see what the next month or 2 brings to the table (which can easily turn into a never ending wait & see scenario)? Stick with the familiar but expensive Intel/Nvidia brands or take a chance on Ryzen/AMD? If you were to spend up to $1000 bucks what is the best way to go. Replacing a core2quad @ 3.6 GHz & a GTX 480. I think it's about 10-12 yrs old. Wow just realizing now how long it's been.. but it served me well. Thanks for any opinions.

 

[cartera]

So for the normal everyday dude looking to build a new PC for web browsing, Kodi streaming, and some light gaming (not looking for "liquid cooled alien technology" but just to be able to play any game for the next couple years at a playable frame rate) what would you do? I've noticed that in a hour-to-hour comparison I spend twice the amount of time streaming movies & tv shows than anything else. I don't even use a monitor anymore, I keep my PC hooked up to a 55' 4k tv. But my computer is very very old now and I've put off upgrading for many years now. There's just so many charts, graphs, reviews & opinions that I get lost trying to research a new system. Where's the best bang for your dollar at this point? Would you even buy now? Does it make sense to wait and see what the next month or 2 brings to the table (which can easily turn into a never ending wait & see scenario)? Stick with the familiar but expensive Intel/Nvidia brands or take a chance on Ryzen/AMD? If you were to spend up to $1000 bucks what is the best way to go. Replacing a core2quad @ 3.6 GHz & a GTX 480. I think it's about 10-12 yrs old. Wow just realizing now how long it's been.. but it served me well. Thanks for any opinions.

Sounds like you're best off looking at the link below:
https://www.techspot.com/guides/buying/

But that is from October last year so it may be worth waiting for the next one which is probably not to far away from being published.

Great article overall from Steve, thanks for the interesting read.

 

[Steve]

So for the normal everyday dude looking to build a new PC for web browsing, Kodi streaming, and some light gaming (not looking for "liquid cooled alien technology" but just to be able to play any game for the next couple years at a playable frame rate) what would you do? I've noticed that in a hour-to-hour comparison I spend twice the amount of time streaming movies & tv shows than anything else. I don't even use a monitor anymore, I keep my PC hooked up to a 55' 4k tv. But my computer is very very old now and I've put off upgrading for many years now. There's just so many charts, graphs, reviews & opinions that I get lost trying to research a new system. Where's the best bang for your dollar at this point? Would you even buy now? Does it make sense to wait and see what the next month or 2 brings to the table (which can easily turn into a never ending wait & see scenario)? Stick with the familiar but expensive Intel/Nvidia brands or take a chance on Ryzen/AMD? If you were to spend up to $1000 bucks what is the best way to go. Replacing a core2quad @ 3.6 GHz & a GTX 480. I think it's about 10-12 yrs old. Wow just realizing now how long it's been.. but it served me well. Thanks for any opinions.

I would say a Ryzen 5 CPU (quad-core) is what you will want.

 

[Raiderman]

I am wondering why TS is benching a 4c 8t Intel cpu against a 8c 16t Amd? To me that is Apples and oranges, and why you wouldn't choose a similar Intel 8c Cpu for comparison? Or why not just omit the Intel results altogether?

 

[Steve]

I am wondering why TS is benching a 4c 8t Intel cpu against a 8c 16t Amd? To me that is Apples and oranges, and why you wouldn't choose a similar Intel 8c Cpu for comparison? Or why not just omit the Intel results altogether?

Here I am, wondering why you are wondering any of that. Removing Intel from the picture would make it impossible to learn what we have. The 7700K is the king of gaming now, which is obviously why it was used for comparison.

 

[Raiderman]

Here I am, wondering why you are wondering any of that. Removing Intel from the picture would make it impossible to learn what we have. The 7700K is the king of gaming now, which is obviously why it was used for comparison.

Simple....The title states, "Does Ryzen perform better on AMD GPU's. The title does not say, how does Ryzen compare to the 7700K on AMD GPU's.

 

[Reikoji]

I simply have one thing to say, and that is about Ghost Recon: Wildlands. The game was made in partnership with NVidia for certain graphic enhancements including Nvidia's HBAO+, so it should be no surprise it performs so poorly on an AMD card compared to the Nvidia card. It's likely to end up the same way with AMD's partnership with Bethesda.

It all comes down to optimization in the end, and with Ghost Recon its clear AMD cards are either hindered or were an afterthought.

 

[Steve]

I simply have one thing to say, and that is about Ghost Recon: Wildlands. The game was made in partnership with NVidia for certain graphic enhancements including Nvidia's HBAO+, so it should be no surprise it performs so poorly on an AMD card compared to the Nvidia card. It's likely to end up the same way with AMD's partnership with Bethesda.

It all comes down to optimization in the end, and with Ghost Recon its clear AMD cards are either hindered or were an afterthought.

That's not the point of this test and that's not what we discussed. You are getting confused and are assuming this is a GPU comparison, it's not.

 

[Jibberish18]

DX12 actually showing improvements? Say it ain't so...