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

[tlink]

Just now, MyName13 said:

Because less number of cores shouldn't increase performance (if clock rate is the same)?

https://cdn.videocardz.com/1/2017/02/AMD-Ryzen-5-Coming-Q2.jpg

 

less cores should increase stability and heat. thats why extreme overclockers disable cores and hyperthreading and stuff like that i thought? and that still is a 6 core 12 thread cpu, not their 4 cores 8 threads, and i can't confirm if that is official or a faked leak from that picture alone

[Bananasplit_00]

20 hours ago, RKRiley said:

Why did people expect an 8 core/16 thread to be as good as a 7700k for gaming in the first place?

 

if thats what anyone was expecing im dissapointed in peoples knowlage of how CPUs work. whats strange is that its being beaten to shits by almost EVERYTHING that isnt FX in games but in benchmarks its up there with the 6900K. 

[Bananasplit_00]

5 minutes ago, tlink said:

less cores should increase stability and heat. thats why extreme overclockers disable cores and hyperthreading and stuff like that i thought? and that still is a 6 core 12 thread cpu, not their 4 cores 8 threads, and i can't confirm if that is official or a faked leak from that picture alone

max overclock is usiually an architectural limit, less cores with the same architecture will clock basically the same. go look at the avrage overclock for the Broadwell-e chips and you will see that they reach verry similar clocks

[tlink]

1 minute ago, Bananasplit_00 said:

max overclock is usiually an architectural limit, less cores with the same architecture will clock basically the same. go look at the avrage overclock for the Broadwell-e chips and you will see that they reach verry similar clocks

yea the max overclock usually is an architectural limit, the max overclock is apparently 5.2 ghz. the overclock on air and water can improve tough because those are not architectural limits but cooling limits.

[Valentyn]

14 minutes ago, MyName13 said:

Because decreasing the number of cores shouldn't increase performance (if clock rate is the same)?

https://cdn.videocardz.com/1/2017/02/AMD-Ryzen-5-Coming-Q2.jpg

 

Ignoring that price is going to be much lower as well.

The 1700 is $329, how much is the 4c/8t Ryzen going to be.

Some reviewers wit Gigabyte motherboards had much better gaming benchmarks than the Asus and Asrock ones; while being able to run memory at 3000Mhz.

I think the R3 and R5 Ryzen cpus will do rather well when they're out; especially if motherboard manufacturers sort out their issues.

[Organized]

I don't think that the chips with fewer cores will overclock any better, because I guess the main problem with overclocking is that the L3 cache is clocked at the frequency of the highest clocked core.

 

It's a bit disappointing though for me as a fan of overclocking that I probably won't be able to get a "wow"-overclock with Ryzen. But that's not a problem with performance, it's just the fact that there will be less fun playing around with the OC. Also the recommended max. Vcore is so low, that you almost have no chance to get a nice overclock without going beyond the recommended boundaries.

 

However, I don't think that the Ryzen R7 launch is a flop, absolutely not, they increased their IPC by A LOT (comparing to Bulldozer) and they have made up massive workhorses for a really reasonable price point. There is no reason for content creators anymore to go and build a X99 system for a hell of a lot more money, except they have really special needs like many PCIe lanes.

I will definately wait for the R5 launch, I think this will be the area where you get the best bang for the buck out of Ryzen. Still 6 cores is more than Intel has to offer (for consumers) and I think it will be a lot cheaper than comparable Intels. Multicore performance will be around the same as the 7700k, singlecore performance will obviously be less, BUT it will also be a lot cheaper. I don't need absolutely maxed out fps, I only have a GTX 1060, therefore I don't need a 7700k which will probably in 2 years has less cores than recommended.

[Organized]

8 minutes ago, tlink said:

yea the max overclock usually is an architectural limit, the max overclock is apparently 5.2 ghz. the overclock on air and water can improve tough because those are not architectural limits but cooling limits.

Well, I don't think they sticked even in the area of the recommended max. Vcore for these 5.2 GHz You do these overclocks just for one benchmark and then shut the system down and be happy. It's not even close to a 24/7 overclock. Overclockers pushed the FX8350 to whopping 8.8 GHz, that's 4 GHz more than I get out of my 8350 for everyday use. Don't compare overclocking records with "max. archievable OCs for everyday use".

[Morgan MLGman]

20 hours ago, Kloaked said:

Stop making excuses for them. They literally talked about how much better than they were than Intel's 6800k and 6900k. I guarantee you based on how this looks, the 6-core and 4-core chips are going to suck.

 

It's going to be Bulldozer all over again it looks like.

Not necessarily. Gaming performance is yet to be fixed from what I know as the architecture is brand new and nothing is optimized for it, also, weird performance results are tied to the SenseMI technology having issues with Windows power management settings and some reviewers set the platform up in a wrong way, this should be fixed with a microcode update in around a month (at least this is what AMD said).

 

About those 4 and 6 cores, there is no reason why they would "suck". Why?

 

4 and 6 core Ryzen chips should come with higher clock speeds. 8 cores are always clocked low due to heat that 8 cores emit, exactly the same practice as Intel is doing, compare the 4790K to 5960X, see the difference in clock speeds and overclocking potential? It should be the same with Ryzen. Those 4 and 6 core parts will most likely come with higher clocks due to lower heat output because of lower core count, and judging that Ryzen's IPC is basically at Broadwell level, they should be the ones gamers choose.

 

If you can get a 4 core / 8 threads or a 6 core / 12 threads CPU with Broadwell IPC at 4GHz base and 4,3Ghz boost for instance with overclocking potential up to 4,7-4,8GHz with an AIO for the price of an i5, then they'd be great for gaming.

 

19 minutes ago, Bananasplit_00 said:

max overclock is usiually an architectural limit, less cores with the same architecture will clock basically the same. go look at the avrage overclock for the Broadwell-e chips and you will see that they reach verry similar clocks

Then why is the 5960X clocked at 3GHz when the 4790K is clocked 1GHz higher? Both are Haswell. You also won't overclock the 5960X nearly as high as the 4790K under normal coling (whether an AIO, Air or WC loop). Both are Haswell, again.

Heat is an issue with high-core count chips and it's probably the reason why R7 doesn't overclock high, temperatures are the issue.

[zMeul]

10 hours ago, N3v3r3nding_N3wb said:

Yes, if we're talking multithreaded.  If we're talking singlethreaded, the performance should be better.

wait .. what?!

[XenosTech]

1 hour ago, Morgan MLGman said:

Then why is the 5960X clocked at 3GHz when the 4790K is clocked 1GHz higher? Both are Haswell. You also won't overclock the 5960X nearly as high as the 4790K under normal coling (whether an AIO, Air or WC loop). Both are Haswell, again.

Heat is an issue with high-core count chips and it's probably the reason why R7 doesn't overclock high, temperatures are the issue.

Not really the temps... More cores means more voltage to oc and they hit the voltage limit pretty fast where as  you could clock a single core to 7 or 8 ghz and they the oc potential gets smaller with the addition of more cores.

[Morgan MLGman]

Just now, XenosTech said:

Not really the temps... More cores means more voltage to oc and they hit the voltage limit pretty fast where as  you could clock a single core to 7 or 8 ghz and they the oc potential gets smaller with the addition of more cores.

Well, with proper cooling you're not limited by voltage though in terms of the CPU, the motherboard will be the first limit as it won't be able to deliver the voltage beyond some point. Overclockers OC'd the R7 1800X to 5,2GHz at over 1,8V, that wasn't the issue, the temperatures were as they were already using LN2 to cool it... Though indirectly: voltage + clocks + amount of cores = heat...

[XenosTech]

20 minutes ago, Morgan MLGman said:

Well, with proper cooling you're not limited by voltage though in terms of the CPU, the motherboard will be the first limit as it won't be able to deliver the voltage beyond some point. Overclockers OC'd the R7 1800X to 5,2GHz at over 1,8V, that wasn't the issue, the temperatures were as they were already using LN2 to cool it... Though indirectly: voltage + clocks + amount of cores = heat...

True but still it's way harder for us consumers to oc the balls out of 8 core chips with the means we have (that are affordable and make sense.)

[Morgan MLGman]

9 minutes ago, XenosTech said:

True but still it's way harder for us consumers to oc the balls out of 8 core chips with the means we have (that are affordable and make sense.)

Yeah, still, it is only logical to conclude that the 4 and 6 core Ryzen chips will come clocked higher out of the box and will overclock higher on consumer-grade cooling solutions than the 8 core chips, considering IPC of almost exactly Broadwell level (actually a tiny bit higher judging by cinebench scores but it's within a margin of error) those chips should really be what gamers look for, not the R7 lineup.

 

5960X and 6900K also come clocked at relatively low frequencies compared to their quad core parts from the same generation, they were never considered for gaming as they cost a RIDICULOUS amount of money, people are looking at R7's gaming performance only because they come in normal prices compared to Intel's offerings.

 

It's also funny cause when people say that AMD is bad in gaming and good in productivity and CPU heavy tasks, they never say that on the other hand Intel is good in gaming and bad in productivity and heavy tasks since you get similar mulithtreaded performance for the money, with a chip that's 2x cheaper (1800X vs 6900K), what's more important you can get the 1800X + GTX 1080 + X370 platform for the price of a 6900K + X99 platform itself... And still have the same productivity potential.

[XenosTech]

13 minutes ago, Morgan MLGman said:

Yeah, still, it is only logical to conclude that the 4 and 6 core Ryzen chips will come clocked higher out of the box and will overclock higher on consumer-grade cooling solutions than the 8 core chips, considering IPC of almost exactly Broadwell level (actually a tiny bit higher judging by cinebench scores but it's within a margin of error) those chips should really be what gamers look for, not the R7 lineup.

 

5960X and 6900K also come clocked at relatively low frequencies compared to their quad core parts from the same generation, they were never considered for gaming as they cost a RIDICULOUS amount of money, people are looking at R7's gaming performance only because they come in normal prices compared to Intel's offerings.

 

It's also funny cause when people say that AMD is bad in gaming and good in productivity and CPU heavy tasks, they never say that on the other hand Intel is good in gaming and bad in productivity and heavy tasks since you get similar mulithtreaded performance for the money, with a chip that's 2x cheaper (1800X vs 6900K), what's more important you can get the 1800X + GTX 1080 + X370 platform for the price of a 6900K + X99 platform itself... And still have the same productivity potential.

You know how this pro-intel and pro-nvidia forum goes already... any and everything AMD do will get taken out of context... Almost every thread people been saying the 1800x is a terrible gamer when compared to the 7700k but they don't say the same for the 6900k or the 6800k but what do I know about the core performance of 4 core chips to 6 and 8 core chips.

[Memories4K]

36 minutes ago, XenosTech said:

You know how this pro-intel and pro-nvidia forum goes already... any and everything AMD do will get taken out of context... Almost every thread people been saying the 1800x is a terrible gamer when compared to the 7700k but they don't say the same for the 6900k or the 6800k but what do I know about the core performance of 4 core chips to 6 and 8 core chips.

That's because Intel didn't just come out with benches on the 6900k showing how it can do as well, if not better, than their competitors comparable CPU (lol as if AMD could even compete with a 6900k a week ago) in games.

Nobody says the same for the 6900k because people already know the 6900k won't do as well but that the 6900k does get close, whereas the 1800X is lagging FAR behind both chips in gaming; contradictory to the 1800X's performance in other areas outside of gaming compared to the 6900k.

[Valentyn]

German review site Golem.de has found that the motherboards do seem to be a big culprit for performance issues. They need new BIOSs before getting better and more consistent performance.

https://translate.google.co.uk/translate?hl=en&sl=de&u=https://www.golem.de/news/ryzen-7-1800x-im-test-amd-ist-endlich-zurueck-1703-125996-4.html&prev=search
 

Quote

The MSI board was delivered with BIOS version 113, until last Friday a new one appeared.

Version 117, which is still up-to-date, improved speed and stability. If we were still able to count on sporadic Bluescreens with the older UEFI, the board is currently stable. Much more important, however, is the drastically higher performance in games and the real pack with 7-Zip. The release notes include, among other things, a fixed problem with the memory act and its timing as well as the voltage.

 

 

Quote

Compared to the original bios, the new UEFI increases the image rate in our game course between plus 4 and plus 26 percent, on the average even plus 17 percent! In view of this tremendous increase in performance, we had to be certain that our values are correct, and have measured with the Asus boards.This gives us a touch more speed in games than with the updated MSI board.

Apart from the low bios performance with the older bios, an option is also missing for the MSI board, as well as for the Asus boards, to switch off the Simultaneous Multithreading (SMT) of the Ryzen 7 1800X. Since AMD according to its own statement similar to formerly Intel under Core-Parking suffers, SMT reduces the picture rate in some games partly. In turn, SMT has accelerated some applications dramatically, so we would have liked to try out how the double thread count affects the performance.
 

 

[Java]

This article just came out from PC World: http://www.pcworld.com/article/3176053/components-processors/amds-su-says-patches-will-boost-ryzen-gaming-performance-it-will-only-get-better.html 

 

Basically Dr.Lisa Su states that performance will get better as developers patch for the new CPU. With that stated and bios updates on the way, I'd wait to buy Ryzen until everythings (hopefully) sorted out.  Remember, this isn't the first time a new processor ecosystem struggled on day 1. 

[XenosTech]

17 minutes ago, Memories4K said:

That's because Intel didn't just come out with benches on the 6900k showing how it can do as well, if not better, than their competitors comparable CPU (lol as if AMD could even compete with a 6900k a week ago) in games.

Nobody says the same for the 6900k because people already know the 6900k won't do as well but that the 6900k does get close, whereas the 1800X is lagging FAR behind both chips in gaming; contradictory to the 1800X's performance in other areas outside of gaming compared to the 6900k.

Any person with any kind of tech sense knows that 6 and 8 core chips don't perform as well in gaming and shouldn't buy them for gaming. So no matter how much AMD would have tooted that it's a good gamer you as the technical person should have know right off the bat it was bs but AMD didn't market that R7's as gaming chips, they said it was good as good as intels $1000 6900k at rendering and multi-threaded tasks. The only chip they're saying is some what good for gaming is the 1700 but they mostly showed that off for the purpose of a content creator i.e a twitch streamer vs a streamer using a 6700k or a 7700k. Nothing about AMD's demos screamed excellent single core performance which the 7700k will roflstomp it and any 6 or 8 core to oblivion and back without breaking a sweat.

[Memories4K]

11 minutes ago, XenosTech said:

Any person with any kind of tech sense knows that 6 and 8 core chips don't perform as well in gaming and shouldn't buy them for gaming. So no matter how much AMD would have tooted that it's a good gamer you as the technical person should have know right off the bat it was bs but AMD didn't market that R7's as gaming chips, they said it was good as good as intels $1000 6900k at rendering and multi-threaded tasks. The only chip they're saying is some what good for gaming is the 1700 but they mostly showed that off for the purpose of a content creator i.e a twitch streamer vs a streamer using a 6700k or a 7700k. Nothing about AMD's demos screamed excellent single core performance which the 7700k will roflstomp it and any 6 or 8 core to oblivion and back without breaking a sweat.

EXCEPT WHEN AMD BROUGHT OUT SYSTEMS WITH 6900KS AND COMPARING THEM IN DIFFERENT VIDEO GAMES ALONG WITH THE 1800X/1700X.

You're missing my point, the 6900k comes very close to the 7700k and in other situations even pulls ahead by an extremely small margin.
The 1800X comes very close to the 6900k and at times even overtakes it by a small margin in productivity tasks (being half the price for so similar performance is what's drawing people in) but STRANGELY enough, the 1800X falls FAR behind the 6900k in gaming. (Keep in mind, 6900k is Broadwell-E)
The 6900k is also an 8-core/16-thread CPU, it can stay pretty close to the 7700k and at times overtake it, so why is it that suddenly there's this giant gap between the 6900k and the 1800X?
That is what we are saying.

[XenosTech]

1 minute ago, Memories4K said:

EXCEPT WHEN AMD BROUGHT OUT SYSTEMS WITH 6900KS AND COMPARING THEM IN DIFFERENT VIDEO GAMES ALONG WITH THE 1800X/1700X.

You're missing my point, the 6900k comes very close to the 7700k and in other situations even pulls ahead by an extremely small margin.
The 1800X comes very close to the 6900k and at times even overtakes it by a small margin in productivity tasks (being half the price for so similar performance is what's drawing people in) but STRANGELY enough, the 1800X falls FAR behind the 6900k in gaming. (Keep in mind, 6900k is Broadwell-E)
The 6900k is also an 8-core/16-thread CPU, it can stay pretty close to the 7700k and at times overtake it, so why is it that suddenly there's this giant gap between the 6900k and the 1800X?
That is what we are saying.

What games are you talking about ? BF1 at 4k ? Becasue we already know that that was gpu bound and the only game I know that showed was Dota and that was for streaming. The sniper elite didn't really prove anything either because that was at 4k also.

 

New platform with teething issues ? Lots of stuff about zen on a whole seems whacky and everyone is pointing fingers and blaming each other. 

Maybe it's something in the bios that holding it back could be something else, I don't work for AMD so I can't tell you much about what's going on. All I know is that for an 8 core cpu on a completely new architect, AMD did a damn good job of making a cheaper 8 core that holds its own vs intels top of the line in workloads that they're typically used for, on their first try.

[Memories4K]

3 minutes ago, XenosTech said:

What games are you talking about ? BF1 at 4k ? Becasue we already know that that was gpu bound and the only game I know that showed was Dota and that was for streaming. The sniper elite didn't really prove anything either because that was at 4k also.

 

New platform with teething issues ? Lots of stuff about zen on a whole seems whacky and everyone is pointing fingers and blaming each other. 

Maybe it's something in the bios that holding it back could be something else, I don't work for AMD so I can't tell you much about what's going on. All I know is that for an 8 core cpu on a completely new architect, AMD did a damn good job of making a cheaper 8 core that holds its own vs intels top of the line in workloads that they're typically used for, on their first try.

It doesn't matter if it was GPU-bound, the point is that AMD did indeed having gaming marketed and advertised on it's R7 line-up.
To shove it off and change the narrative to "Oh well, these chips were never meant to have gaming in mind, so we can just look over that" i believe is irresponsible.

That's the thing!
That is why it's doing so much worse.
Early X99 stuff had similar issues and reviewers are having conflicting results, there is something wrong here because i don't see why the same CPU that can keep up in this workload with the competitors similar chip to such a degree is now suddenly dramatically behind in this other workload.
I'm not trying to say the R7 line-up is Bulldozer 2.0 because outside of gaming it really is a great, great deal compared to the 6900k.
What i am saying is that the results we've gotten so far in regards to gaming should absolutely be heavily scrutinized and looked at; questions should be asked to why this is happening and either AMD, developers, and/or board manufacturers be held accountable for explaining how it is that the 1800X is so weak in this workload compared to the 6900k where it was doing very well in other workloads.

[XenosTech]

5 minutes ago, Memories4K said:

It doesn't matter if it was GPU-bound, the point is that AMD did indeed having gaming marketed and advertised on it's R7 line-up.
To shove it off and change the narrative to "Oh well, these chips were never meant to have gaming in mind, so we can just look over that" i believe is irresponsible.

That's the thing!
That is why it's doing so much worse.
Early X99 stuff had similar issues and reviewers are having conflicting results, there is something wrong here because i don't see why the same CPU that can keep up in this workload with the competitors similar chip to such a degree is now suddenly dramatically behind in this other workload.
I'm not trying to say the R7 line-up is Bulldozer 2.0 because outside of gaming it really is a great, great deal compared to the 6900k.
What i am saying is that the results we've gotten so far in regards to gaming should absolutely be heavily scrutinized and looked at; questions should be asked to why this is happening and either AMD, developers, and/or board manufacturers be held accountable for explaining how it is that the 1800X is so weak in this workload compared to the 6900k where it was doing very well in other workloads.

Read up someone posted a link with a site saying that it's pointing to a bios problem but I don't know how true that is but we'll have to wait and see. 

 

I didn't care about the gaming performance demoed because it was all done at 4k and at 4k the gpu is the work horse not the cpu so those were pretty much irrelevant as proof that the cpu's were good at gaming, but yes questions have to be asked and people have to be held accountable for this oversight but the r7 by no means is a failure by AMD as people make it out to be.

[Guest Darth Revan]

10 hours ago, Kobz360 said:

The difference this time between Zen and Bulldozer is that Ryzen has shown to have decent single-threaded performance across the board (while I do understand that they are synthetic benchmarks, they are representative of CPU performance in rendering, encoding, decompression etc. otherwise what would be the point of reviewers using these as indicators of performance?), yet for some strange reason, is lacking in gaming applications, not to mention it's memory issues and negative scaling with SMT enabled in quite a few circumstances. If there were (non-gaming) situations showing lower-tier intel chips beating Ryzen 7, while simultaneously being beaten by the 6900K regardless of it's lower clockspeed (i.e. 6900K > i5s and i3s > Ryzen 7), then I would say there's cause for concern. But looking at Anandtech, Tom's, Gamer Nexus and Linus' reviews, it seems Ryzen CPUs itself are great for heavy workstation use, but strangely suffer in gaming scenarios where the 6900K, which it often trades blows with in non-gaming situations, remains extremely effective regardless of it's lower clock speed. 

 

AMD still needs to iron out some of these smaller issues and it seems like it will be done so in the coming weeks and months.  You continue to bring up the real-world, but at current time, the only real-world scenario in which Ryzen seems to be struggling is in gaming. Of course, for many, gaming performance is the most important and I completely understand that, so of course it makes sense that people are not happy. For now, if those people are seeking the best gaming CPU, they can't look any further than a 7700K. 

 

 

There's nothing AMD can do about an architecture that doesn't deliver.

Apart from of course making a new one that actually does deliver.

It's in the hands of software developers to optimize their programs to work better on Zen.

Which no one will do.

The only thing that will be ironed out, will be the memory issue since that's more on the motherboard manufactures than AMD. And when it's fixed it won't magically give Zen a huge boost in performance.

[FratStar]

36 minutes ago, XenosTech said:

Read up someone posted a link with a site saying that it's pointing to a bios problem but I don't know how true that is but we'll have to wait and see. 

 

I didn't care about the gaming performance demoed because it was all done at 4k and at 4k the gpu is the work horse not the cpu so those were pretty much irrelevant as proof that the cpu's were good at gaming, but yes questions have to be asked and people have to be held accountable for this oversight but the r7 by no means is a failure by AMD as people make it out to be.

Do you really expect people to have a proper objective outlook on this? In another thread people were talking about how AMD could have possibly send "golden samples" Since they were chips used in the press builds even though they were taking from the retail pool.

 

Ryzen is pretty much exactly where it should have been performance with with multi threading and even in the single threaded productivity performance it's right where it should be. There are a few benchmarks (non gaming) where it gets its ass handed to it, but for the most part it keeps up with intel's top offerings, Sans-6950X. There are improvements that can be made and features that can come down the pipeline as Zen as whole matures.

 

 

[XenosTech]

2 minutes ago, Darth Revan said:

There's nothing AMD can do about an architecture that doesn't deliver.

Apart from of course making a new one that actually does deliver.

It's in the hands of software developers to optimize their programs to work better on Zen.

Which no one will do.

The only thing that will be ironed out, will be the memory issue since that's more on the motherboard manufactures than AMD. And when it's fixed it won't magically give Zen a huge boost in performance.

There's nothing wrong with the architecture, it does what it was designed to do, improve ipc over the last architecture and it matches the 6900k and all those op hedtp from intel in the respective workloads.