For those of you that very rightly pointed out that the video demonstrated an increase in performance over Ryzen in only two games, the following video might be of interest.

This guy managed to get Ryzen RAM running at 3600 MHz. The result was that quite a number of games were close, equalled or beat the 7700k. And this was running the 7700k at 5 GHz.

His theory is that the Ryzen inter CCX communications are highly dependant on RAM speed. 

One game was showing a massive 40% increase in frame rate.

When Ryzen is more consistently handling high frequency RAM, it will be very interesting to see the results. 

 

 

Sorry, but the only games where Ryzen is *ever* going to beat a highly-clocked i7 are those which are extremely well-threaded, aka the vast MINORITY of games.  Any sampling of games in a pseudo-review such as this which shows any significant quantity of games running faster on Ryzen is either an anomaly or a purposeful choice made to mis-represent the quantity of these games in the market and shine a better light on a particular product.

AMD themselves have admitted Kaby Lake has better IPC (instructions per clock cycle) than Ryzen.  They've also admitted that they can't match the clock speed of Intel's chips due to both design and manufacturing differences.  You can hear AMD's Robert Hallock, their global technical marketing lead (i.e. the guy whose job it is to speak to the press and partners about technical details of their products) admit both of those points beginning at the 8:25 mark of this video here: 

When you add it all up, and account for overclocking differences which exacerbate the clock speed differential, the i7 7700k is as much as 36.9% faster than Ryzen in workloads which are not well-threaded.

 

Edit: just watched the video you posted.  The author is disingenuous, or does not know how to compare on an even footing.  He tested the Ryzen chip with RAM up to 3600MHz but for some reason only tested the 7700k with 3200MHz RAM.  Now, the actual degree of performance difference there is irrelevant, the fact that he tested with unequal conditions is undeniable.  Also, his choice of graphics card (a GTX 1070) means that many games will be limited by the GPU much of the time, even @ only 1080p.  In order to test a CPU properly, the bottleneck needs to be shifted to the CPU as much as possible.  This has historically been accomplished one of two ways:

1) run games with very low settings/resolutions (best way to shift bottleneck but unrealistic)

2) use the fastest graphics card available (far more realistic and therefore meaningful to end users)

15 hours ago, TechguyMaxC said:

1) run games with very low settings/resolutions (best way to shift bottleneck but unrealistic)

2) use the fastest graphics card available (far more realistic and therefore meaningful to end users)

Even with the fastest graphics card available, running at 4K at max settings (like anyone who spent a lot of money on the fastest card available would do) shifts the bottleneck to the GPU.

The takeaway point from this is that Ryzen is "fast enough" in gaming to utilize the fastest GPUs at the settings people typically use. The i7 7700K is absolutely faster in the majority of games today, but this doesn't make Ryzen "slow".

https://arstechnica.com/gadgets/2017/03/intel-still-beats-ryzen-at-games-but-how-much-does-it-matter/

Quote

The author is disingenuous, or does not know how to compare on an even footing.  He tested the Ryzen chip with RAM up to 3600MHz but for some reason only tested the 7700k with 3200MHz RAM. 

 

Well yes, but the title of the video is... "closing the 7700K gap in gaming". Thus, the point is that Ryzen benefits from faster RAM and helps to close the gap, at least in titles he tested. I don't think he stated anywhere that Ryzen was generally Superior to Kaby Lake in gaming or that the benefits he saw from faster RAM were across the board. I don't know of anyone that claims that.

1:31 he explains his RAM choices. He states that according to his own testing, going to 3600 on the i7 is minimal to non-existent in terms of performance improvements. You will probably disagree but there you go.

 

Quote

Sorry, but the only games where Ryzen is *ever* going to beat a highly-clocked i7 are those which are extremely well-threaded, aka the vast MINORITY of 

 

The point he was making was that, according to his tests, Ryzen benefits considerably from higher frequency RAM. That's it. And of course he did specify in the video that faster RAM with Ryzen will help performance in "extremely CPU limited scenarios". 4:12. 

7:01 he clearly states that he's not pitching for either platform, just exploring ways we can improve gaming performance re Ryzen CPU's. Thus I think you may have been somewhat unfair in your criticism. 

Quote

AMD themselves have admitted Kaby Lake has better IPC (instructions per clock cycle) than Ryzen.  They've also admitted that they can't match the clock speed of Intel's chips due to both design and manufacturing differences.

 

Err... yes we know Max. 

Another video to watch, Ryzen 5 this one, compared with 7600K. Barely any difference in frame rate between games. Minimal in percentage terms.

13:13 "Gaming performance up to par with 7600K". 

13:21 "Clear evidence in past month that Ryzen has a lot of power left under the hood. "And game optimisations and faster RAM have been shown to draw out more performance".

But of course, lots of variables for Max to get a bee in his bonnet about. 

 

 

8 hours ago, martin-w said:

 

. I don't think he stated anywhere that Ryzen was generally Superior to Kaby Lake in gaming......... I don't know of anyone that claims that.

 

The title of this thread could sure lead someone to believe just that..

14 hours ago, Bert Pieke said:

The title of this thread could sure lead someone to believe just that..

 

It could. To some perhaps. But Max was referring to the video being disengenous, not my title. The entire debate was about that, not my title.

Having said that, my title is correct in that Ryzen, in his video, did beat Intel in games. Don't think my title said "all games" did it? 

 

 

A never ending argument.. Yet no one (no one) has brought a video comparing Intel vs Ryzen  (using the exact same settings) on P3D (Again... On P3D) 

 

That would be worth seeing. 

6 hours ago, joemiller said:

Yet no one (no one) has brought a video comparing Intel vs Ryzen  (using the exact same settings) on P3D (Again... On P3D)

Of course, we all know the "professional" testers (i.e. the guys making $ creating YouTube videos/blog columns centered on the current Intel/AMD state of affairs) couldn't care less about P3D... or any sim.  If a valid comparison ever does happen it will come from within our community.  Doesn't hurt to hope...

Greg

18 hours ago, lownslo said:

Of course, we all know the "professional" testers (i.e. the guys making $ creating YouTube videos/blog columns centered on the current Intel/AMD state of affairs) couldn't care less about P3D... or any sim.  If a valid comparison ever does happen it will come from within our community.  Doesn't hurt to hope...

Greg

I'm being honest... Would love to see the performance, with same settings, same version, same RAM, etc.etc.etc.I almost build a Ryzen system but got carried away by our "knowledgeable"  sim guys. But oh well... too bad. 

2 hours ago, joemiller said:

I'm being honest... Would love to see the performance, with same settings, same version, same RAM, etc.etc.etc.I almost build a Ryzen system but got carried away by our "knowledgeable"  sim guys. But oh well... too bad. 

To be honest,  it's your responsibility, if you ar dissipointed with kabylake you can buy a  Ryzen.

I have a Ryzen system, nothing a install P3D on have only one licence not worth the work to transfer and retransfer the licence.

The memspeed on Ryzen is depending on mems,  mobo, bios and last the memcontroller on the cpu.

Not all do 3600mhz 

Good luck with your Ryzen system, I am happy with it , run P3D on Intel Kabylake for some reason .

11 hours ago, westman said:

To be honest,  it's your responsibility, if you ar dissipointed with kabylake you can buy a  Ryzen.

I have a Ryzen system, nothing a install P3D on have only one licence not worth the work to transfer and retransfer the licence.

The memspeed on Ryzen is depending on mems,  mobo, bios and last the memcontroller on the cpu.

Not all do 3600mhz 

Good luck with your Ryzen system, I am happy with it , run P3D on Intel Kabylake for some reason .

ok

I had a Lincoln V12. A new Ford V6 Ecoboost 2,7 will run all over it, they both gave you a great ride.

Intel will give you higher FPS's. But their is more to it than raw FPS's. The average fps is as important as is the presence or lack of stutters, the end result is smoothness or choppy frame rates. When FSX first arrived a friend got it and asked me to check it out. He had an old rig. the takeoff roll would stop and then resume. Prior to that event he came to my shop because he could not download updates. The status bar would stop halfway thru the download. It simply ran out of memory, didn't crash, just stopped. Added more memory and had no more problems.

When you have enough resources to generate enough FPS's for smooth gameplay that's what counts. If all you do is play single threaded games then go for intels CPU's. I have a customer that own a machine shop, I built him a new system (supposed to be retired) it basically runs the whole shop, bill of materials accounting, print a barcoded shop follower for each job to swipe when each operation is completed. Each job has to be coded for the CNC boring mills etc. He used to have to take an hour at times to find a job and tell his customer where it was and a completion estimate. Now he does it by typing in the job #, takes about three seconds! After work he like to fly, he does. He tells me he has no problems and is talking about  3PD. He runs photo shop faster than his friends $5,000 Mac pro, like twice as fast?

I said a long time ago, It would be close enough. I also stated a few more frames at what cost.

Any way good luck to all

BaidyB