AMD Ryzen 7 2700X 3.7 GHz

[Charcharo]

I personally think 720P benches are useless unless they are used for E-Sport titles. Then they make sense.

 

[cucker tarlson]

I personally think 720P benches are useless unless they are used for E-Sport titles. Then they make sense.

Wrong. And personal beliefs should not be a part of this discussion.

Well clearly I do as I live in 2018

Seems like I have to to get the point across for those who clearly dont get it and still live back in 2003.

stop spreading fud and get back to the yr 2018 for god sake.

"It's 2018, get real" says a guy who runs a FX cpu in 2018. With a friggin sli setup.
You obviously don't know much about any sort of cpu limitations, neither do you know how to test them.

 

[Charcharo]

Wrong. And personal beliefs should not be a part of this discussion.

OK. My informed real world opinion as someone who plays old games and the most CPU demanding titles in existence, who until lat least year was using a 1440x900 screen to play on is that anything under 1080p/900p (at the least) is de facto irrelevant for high end modern parts.

If you disagree you must give me an objective reason that is not about "future performance" as that simply should not depend on single-threaded scenarios in this year. It is an issue of the developers at that point as it means their game will never (likely) run good, no matter what Intel or AMD do. If it is about 240hz gaming, I accept E-Sport titles. If it is about the insanely small, tiny minority of people who have ultra high end gear and play at 720p because an ugly blob bothers them less than not getting 200+ Fps... egh ok that is valid but damn is it niche.

 

[trparky]

Benchmarking at 720p puts the game performance squarely on the processor's shoulders as versus the GPU. It damn near removes the GPU from the equation completely. It's about the closest you can get to a pure CPU benchmark.

 

[Charcharo]

Benchmarking at 720p puts the game performance squarely on the processor's shoulders as versus the GPU. It damn near removes the GPU from the equation completely. It's about the closest you can get to a pure CPU benchmark.

I obviously know that. But it is academic, the difference between science and engineering . It tells you almost nothing real and nothing about the future.

Also I am unsure if it perfectly benches the CPUs as different operations in a scene take different time/resources to be calculated. CPUs are lightly impacted by resolution, but the lightly part still exists, and how they communicate with a GPU /GPU driver matters too.

 

[Vayra86]

Seems like I have to to get the point across for those who clearly dont get it and still live back in 2003. You just said it yourself, its a CPU review....but we are going to show you gaming benchmarks on a res from 2003? cant you see how irrelevant that is? it still is gaming benchmarks........at a res no one uses! with said hardware, so please stop spreading fud and get back to the yr 2018 for god sake.

Again im still waiting for that person that games at 720p with a i7/Ryzen 7 and a GTX 1080 Ti.

Another thick plank here then... Are you going to repeat the same broken record every few pages?

Spreading 'FUD'? If this is FUD to you, then you suffer from a serious lack of insight and experience. The line about 2003 and actually using said hardware on 'this res' shows you really don't get it. I've already supplied multiple examples of recent or still heavily played games that still rely on the same things as they did back in 2003 - and that should not be a surprise to anyone if you know how genuinely old most game engines really are - even today. Many of them are just new iterations of the same old stuff. Re read my earlier response on this. You don't have to agree, and you don't have to see it or believe it, that's fine, skip the 720p tests in that case and move on. I have better things to do than to keep repeating myself. Its your delusion not mine. The reality is, you bought an FX at some point, which is evidently one of the worst gaming CPUs for its time and it was a well known fact even then - and people who bought them used the same arguments as you're doing now. 'Everything's gonna be multi threaded anyway'. Its quite hilarious and even in 2018 its still completely wrong.

https://www.techpowerup.com/forums/threads/amd-ryzen-7-2700x-3-7-ghz.243209/page-10#post-3832275

I obviously know that. But it is academic, the difference between science and engineering . It tells you almost nothing real and nothing about the future.

Also I am unsure if it perfectly benches the CPUs as different operations in a scene take different time/resources to be calculated. CPUs are lightly impacted by resolution, but the lightly part still exists, and how they communicate with a GPU /GPU driver matters too.

Its not purely academic, low res benchmarks are perfect to see the relative performance between CPUs. Not more than one page ago we looked at Starcraft 2 and worst-case scenario benchmarks and we also concluded that even in modern games, the min. FPS hinges on that one big game thread no matter how many cores the game can use. The vast majority of games and engines still work this way, and there are a few exceptions to this rule, DOOM being a good example of an engine that really extracts high FPS from increased core counts as well as clocks, but the reality is that this game also barely uses any CPU whatsoever.

The abundance of shooters and first/third person games in many reviews are showing a twisted reality where most CPUs perform largely the same. But gaming is a lot more than shooters and I so happen to play very few shooters and much more of other game types, and many of them are capable of using every Ghz I throw at it but hardly benefit from core count. Stellaris being (another) good example of something really quite recent that still hinges on the same performance bottleneck as games did ten years back.

 

[Melvis]

If we can get 4.0-4.2ghz stable or higher while waiting for Ryzen 2 i'll be happy. I still on my FX-8350 waiting for Ryzen +

Wrong. And personal beliefs should not be a part of this discussion.


"It's 2018, get real" says a guy who runs a FX cpu in 2018. With a friggin sli setup.
You obviously don't know much about any sort of cpu limitations, neither do you know how to test them.

You find it funny that in 2018 im running a FX CPU with a SLi setup thats still able to play games at 1440P? I think the jokes on you kid
You obviously have no clue at all how a CPU works with modern software in 2018. Might need to go do some research before commenting next time?

Here I shall give you a few links so you can go looking for yourself and find out how even a older CPU is still able to play modern games even in 2018.

https://www.google.com.au/
https://www.youtube.com/

Another thick plank here then... Are you going to repeat the same broken record every few pages?

Spreading 'FUD'? If this is FUD to you, then you suffer from a serious lack of insight and experience. The line about 2003 and actually using said hardware on 'this res' shows you really don't get it. I've already supplied multiple examples of recent or still heavily played games that still rely on the same things as they did back in 2003 - and that should not be a surprise to anyone if you know how genuinely old most game engines really are - even today. Many of them are just new iterations of the same old stuff. Re read my earlier response on this. You don't have to agree, and you don't have to see it or believe it, that's fine, skip the 720p tests in that case and move on. I have better things to do than to keep repeating myself. Its your delusion not mine. The reality is, you bought an FX at some point, which is evidently one of the worst gaming CPUs for its time and it was a well known fact even then - and people who bought them used the same arguments as you're doing now. 'Everything's gonna be multi threaded anyway'. Its quite hilarious and even in 2018 its still completely wrong.

https://www.techpowerup.com/forums/threads/amd-ryzen-7-2700x-3-7-ghz.243209/page-10#post-3832275

Just another kid know it all we have here, and yes I will over and over again till you finally catch up with reality and get out of the dark ages.

Look we know your new to the computer world its ok we are here to educate you on these things so you can be more respected on forums when replying. We know that if you dont understand that 720p is a dead res from 2003 and that in the real world no one uses it with said hardware, its just a fact otherwise you or anyone else would be linking me to someone that actually games at 720p with said hardware, but alas im still waiting. You dont have to agree or like the facts thats totally up to you and you can just skip this and stop responding, its probably in your best interest honestly. So your going to skip the facts and comment on my FX system? which has no baring on this subject at all, now I know your reaching. Maybe before attacking my computer you should go back to when they came out and see the threads (yes ill be there) and you might learn something, but again we can tell your new to this since you only been on here since 2014 so if you cant find your answers I will be more then happy to explain it to you, so you can catch up with the rest of us oldies. FYI Multi-threaded programs/games is alot more now then when it was back in 2012.

 

[W1zzard]

Just had a thought about those YouTube videos. Doesn't sensor recording, overlay, video recording add additional load on the system which introduces some bias toward systems with more cores?

 

[Charcharo]

Another thick plank here then... Are you going to repeat the same broken record every few pages?

Spreading 'FUD'? If this is FUD to you, then you suffer from a serious lack of insight and experience. The line about 2003 and actually using said hardware on 'this res' shows you really don't get it. I've already supplied multiple examples of recent or still heavily played games that still rely on the same things as they did back in 2003 - and that should not be a surprise to anyone if you know how genuinely old most game engines really are - even today. Many of them are just new iterations of the same old stuff. Re read my earlier response on this. You don't have to agree, and you don't have to see it or believe it, that's fine, skip the 720p tests in that case and move on. I have better things to do than to keep repeating myself. Its your delusion not mine. The reality is, you bought an FX at some point, which is evidently one of the worst gaming CPUs for its time and it was a well known fact even then - and people who bought them used the same arguments as you're doing now. 'Everything's gonna be multi threaded anyway'. Its quite hilarious and even in 2018 its still completely wrong.

https://www.techpowerup.com/forums/threads/amd-ryzen-7-2700x-3-7-ghz.243209/page-10#post-3832275



Its not purely academic, low res benchmarks are perfect to see the relative performance between CPUs. Not more than one page ago we looked at Starcraft 2 and worst-case scenario benchmarks and we also concluded that even in modern games, the min. FPS hinges on that one big game thread no matter how many cores the game can use. The vast majority of games and engines still work this way, and there are a few exceptions to this rule, DOOM being a good example of an engine that really extracts high FPS from increased core counts as well as clocks, but the reality is that this game also barely uses any CPU whatsoever.

The abundance of shooters and first/third person games in many reviews are showing a twisted reality where most CPUs perform largely the same. But gaming is a lot more than shooters and I so happen to play very few shooters and much more of other game types, and many of them are capable of using every Ghz I throw at it but hardly benefit from core count. Stellaris being (another) good example of something really quite recent that still hinges on the same performance bottleneck as games did ten years back.

Oh mate I think we did not understand each other. I agree and even stated that games like Starcraft 2 (which I play, pretty well might I add as long as I am drunk!) do matter. I play such titles, I even mentioned Call of Chernobyl which when configured for max simulation simply slaughters CPUs. Starcraft 2 looks like an idle test in comparison for single-threaded benches in comparison, it is that bad.

My point is that the performance profile at 720p and at 1080p or above is not QUITE the same. People have this idea that CPU resources are unaffected by resolution (incorrect) and that all things a game needs require similar percentile resources or time to complete. That is not true, unfortunately. Now you will say "OK Alex, fair enough you are right. I will add that nuance. Still at 720P or low resolutions in general we see more of the CPU at work even with that nuance added in." and I would agree with you. I just do not think it matters for most titles bar E-Sports ones. That was my point. It is not me not knowing or disagreeing with you in general.

As for Starcraft and Stellaris (which I admit I do not play) ... egh I am divided. You see old games we can not expect to run perfectly well on new CPUs. I know that and that is the reason why I personally, whilst happy with my 1500X (at least it is never slower than my old Haswell i5!) will wait for Ryzen 3000/Zen 2 before I consider an upgrade. I hope for higher clocks and IPC especially in niche scenarios, from AMD. I play such titles a lot and want good performance in them. What I mean though is that the big onus is on those developers to work on their engines and games. In the case of Starcraft, it still gets dev time and is a popular title even to this day. Stellaris I have not played so I dont know. But if Ashes of the Singularity and Total War can do multi-threading well whilst using MUCH more CPU resources on every single front, and a low budget game like Men of War too can at least use 8 threads well (whilst simulating ballistics, destructible terrain, AI for every soldier, Inventory, armor penetration, destructible buildings and more) ... well there is no excuse. Starcraft 2 should receive an update in general and that would improve performance for all players. We should demand more as gamers too because IPC and clocks are at a wall. They can not improve by much, no matter what Intel or AMD do. At least not with silicon. That was my point.

Just had a thought about those YouTube videos. Doesn't sensor recording, overlay, video recording add additional load on the system which introduces some bias toward systems with more cores?

Should have a small impact but not for Digital Foundry. There it should be zero.
IMHO it would be cool if you can do some CPU tests under ... more unclean conditions. I should likely not mention how many things I have open most of the time else people will mock me !

 

[cucker tarlson]

You find it funny that in 2018 im running a FX CPU with a SLi setup thats still able to play games at 1440P? I think the jokes on you kid
You obviously have no clue at all how a CPU works with modern software in 2018. Might need to go do some research before commenting next time?

Here I shall give you a few links so you can go looking for yourself and find out how even a older CPU is still able to play modern games even in 2018.

https://www.google.com.au/
https://www.youtube.com/



Just another kid know it all we have here, and yes I will over and over again till you finally catch up with reality and get out of the dark ages.

Look we know your new to the computer world its ok we are here to educate you on these things so you can be more respected on forums when replying. We know that if you dont understand that 720p is a dead res from 2003 and that in the real world no one uses it with said hardware, its just a fact otherwise you or anyone else would be linking me to someone that actually games at 720p with said hardware, but alas im still waiting. You dont have to agree or like the facts thats totally up to you and you can just skip this and stop responding, its probably in your best interest honestly. So your going to skip the facts and comment on my FX system? which has no baring on this subject at all, now I know your reaching. Maybe before attacking my computer you should go back to when they came out and see the threads (yes ill be there) and you might learn something, but again we can tell your new to this since you only been on here since 2014 so if you cant find your answers I will be more then happy to explain it to you, so you can catch up with the rest of us oldies. FYI Multi-threaded programs/games is alot more now then when it was back in 2012.

Referring me to Youtube.com really says it about how comprehensive your knowledge is. No wonder people build rigs like yours, with all sorts of bottlenecks, if they have Youtube as reference. It's just a shame that they later go on quality forums like tpu and think they can lecture the reviewer

 

[intrepid3d]

Just had a thought about those YouTube videos. Doesn't sensor recording, overlay, video recording add additional load on the system which introduces some bias toward systems with more cores?

MSI After Burner: no
Shadow Rlay recording: very little, about 2% if anything

 

[Melvis]

Referring me to Youtube.com really says it about how comprehensive your knowledge is. No wonder people build rigs like yours, with all sorts of bottlenecks, if they have Youtube as reference. It's just a shame that they later go on quality forums like tpu and think they can lecture the reviewer

Oh dear you have got alot to learn and shows us how much ignorance and little knowledge you do have in the tech industry thats now a given fact. You tube videos are more credible then reviews done on tech sites, you get to see the tests been done and carried out infront of your eyes with live testing and benchmarking etc, alot harder to fake or give false info then just taking a screen shot and putting it on a website.
IF this is all you have as a comeback and all you can do is complain about youtube then you have already lost

 

[cucker tarlson]

Oh dear you have got alot to learn and shows us how much ignorance and little knowledge you do have in the tech industry thats now a given fact. You tube videos are more credible then reviews done on tech sites, you get to see the tests been done and carried out infront of your eyes with live testing and benchmarking etc, alot harder to fake or give false info then just taking a screen shot and putting it on a website.
IF this is all you have as a comeback and all you can do is complain about youtube then you have already lost

Your purchase is not my problem.
So you're saying no bottlenecks ?
Not with fx cpu ? Not with 2.0 lanes ?

 

[W1zzard]

MSI After Burner: no

Did some quick testing

Seems to be around 1-2% for the OSD (values vary a bit, but there's a clear difference)

More at higher FPS:

 

[intrepid3d]

Did some quick testing

Seems to be around 1-2% for the OSD (values vary a bit)

More at higher FPS:

1 - 2% is Margin of error, that's why values vary, no one run is identical to another.

Oh and you're using FRAPS with MSI After Burner, using two sensory monitoring applications is not a good idea as they can conflict.

 

[Charcharo]

Guys, if the man with the FX CPU is happy... then maybe he is correct for himself? Maybe it is OK for him? I mean it is ugly from the outside to see that even if you are technically correct.

Same way I tell people with Pentiums or i3s or FXes that if they are happy... they do not need to change their set up yet. Every single person is unique in what they do and how they do it with gaming.

 

[intrepid3d]

Guys, if the man with the FX CPU is happy... then maybe he is correct for himself? Maybe it is OK for him? I mean it is ugly from the outside to see that even if you are technically correct.

Same way I tell people with Pentiums or i3s or FXes that if they are happy... they do not need to change their set up yet. Every single person is unique in what they do and how they do it with gaming.

This, i'm mean for Christ Sake i have never understood pepole who attack others for thier hardware choices, its hardware snobbery and out right church mentality of the highest order.

 

[W1zzard]

1 - 2% is Margin of error, that's why values vary, no one run is identical to another.

Oh and you're using FRAPS with MSI After Burner, using two sensory monitoring applications is not a good idea as they can conflict.

values don't vary that much, test yourself and you'll see. i did not cherry pick results

fraps was the easiest for me to do a/b testing. a game with integrated fps counter would work too.

 

[intrepid3d]

values don't vary that much, test yourself and you'll see. i did not cherry pick results

fraps was the easiest for me to do a/b testing. a game with integrated fps counter would work too.

Wait, are you saying you don't get a 1 to 2% variation when runnig multiple game benchmarks, or any benchmark, because that is the understood norm, most reviewers run multiple runs and take avrages from the results. even cinebench which is recognized as the most consistent varies 1 or 2% between runs.

No. i don't trust your runs are 100% consistent, no ones are.

 

[W1zzard]

Wait, are you saying you don't get a 1 to 2% variation when runnig multiple game benchmarks, or any benchmark, because that is the understood norm, most reviewers run multiple runs and take avrages from the results. even cinebench which is recognized as the most consistent varies 1 or 2% between runs.

No. i don't trust your runs are 100% consistent, no ones are.

I'm not saying that. What I'm saying is that monitoring, OSD and capture add some CPU load (not a random variation).

Of course trying to measure the impact has some variation (a small percentage of the impact itself)

 

[cucker tarlson]

I'm not saying that. What I'm saying is that monitoring, OSD and capture add some CPU load (not a random variation).

Of course trying to measure the impact has some variation (a small percentage of the impact itself)

Recording and OSD - negligible
Streaming - yuge

This, i'm mean for Christ Sake i have never understood pepole who attack others for thier hardware choices, its hardware snobbery and out right church mentality of the highest order.

Lol I got nothing against the man's choice, I could not care less. Apologies if I came across too strong, but we gotta draw a line when he tries to discredit the review results just cause he's happy living in ignorance. I was pointing out what he was saying about us living in 2003 and him being so damn forward thinking was ridiculous and ironic.

Oh yeah, I missed it was you again. Well that explains A LOT cause I was just thinking why would anyone accuse me of hardware snobbery.

 

[Charcharo]

I'm not saying that. What I'm saying is that monitoring, OSD and capture add some CPU load (not a random variation).

Of course trying to measure the impact has some variation (a small percentage of the impact itself)

Egh if it is the same load for all CPUs, IMHO it don't matter.

I dislike clean install CPU benches too, but only because that is not how I work/play on my PC. I completely understand why they are done and do not want it to be changed if it is too much workload for both.

 

[intrepid3d]

I'm not saying that. What I'm saying is that monitoring, OSD and capture add some CPU load (not a random variation).

Of course trying to measure the impact has some variation (a small percentage of the impact itself)

Ah ok, its very low load, a couple of % at most and its the same for all CPU's.

 

[W1zzard]

Egh if it is the same load for all CPUs, IMHO it don't matter.

The load should be proportional to FPS, so not exactly constant, but probably fine, because it's just percentages on top of percentages.

My main thought is this though: This load can probably migrate to an unloaded core in a multi-core system. On a system with few cores, that are fully loaded already by the game, this will eat into the CPU time available to the game.

 

[Charcharo]

The load should be proportional to FPS, so not exactly constant, but probably fine, because it's just percentages on top of percentages.

My main thought is this though: This load can probably migrate to an unloaded core in a multi-core system. On a system with few cores, that are fully loaded already by the game, this will eat into the CPU time available to the game.

Seems fair enough to me. Wont affect benches noticeably IMHO. And is actually closer to real world results (even if not for the correct reasons).
Of course the DF way seems the best but it is harder.