AMD Ryzen reviewers say: - Either experiencing weird results or not recommened for gaming

[xAcid9]

3 minutes ago, Taf the Ghost said:

Except that the R7s are beasts at compile & render tasks.  These things, especially the normal 1700, is going to be really popular in a lot of smaller companies that need either render or compile boxes.  I'm really curious about LTT playing with it for their editing systems over the next year.

 

Also, the whole "It didn't beat the 7700k in gaming, so it's trash!" seems like the fanboy response from the Blue team more than anything.  Especially when you point out that most games don't even need over a i5 7600k, not an i7 7700k, to max out at anything but 1080p settings.  The ~110USD you save (NewEgg USA pricing) by going with the i5 is much better spent on going up a Graphics Card tier.  Going from a 1060 to a 1070 is worth far more than going up to the i7.

 

The one issue, though, since we have enough data now, is some something is "off" with the R7 numbers in gaming.  It's maybe 10% slower than it should be.  That's because the Min FPS never seems to drop in line with the way Intel chips do.  Which either points to ways in which Intel chips "peak" in games or it points to AMD some how having a different scaling within game engines for reasons we don't know yet.  (Aside from the clear Memory issues that seem to be plaguing testing.)

In GN review Ryzen 1% and 0.1% framerate percentile is much worst compare to Intel proc though, what's up with that?

[TechGod]

10 minutes ago, xAcid9 said:

In GN review Ryzen 1% and 0.1% framerate percentile is much worst compare to Intel proc though, what's up with that?

Check my thread that I made yesterday but essentially Windows Scheduler is treating the Ryzen cores incorrectly. It thinks you have a 16 core CPU instead of a 8c/16t but don't worry. A Windows update should fix the compatibility issues and also boost minimums (W7 perf is way better with Ryzen than W10)

[Kinda Bottlenecked]

Has anyone reviewed multi tasking on the Ryzen? 

 

7700K has high frames in games, yes. But someone thats multi tasking will ruin the 7700K's towering lead it has over the rest of the CPUs. Drops and stutters will start sprouting when you try to(i suspect). 

 

Where did this go? 

 

Why is no one testing the consistency of the fps in game when there are more intensive tasks running in the background? 

Why is no one challenging their claim of their superior multitasking? 

It was shown that the 7700K drops frames everywhere while streaming vs the 1700 so why is no one challenging that? 

 

I'd like to see this again^ with current CPUs since now we do have really fast 5Ghz 4 cores. 

[ZeGhostrider]

I really don't care for all this "not good for gaminglalala", just ordered a R7 1700.

 

Reviews are all over the place mixed with rumors of Intel bribing, going to see for myself if Ryzen is any good or not.

(if you have the means I recommend you doing the same)

 

IF it performs anywhere near my 5820K, I'll keep it mainly to support AMD.

 

And IF Vega manages to give me slightly more than GTX 1080 level of performance, I will finally have a proper AMD build.

But honestly I STILL mostly use my old i5 4690K + 280X (console substitude build) in my living room to game anyway, it's still very capable.

[Valentyn]

5 hours ago, xAcid9 said:

In GN review Ryzen 1% and 0.1% framerate percentile is much worst compare to Intel proc though, what's up with that?

GN is in the minority of reviewers in that regard. Far more find it gives better overall minimums, and a smoother gameplay experience than the 7700K; and viewers actually ask about it in the lengthy video below as well ( great watch between Wendell and more techies). Plenty of people have been noticing stuttering; and hitching on the Kabylake vs Ryzen.

 

Then add in the fact that people testing Ryzen in Windows 7 gets far better minimums as well, and it really looks like the Windows 10 scheduler is messing up badly for performance in some programs and games.

 

https://forums.anandtech.com/threads/ryzen-strictly-technical.2500572/

Quote

 

"I did some 3D testing and even thou there is not nearly enough data to confirm it, I'd say the SMT regression is infact a Windows 10 related issue.
In 3D testing I did recently on Windows 10, the title which illustrated the biggest SMT regression was Total War: Warhammer.

All of these were recorded at 3.5GHz, 2133MHz MEMCLK with R9 Nano:

Windows 10 - 1080 Ultra DX11:

8C/16T - 49.39fps (Min), 72.36fps (Avg)
8C/8T - 57.16fps (Min), 72.46fps (Avg)

Windows 7 - 1080 Ultra DX11:

8C/16T - 62.33fps (Min), 78.18fps (Avg)
8C/8T - 62.00fps (Min), 73.22fps (Avg)

 

 

https://www.computerbase.de/2017-03/amd-ryzen-1800x-1700x-1700-test/4/#diagramm-battlefield-1-dx11-multiplayer-frametimes-ryzen-7-1800x-gegen-core-i7-7700k

 

 

The only line from Intel that offers comparable, and better minims is x99 so far.

 

 

 

 

 

 

[cj09beira]

5 hours ago, xAcid9 said:

In GN review Ryzen 1% and 0.1% framerate percentile is much worst compare to Intel proc though, what's up with that?

 

5 hours ago, Taf the Ghost said:

Except that the R7s are beasts at compile & render tasks.  These things, especially the normal 1700, is going to be really popular in a lot of smaller companies that need either render or compile boxes.  I'm really curious about LTT playing with it for their editing systems over the next year.

 

Also, the whole "It didn't beat the 7700k in gaming, so it's trash!" seems like the fanboy response from the Blue team more than anything.  Especially when you point out that most games don't even need over a i5 7600k, not an i7 7700k, to max out at anything but 1080p settings.  The ~110USD you save (NewEgg USA pricing) by going with the i5 is much better spent on going up a Graphics Card tier.  Going from a 1060 to a 1070 is worth far more than going up to the i7.

 

The one issue, though, since we have enough data now, is some something is "off" with the R7 numbers in gaming.  It's maybe 10% slower than it should be.  That's because the Min FPS never seems to drop in line with the way Intel chips do.  Which either points to ways in which Intel chips "peak" in games or it points to AMD some how having a different scaling within game engines for reasons we don't know yet.  (Aside from the clear Memory issues that seem to be plaguing testing.)

It seems like there is a problem with the scheduler in windows with is not handling the ryzen chip as a 8c 16thr but as a 16c/16thr which makes it underperform and for some odd  reason win 7 is working just fine. 15fps increase in the minimums which is huge.

[xAcid9]

8 hours ago, TechGod said:

Check my thread that I made yesterday but essentially Windows Scheduler is treating the Ryzen cores incorrectly. It thinks you have a 16 core CPU instead of a 8c/16t but don't worry. A Windows update should fix the compatibility issues and also boost minimums (W7 perf is way better with Ryzen than W10)

I see, then i hope the update will be ready before Ryzen R5 release because that is the only good timing for AMD to patch up this mess in my opinion.

3 hours ago, Valentyn said:

GN is in the minority of reviewers in that regard. Far more find it gives better overall minimums, and a smoother gameplay experience than the 7700K; and viewers actually ask about it in the lengthy video below as well ( great watch between Wendell and more techies). Plenty of people have been noticing stuttering; and hitching on the Kabylake vs Ryzen.

 

Then add in the fact that people testing Ryzen in Windows 7 gets far better minimums as well, and it really looks like the Windows 10 scheduler is messing up badly for performance in some programs and games.

 

https://forums.anandtech.com/threads/ryzen-strictly-technical.2500572/

-snip-

Cool, that make GN review somewhat invalid then? 

 

I read The Stilt's thread, many good info there. Hopefully he will make another world record with AMD CPU. 

[alphaproject]

The Ryzen chip is 50% cheaper than the 6900 though... It beats it in many applications for 50% of the cost except gaming...so what?

 

If you have not upgraded like myself since the I5-2500K or I7-2600K the Ryzen is one hell of an upgrade still. The Multicore performance is what gets me excited. Sure the gaming performance doesn't blow me away but it doesn't need to since I'm using an old 2500K anyhow.

 

Of course if I was an early adopter and bought a 7700K or something... yea I guess Ryzen might have let me down, but then again...If I had the 7700K I wouldn't care about Ryzen now would I.

 

Who here owns the higher up Intel CPU's and really thought the Ryzen would absolutely change the CPU world? I never once thought that. I just thought it would make a great editing rig for video/audio and would be great to game on as well. The one thing Intel did do was lower prices..so thank Ryzen for that.

[Valentyn]

49 minutes ago, xAcid9 said:

I see, then i hope the update will be ready before Ryzen R5 release because that is the only good timing for AMD to patch up this mess in my opinion.

Cool, that make GN review somewhat invalid then? 

 

I read The Stilt's thread, many good info there. Hopefully he will make another world record with AMD CPU. 

I wouldn't say it makes it invalid, results is extremely varied between reviewers. Based on Motherboards, their BIOSs, RAM, and other factors.

The best thing is still to wait a month or two to see if the Windows 10 Scheduler, and BIOSs get sorted, and if that bring any real performance improvements.

As it stands now, Ryzen R7 is a fantastic Price to Performance winner if you need a Workstation that competes with X99.

For Pure gaming I wouldn't consider it just yet, then again who would also consider an X99 5960X or 6900K for pure gaming either considering the price to performance versus the 7600 and 7700K.

[Guest Kloaked]

On 3/4/2017 at 1:19 PM, Darth Revan said:

I had no idea AMD had so many fanboys.

I take it you're new to the tech community?

[79wjd]

6 hours ago, ZeBuzzSaw said:

mixed with rumors of Intel bribing

Don't forget that AMD is just as guilty (asking reviewers to gimp Intel setups to make AMD look better).

 

Also, the actions of a salesperson trying to protect his client base by offering then a compelling deal isn't exactly a "dick" move.

[dtaflorida]

On 3/2/2017 at 8:36 AM, kladzen said:

And frankly i'm quite scared about the results... Either it's not as good as we thought or it's some other underlying issues...

*GASP* *SHOCK* *UTTER AMAZEMENT*

/S

 

Same old same old. Which is worse: Disappointment when reality being greater than marketing? Or disappointment from reality being status quo? 

[Jahramika]

What a joke Thread I have the 6700k and 1800x both game very well not a problem.

[79wjd]

8 hours ago, Jahramika said:

What a joke Thread I have the 6700k and 1800x both game very well not a problem.

No one is saying an 1800x won't game well. It will, even a 2600k games well. That doesn't make it a smart purchasing decision when the sole purpose is gaming. 

 

Would I personally buy a 1700 over a 7700k and lose performance for the time being in the hopes that I'll gain much more (than a 7700k) in the future? Yes. Does that make it any smarter of a decision? Not really, it's still a complete gamble betting on future performance that may very likely never come. Is it still a risk I would personally be willing to take? yes. At the end of the day, a 6700k/7700k costs about the same, or slightly less, than a 1700 and offers better performance for gaming. 

[MoraisGT]

Interesting article from eTeknix, where they compare the 1800X with the 5820K and 7700K in gaming at 1080p, 1440p and 4K.

 

http://www.eteknix.com/nvidia-gtx-1080-ti-cpu-showdown-i7-7700k-vs-ryzen-r7-1800x-vs-i7-5820k/all/1/

 

One very impressing aspect is the minimum FPS, in some cases the Ryzen chip has over a 10fps lead even comparing with the 7700K.

 

And the bad performance at 1080p seems to be almost gone now...

[Valentyn]

6 minutes ago, MoraisGT said:

Interesting article from eTeknix, where they compare the 1800X with the 5820K and 7700K in gaming at 1080p, 1440p and 4K.

 

http://www.eteknix.com/nvidia-gtx-1080-ti-cpu-showdown-i7-7700k-vs-ryzen-r7-1800x-vs-i7-5820k/all/1/

 

One very impressing aspect is the minimum FPS, in some cases the Ryzen chip has over a 10fps lead even comparing with the 7700K.

 

And the bad performance at 1080p seems to be almost gone now...

Have you seen @LinusTech video review for the GTX 1080Ti?
The 1800X gets double the minimum FPS in GTA V compared to the 7700K at 4K.

In fact the 1800X has better minimums for all GPUs in GTA V

[MoraisGT]

Just now, Valentyn said:

Have you seen @LinusTech video review for the GTX 1080Ti?
The 1800X gets double the minimum FPS in GTA V compared to the 7700K at 4K.

In fact the 1800X has better minimums for all GPUs in GTA V

--image_snip--

I honestly didn't pay much attention to the numbers but now that you mention it, it is truly remarkable, especially considering it's at 4k, where the CPU should not have a stronger influence in FPS as in 1080p.

[SpaceGhostC2C]

18 minutes ago, MoraisGT said:

Interesting article from eTeknix, where they compare the 1800X with the 5820K and 7700K in gaming at 1080p, 1440p and 4K.

 

http://www.eteknix.com/nvidia-gtx-1080-ti-cpu-showdown-i7-7700k-vs-ryzen-r7-1800x-vs-i7-5820k/all/1/

 

One very impressing aspect is the minimum FPS, in some cases the Ryzen chip has over a 10fps lead even comparing with the 7700K.

 

And the bad performance at 1080p seems to be almost gone now...

It's kind of strange, though. I mean, forget about Ryzen for a moment: the stock 5820K is also almost always above the 7700K (sometimes loses on the average but still leads in minimum). Wasn't the 7700K the top Intel gaming cheap, with very few exceptions (the few games that for one reason or another use a lot of cores)? 

[Valentyn]

5 minutes ago, SpaceGhostC2C said:

It's kind of strange, though. I mean, forget about Ryzen for a moment: the stock 5820K is also almost always above the 7700K (sometimes loses on the average but still leads in minimum). Wasn't the 7700K the top Intel gaming cheap, with very few exceptions (the few games that for one reason or another use a lot of cores)? 

Few exceptions? Hmm, more modern games do like the cores

 

 

 

https://www.computerbase.de/2017-03/amd-ryzen-1800x-1700x-1700-test/4/#diagramm-battlefield-1-dx11-multiplayer-fps
 

 

 

 

 

 

http://pclab.pl/art73043-4.html
 

 

[MoraisGT]

11 minutes ago, SpaceGhostC2C said:

It's kind of strange, though. I mean, forget about Ryzen for a moment: the stock 5820K is also almost always above the 7700K (sometimes loses on the average but still leads in minimum). Wasn't the 7700K the top Intel gaming cheap, with very few exceptions (the few games that for one reason or another use a lot of cores)? 

Like @Valentyn said. Modern games are taking advantage of more and more cpu cores.

 

In BF1 for example you can see the 7700K hit usages of over 90%. That's why you see the 5820K pull ahead in some cases.

[SpaceGhostC2C]

1 minute ago, Valentyn said:

Few exceptions? Hmm, more modern games do like the cores

 

 

  Reveal hidden contents

 

 

https://www.computerbase.de/2017-03/amd-ryzen-1800x-1700x-1700-test/4/#diagramm-battlefield-1-dx11-multiplayer-fps

 

Battlefield 1

 

 

 

 

 

 

http://pclab.pl/art73043-4.html

 
 

 

 

Your second link has mainstream  i7s and i5s at the top, no 2011 CPUs even on the list 

 

Anyway, my point remains: when people do benchmarks, the 7700K tends to come on top in most games, isn't it? Or is everyone now benchmarking against the 2011-3 chips? I don't really care about game benchmarks so I don't follow too close, but my impression was that the conventional wisdom was "7700K=highes SC performance + 4C/8T = gaming champion", and the whole "games use more threads" was barely moving things form i5 to i7.

 

iIn any case, I'm not trying to argue in any direction, I'm simply asking if the i7 vs i7 results in that link are business as usual or are also different.

[SpaceGhostC2C]

Just now, MoraisGT said:

In BD1 for example you can see the 7700K hit usages of over 90%. That's why you see the 5820K pull ahead in some cases.

But that doesn't mean anything: the game could max out the existing cores, yet be incapable to expand to additional cores if available.

 

Anyway, I clarified my question above: are these results consistent with others or not?

[MoraisGT]

Just now, SpaceGhostC2C said:

But that doesn't mean anything: the game could max out the existing cores, yet be incapable to expand to additional cores if available.

 

Anyway, I clarified my question above: are these results consistent with others or not?

But BF1 is capable of using more cores that what the 7700K has.....that's why I mentioned it.

 

It is consistent with some. Including Linus's video of the 1080Ti. But at this point it's a bit difficult to compare because of all the bugs both from Windows and the boards.

[Valentyn]

5 minutes ago, SpaceGhostC2C said:

Your second link has mainstream  i7s and i5s at the top, no 2011 CPUs even on the list 

 

Anyway, my point remains: when people do benchmarks, the 7700K tends to come on top in most games, isn't it? Or is everyone now benchmarking against the 2011-3 chips? I don't really care about game benchmarks so I don't follow too close, but my impression was that the conventional wisdom was "7700K=highes SC performance + 4C/8T = gaming champion", and the whole "games use more threads" was barely moving things form i5 to i7.

 

iIn any case, I'm not trying to argue in any direction, I'm simply asking if the i7 vs i7 results in that link are business as usual or are also different.

 

Well in the computerbase tests you can clearly see the 7700K losing a lot against X99 in modern titles.

"7700K=highes SC performance + 4C/8T = gaming champion"

Only really applies to some current, and older games it seems, as more games are using more threads.

 

Also Ryzen R7's primary competition is X99/2011-3 after all, so it makes sense to bench against those.

[SpaceGhostC2C]

Just now, MoraisGT said:

It is consistent with some. Including Linus's video of the 1080Ti. But at this point it's a bit difficult to compare because of all the bugs both from Windows and the boards.

Which bugs? I'm talking 5820K vs 7700K, not Ryzen. I haven't seen the 1080ti video, but the other link is other about the 1080i. Therefore my follow-up question would be: are these in line with what people was getting for other GPUs?

Just now, Valentyn said:

 

Well in the computerbase tests you can clearly see the 7700K losing a lot against X99 in modern titles.

"7700K=highes SC performance + 4C/8T = gaming champion"

Only really applies to some current, and older games it seems, as more games are using more threads.

 

Also Ryzen R7's primary competition is X99/2011-3 after all, so it makes sense to bench against those.

OK, but you all go too far with things I didn't referred to  Let's pull a Philadelphia and explain it like I'm 4 years old:

 

7700K vs 5820K results: tick the correct answer

A) "Wow, that's surprising and unusual!"

B) "Same as always, nothing to see here"