• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

AMD Ryzen Thread: Affordable Core Act

New page, no images.

https://git.kernel.org/cgit/linux/k.../?id=08b259631b5a1d912af4832847b5642f377d9101

x86/CPU/AMD: Fix Zen SMT topology

After:
a33d331761bc ("x86/CPU/AMD: Fix Bulldozer topology")

our SMT scheduling topology for Fam17h systems is broken, because the ThreadId is included in the ApicId when SMT is enabled.

So, without further decoding cpu_core_id is unique for each thread rather than the same for threads on the same core. This didn't affect systems with SMT disabled. Make cpu_core_id be what it is defined to be.

Signed-off-by: Yazen Ghannam <Yazen.Ghannam@amd.com> Signed-off-by: Borislav Petkov <bp@suse.de> Cc: <stable@vger.kernel.org> # 4.9 Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/20170205105022.8705-2-bp@alien8.de Signed-off-by: Ingo Molnar <mingo@kernel.org>
 

StaSeb

Member
Oh wow... I did not know about all that UEFI/GPT-WIndows stuff at all. My last HDD had 320GB, so I never encountered that problem. So I am creating a new boot-USB-Stick now, also I will disconnect the HDD to ensure the System will be installed on the SSD only - it seems to cache stuff to the HDD right now.

Well... at least I learned stuff today!
 

Randam

Member
Wasn't aware of that, but, despite that, there's no guarantee that AMD will deliver on their promises that it'll get better with patches. For all we know, those could end up being just a minor improvement.

If they end up improving alot, that'll be good for everyone, especially since Intel does need competition, but until then, Ryzen can't really be recommended for those focused on gaming based on promises, in my opinion.
None of them?
 

spyshagg

Should not be allowed to breed
·feist·;231542559 said:
ComputerBase tested Zen against Phenom II (alongside some Bulldozer revisions and similar vintage Sandy Bridge and Ivy Bridge)

Ryzen -vs- Phenom II -vs- Piledriver/+ (Bulldozer Gen2/2.5) & Steamroller (Bulldozer Gen3)


Phenom II X6 and AMD Ryzen 7 in comparison [ComputerBase]
https://www.computerbase.de/2017-03/benchmarks-phenom-ii-x6-ryzen-7-vergleich/




Totalrating Applications & Games (720p) ******* Totalrating Applications & Games (Full HD) ******* Total Applications (Windows)



Overall Rating Games (Full HD) ******* Overall Rating Games (720p) ******* SunSpider



Blender ******* 7-zip ******* Adobe Photoshop



3DParticle Movement Benchmark******* Cinebench R15 ******* DBpoweramp Music Converter



PCMark 8



Dolfyn ******* Dolphin CPU benchmark ******* Euler3d CFD Benchmark



FLAC Audio Encoding ******* Octopuses



OpenSSL ******* POV-ray ******* TrueCrypt



Handbrake ******* Power consumption

that is impressive
 
Cumulative gaming performance against similarly clocked 8c/16t Haswell-E and Broadwell-E, on un-optimised and un-patched software with a Windows OS that has yet to receive a Ryzen update.

Average 24/7 max OC on those two Intels is ~10%+/- higher than 8c/16t Ryzen, with the very best samples being ~12-15%+/-.


AMD Ryzen Gaming Performance In-Depth: 16 Games Played at 1080p & 1440p [TechSpot]
http://www.techspot.com/review/1348-amd-ryzen-gaming-performance/

AMD Ryzen processors made a strong impression last week, however a number of technical difficulties and time constraints resulted in more questions than answers when it came to the four games we managed to benchmark in time for launch. As promised, we're back to follow up on our initial 1080p testing with a more in-depth look at Ryzen's gaming performance across a 16 titles played at 1080p and 1440p resolutions.

In addition to including more games, we're also adding results for the 1800X and 1700X with SMT disabled as Anandtech forum-goers have discovered a problem with the Windows 10 scheduler that can cause Ryzen to perform worse in lightly-threaded applications with SMT enabled. Apparently Windows 10 treats all Ryzen threads the same (not identifying SMT from physical cores) and thus the operating system thinks all threads have access to their own L2 and L3 cache when in fact they don't.
settingspok8a.png



Individual game benches at the link, here are the totals:


 

Thraktor

Member
Regarding how broken Windows 10's scheduler for SMT is compared to Windows 7 (very relevant for gaming I guess):
https://forums.anandtech.com/threads/ryzen-strictly-technical.2500572/page-5#post-38773988

Hope Microsoft unbreaks this asap.


Yes. My response was just about ethomaz equating stock turbo speed with OC speed which is wrong as stock turbo speed is per core, OC speed is for all cores, always. And since you want the latter... ;)

That's an insane performance drop from Win7 to Win10. I'd be interested to see someone do a proper set of gaming performance comparisons between Windows 7 and 10 to investigate what kind of real-world effect this might have.
 

kotodama

Member
https://www.youtube.com/watch?v=ylvdSnEbL50&feature=youtu.be

Not sure if this video was posted. Most of what he's kind of goes past me as I don't understand all the terms, but I hope his prediction will follow through.

Cool was just about to post that myself. Interesting singular data point that maybe the assumption that using low res, etc to test for CPU bound situations is just a bad real world test. Not enough data points to be sure, but food for thought mayhaps.

Hmm, if his predictions and that historical data point follow through then there is FineWine Technology in AMD CPUs as well.
 

Steel

Banned
·feist·;231572557 said:
Cumulative gaming performance against similarly clocked 8c/16t Haswell-E and Broadwell-E, on un-optimised and un-patched software with a Windows OS that has yet to receive a Ryzen update.

Average 24/7 max OC on those two Intels is ~10%+/- higher than 8c/16t Ryzen, with the very best samples being ~12-15%+/-.


AMD Ryzen Gaming Performance In-Depth: 16 Games Played at 1080p & 1440p [TechSpot]
http://www.techspot.com/review/1348-amd-ryzen-gaming-performance/






Individual game benches at the link, here are the totals:

So, with the broken scheduler and SMT, across that many games the difference is only about 20% with the 7700k and performs about on par with the 7600k. If the 1600x can also get within 7600k range with the two extra cores even before updates then that'd be a pretty good value.
 

ezodagrom

Member
Cool was just about to post that myself. Interesting singular data point that maybe the assumption that using low res, etc to test for CPU bound situations is just a bad real world test. Not enough data points to be sure, but food for thought mayhaps.

Hmm, if his predictions and that historical data point follow through then there is FineWine Technology in AMD CPUs as well.
He's wrong about using games at low res to test the CPU being bad, the purpose is to test the CPU, using high settings and high resolution would only really test the GPU.

Obviously the results show the performance of the CPUs today, when taking into consideration that this is a new architecture and optimization in patches and/or future games can make a huge difference, today's results may not apply in the future.

But all we have today is promises that it'll get better, we don't actually know how much of a difference those optimizations will make, so, as of today, the 7700K is just a better CPU for gaming overall, be it for games focused on less cores or games that make use of more cores.

The current results are AMD's fault, feels like they rushed the CPUs out too fast (especially when seeing the BIOS issues that reviewers had and the poor memory support).
I wonder if it would have worked out better for AMD if they released Ryzen together with Windows 10's creators update, that is, if they would have worked together with Microsoft to at least fix the scheduler issues that people have been pointing out.

EDIT: Well, they do have a chance to change this first impression when it comes to gaming with the 1600X, since it's a CPU targetted at a more gaming oriented price point. They better have ironed out some of the issues by then.
 
I don't post to Neogaf very often, but I'm finding the FUD about Ryzen pretty sad and would like to help clarify what's happened and what will probably happen.

First off, I'm an engineer; I design roads for a living. A LOT of thought goes into a road design. I'm also a mountain biker. I'll make that relevant below. Engineer's design to constraints and limits with performance targets for the applications to which their product is meant to be used.

If I need to design a road for 70 MPH traffic, that means longer curves, higher cross slopes, longer line of sight distances, etc. A 25 MPH road needs far less. A 5 MPH needs very little geometric consideration at all. A racetrack is a completely different animal. The structures have to be design to handle certain loads--primarily large tractor trailers weighting anywhere from 50,000 lbs to 100,000 lbs. All things are designed with a factor of safety as well.

What's my point? The road has to be designed for the application it is to be used for. Do you supposed CPU's are different?

Consider a mountain bike. It has a gear range that lets me climb the mountain at a snails pace. My bike, in particular, doesn't have very high gearing. I let gravity bring me down at a pace I could never achieve pedaling. Now a road bike, it doesn't have the low gear range of my mountain bike, but it has a high gear range that allows the cyclist to ride much faster.

Now suppose I review a road bike as though I want to use it to ride up a mountain. How would that look? I'd say that $10,000 bike is a piece of junk and not worth anyone's time. Now if I were to review a mountain bike for road cycling, what would that look like? I would say my $4,000 bike is a piece of junk.

The latest reviews of Ryzen using low resolutions is like reviewing a mountain bike on the road. It's not a real use case. It's similar to reviewing a mountain bike by seeing how fast it rolls straight down a mountain with no turns compared to another bike that does so a little bit faster. It's such a narrow use case that is unrealistic that it's surprising anyone would use it for comparison.

Yes, getting the GPU bottleneck out of the way to determine the actual CPU performance is good, using games to do it is bad. They require a GPU and are designed around one, so it's impossible to eliminate a GPU as a factor from a system designed around GPUs. Taking the GPU out of the GPU constrained system is like riding a mountain bike downhill in the lowest gear. All you'll end up doing is spinning out without actually contributing to your speed. Does it mean your bike sucks? Does it mean you have no power to contribute as a rider? No, it means the system isn't meant to work the way you're using it. Ryzen smashes a homerun in nearly all tests that aren't games where there are no GPUs. It isn't much weaker on the IPC front, and it's very solid on the multitasking front. It's not at all a weak CPU and is a complete bargain in the HEDT world.

Would you buy a car that can do 300 MPH when you can only drive it on the street at 75 MPH? If you were in the market for a grunt workhorse of a vehicle and saw that a Ford F-350 1-Ton truck only drives at 90 MPH and a Honda Civic can do 120 MPH, would you buy the Civic? Who do you think would try to convince you that the Civic is better for you? Perhaps the guy who's marketing income is paid for by Honda.

Ryzen is like a good pickup truck, it has all the grunt in the world for real work, but still drives well beyond the posted speed limit of modern GPUs. The Honda is great for people that don't have any use whatsoever for a pickup truck, but it's junk when you want to do real work.

I've never build an AMD system, but Ryzen is making me want one for the video editing I've started to do. I had no intention of getting into video editing a year ago, so buying a weaker CPU just for gaming would have been a mistake. It's as clear as day, the R7 1700 is the best CPU for the money out there today for a broad use case that includes gaming. It's not for every use case, however.

Instead of being dissapointed, be happy as hell that there is an option now. Make an informed decision, and not what the marketing world is telling you to do. If you really want the most bang for your buck, I recommend older corporate machines being sold out of warranty with Ivy Bridge and Haswell CPUs in them. Get the entire system with HDD and RAM for the cost of a new Ryzen. Throw in a low power GPU like an RX480 and you have a real gaming machine for under $500.

TLDR: AMD just did something awesome and you're too blind to know how or why.
 

Ac30

Member
Cool was just about to post that myself. Interesting singular data point that maybe the assumption that using low res, etc to test for CPU bound situations is just a bad real world test. Not enough data points to be sure, but food for thought mayhaps.

Hmm, if his predictions and that historical data point follow through then there is FineWine Technology in AMD CPUs as well.

I got 4 years of FineWine with a 290. Yeah it was awesome that the thing launched to compete with a 780 was equal to a 970 in many ways, but AMD should be aiming for 95+% of possible performances day 1, not 80% and then telling us "think how awesome this will be 2 gens out!". That doesn't sell anywhere near as well. They should have kept it a little longer in the oven (even longer, sigh)
 

ezodagrom

Member
That doesn't make any sense.
It doesn't matter if it's not a real use scenario, the whole point is testing how the different CPUs compare to each others in games.
That's not gonna happen if the GPU is the limiting factor, all you're gonna get is similar results between the different CPUs in that case, so they gotta use powerful GPUs and/or lower resolutions so they can actually test the CPUs.
 
That doesn't make any sense.
It doesn't matter if it's not a real use scenario, the whole point is testing how the different CPUs compare to each others in games.
That's not gonna happen if the GPU is the limiting factor, all you're gonna get is similar results between the different CPUs in that case, so they gotta use powerful GPUs and/or lower resolutions so they can actually test the CPUs.

If you want to actually test CPUs, why use GPUs at all? That's my point.

If you consider a game loop in coding, do you suppose a game designer asks himself how the loop will perform if they take all the time and resources allotted for a GPU out? Or do you supposed the designer knows there's a GPU and he's ok with the performance of his code when he knows the GPU is what's holding the loop up? Do you suppose game companies spend money optimizing a game loop for no GPU?

Yeah, Ryzen's L2 cache or CCX memory latency or something else becomes a bottleneck in some unrealistic gaming scenario. Could a game developer design around that so that it isn't the bottleneck? Yep. Should they? No. Something else becomes the bottleneck. There is always a bottleneck; might as well leave it where it doesn't make a difference. Turns out the unrealistic bottleneck shows up earlier in Ryzen than in Intel CPUs when this unrealistic use case is tested.

If you absolutely have to have the highest FPS regardless of all other considerations, Intel won't fail you there. But the dialogue currently going on is so far from the reality of the situation it irritates me.

Quick question, how does Ryzen compare to Jaguar cores? Can developers make console games work around Jaguar cores or not? If I took GPUs out of consoles would they look like crap? Do people benchmark consoles this way? Consoles are great because they are a closed system and designers can work within the constraints of that system to get the most performance possible out of them. A PC running modern games at 720p with a powerful graphics card isn't a system anyone built anything for intentionally. Designers don't hope for performance from the hardware, they design to the constraints of the hardware (unless they are terrible at engineering). The results there aren't a good indicator of the performance of Ryzen as a CPU. The tests that are good indicators show it to be a great CPU.

The dialogue has been shifted and controlled by the powers that be to make Ryzen look far far worse than it is. Intel can spend more on marketing to buy the mindshare of people like you than AMD could ever dream of. Don't be bought by the stuff you hear being said by people who make a living by being paid by companies that want you to think a certain way.
 

Paragon

Member
If you want to actually test CPUs, why use GPUs at all? That's my point.
Because some of us care how games perform. You can't always infer game performance from other CPU tests.
If you could, Ryzen would be performing much better than it does, as Intel's IPC only appears to be 7% higher.

I would say that the tests could be better though, because it seems that sites either test one extreme or the other: ultra settings at 1440p/4K where you are GPU limited, or &#8804;720p at the lowest settings possible.
What I would prefer would be using ultra settings at the highest resolution which keeps peak GPU load around 80%.
That way your test is not going to be running at 400 FPS, but still avoids being GPU-limited.

It does seem like there is potential for some good performance improvements when Windows' scheduler is updated. There's been some interesting testing done with that, where some tasks were running faster on Ryzen CPUs when they were limited to 4 cores/8 threads on a single CCX, instead of having access to all 8 cores/16 threads.
 

ezodagrom

Member
Because some of us care how games perform. You can't always infer game performance from other CPU tests.
If you could, Ryzen would be performing much better than it does, as Intel's IPC only appears to be 7% higher.

I would say that the tests could be better though, because it seems that sites either test one extreme or the other: ultra settings at 1440p/4K where you are GPU limited, or &#8804;720p at the lowest settings possible.
What I would prefer would be using ultra settings at the highest resolution which keeps peak GPU load around 80%.
That way your test is not going to be running at 400 FPS, but still avoids being GPU-limited.

It does seem like there is potential for some good performance improvements when Windows' scheduler is updated. There's been some interesting testing done with that, where some tasks were running faster on Ryzen CPUs when they were limited to 4 cores/8 threads on a single CCX, instead of having access to all 8 cores/16 threads.
Well, from what I see most reviewers do these tests at 1080p and high or ultra settings, some of them including higher resolution tests as well such as 1440p.
Those that do test at 720p also usually have higher resolution tests.

Personally what I want to see the most is DigitalFoundry's take, their videos on CPUs and GPUs tend to be pretty nice.
 
The point is, that if you want to make Ryzen look bad at gaming, you use settings nobody would to highlight it's disadvantages. Just like game design, marketing is by design. It's not magic that people use odd settings trying to highlight something in a review.

At real world usage settings Ryzen iss not far enough behind Intel chips to negate all the advantages Ryzen offers with large 8 cores and 16 threads over the 4 core 8 thread intel chips. Everyone currently is "OMG Ryzen HAS LOW FPS!!!!!!" When in reality it doesn't; it has high FPS that aren't quite as high as the other brand. At the resolutions and settings people actually use, most people couldn't tell the difference in a blind test. Not quite as fast as Intel, but nowhere near as far behind as people are trying to paint it out to be.

Gamers are competitive, I am one. I like to win. I like to be the best, not the second best. It's easy to use that attitude to make Ryzen look like a failure when it isn't. Gamers are also vocal free marketing if you play them right. People are acting like second place is equivalent to last place in every benchmark when it comes to FPS and AMD. If AMD is slightly ahead in anything, why isn't Intel treated the same way? Again I'm just irritated at the dialogue here.
 
Well, at least this "storm" is making so much people in microarchitecture and processors in general.

Hopefully AMD can fix or mitigate part of this scheduler problem.
 
The point is, that if you want to make Ryzen look bad at gaming, you use settings nobody would to highlight it's disadvantages. Just like game design, marketing is by design. It's not magic that people use odd settings trying to highlight something in a review.

At real world usage settings Ryzen iss not far enough behind Intel chips to negate all the advantages Ryzen offers with large 8 cores and 16 threads over the 4 core 8 thread intel chips. Everyone currently is "OMG Ryzen HAS LOW FPS!!!!!!" When in reality it doesn't; it has high FPS that aren't quite as high as the other brand. At the resolutions and settings people actually use, most people couldn't tell the difference in a blind test. Not quite as fast as Intel, but nowhere near as far behind as people are trying to paint it out to be.

Gamers are competitive, I am one. I like to win. I like to be the best, not the second best. It's easy to use that attitude to make Ryzen look like a failure when it isn't. Gamers are also vocal free marketing if you play them right. People are acting like second place is equivalent to last place in every benchmark when it comes to FPS and AMD. If AMD is slightly ahead in anything, why isn't Intel treated the same way? Again I'm just irritated at the dialogue here.


You realize these CPU tests aren't some new invention right? Gaming CPU comparisons have been done at lower settings to highlight the CPU over the GPU for ages. It's not a newfangled conspiracy against AMD, it's the generally accepted way to benchmark gaming performance of CPUs.

Ryzen is a success for AMD in many ways, you are right. But the long and short is, as of now it is not the best for gaming specifically. This is a gaming forum. So, people here are more interested in gaming performance than in media production or other workloads.
 

Steel

Banned
This thing is dropping like a stone.

I'm out, fellas.

Good luck to the rest of you holding.

Well, I told people not to invest unless you could afford to piss away half of what they invested(That's definitely not going to happen with the CPUs as they are in reality, though). Though, considering that it's nearly back to opening price, I have the feeling you're gonna be kicking yourself later.

The point is, that if you want to make Ryzen look bad at gaming, you use settings nobody would to highlight it's disadvantages. Just like game design, marketing is by design. It's not magic that people use odd settings trying to highlight something in a review.

At real world usage settings Ryzen iss not far enough behind Intel chips to negate all the advantages Ryzen offers with large 8 cores and 16 threads over the 4 core 8 thread intel chips. Everyone currently is "OMG Ryzen HAS LOW FPS!!!!!!" When in reality it doesn't; it has high FPS that aren't quite as high as the other brand. At the resolutions and settings people actually use, most people couldn't tell the difference in a blind test. Not quite as fast as Intel, but nowhere near as far behind as people are trying to paint it out to be.

Gamers are competitive, I am one. I like to win. I like to be the best, not the second best. It's easy to use that attitude to make Ryzen look like a failure when it isn't. Gamers are also vocal free marketing if you play them right. People are acting like second place is equivalent to last place in every benchmark when it comes to FPS and AMD. If AMD is slightly ahead in anything, why isn't Intel treated the same way? Again I'm just irritated at the dialogue here.

This isn't what's going on. People recognize that Ryzen is great for non-gaming applications and reasonably decent for gaming, but if you're buying for gaming it's not a great price/performance at the moment. Given that SMT on windows 7 seems to give a 17% boost in performance and the scheduler issues, that could very well change.

Even with those problems, however, it's still within spitting distance(<10%) of its direct competitor the 6900k in most games at half or less of the price depending on model. On top of that the 1600x is likely to have similiar gaming performance for a much lower price with still great HEDT performance with the extra 2 cores, so that will likely be more disruptive of the gaming side of things.

My point is, the tests aren't the problem here, and there is some overreaction but there's nothing wrong with testing at low resolutions as that shows the actual difference of these cpus that will be shown overtime at higher resolutions and settings.
 

ezodagrom

Member
The point is, that if you want to make Ryzen look bad at gaming, you use settings nobody would to highlight it's disadvantages. Just like game design, marketing is by design. It's not magic that people use odd settings trying to highlight something in a review.

At real world usage settings Ryzen iss not far enough behind Intel chips to negate all the advantages Ryzen offers with large 8 cores and 16 threads over the 4 core 8 thread intel chips. Everyone currently is "OMG Ryzen HAS LOW FPS!!!!!!" When in reality it doesn't; it has high FPS that aren't quite as high as the other brand. At the resolutions and settings people actually use, most people couldn't tell the difference in a blind test. Not quite as fast as Intel, but nowhere near as far behind as people are trying to paint it out to be.

Gamers are competitive, I am one. I like to win. I like to be the best, not the second best. It's easy to use that attitude to make Ryzen look like a failure when it isn't. Gamers are also vocal free marketing if you play them right. People are acting like second place is equivalent to last place in every benchmark when it comes to FPS and AMD. If AMD is slightly ahead in anything, why isn't Intel treated the same way? Again I'm just irritated at the dialogue here.
When reviewing a GPU when it comes to games, strong CPUs should be used to eliminate a possible CPU bottleneck. The goal is to test the GPUs performance, so you can't have the CPU being a limiting factor.

When reviewing a CPU, strong GPUs and low enough resolution/settings should be used to eliminate a possible GPU bottleneck. The goal is to test the CPUs performance, so you can't have the GPU being a limiting factor. Such reviews sometimes include higher resolution tests as well just for the sake of showing it in expected usage scenarios, but extreme resolutions are avoided since there's no point in showing a situation where the majority of the tested CPUs perform the same, due to a GPU bottleneck.

That's all there is to it. Ryzen, as it currently stands, is not delivering on gaming, even in games that use many cores/threads. Sure, in situations where the GPU is a bottleneck, such differences are not noticeable, but people upgrade GPUs more often than they do CPUs, what is currently a GPU bottleneck will eventually shift to the CPU.

It's a new architecture though, if AMD can deliver on their promises of patches for games and/or windows to better support the new CPUs, the situation could change, but there's no guarantee on what kind of results they'll achieve. All the reviewers can do is judge the situation as it is at the moment.

AMD has the chance to change these first impressions with their future releases though, especially the 1600X, since that one is targetted at a more gaming oriented price.
 
You realize these CPU tests aren't some new invention right? Gaming CPU comparisons have been done at lower settings to highlight the CPU over the GPU for ages. It's not a newfangled conspiracy against AMD, it's the generally accepted way to benchmark gaming performance of CPUs.

Ryzen is a success for AMD in many ways, you are right. But the long and short is, as of now it is not the best for gaming specifically. This is a gaming forum. So, people here are more interested in gaming performance than in media production or other workloads.

I've been following CPUs and graphics cards since, well a decade before GPUs were called GPUs. I understand this testing methodology is old and more or less the media standard. Most reviews come out strongly in favor of what Ryzen actually is. I'm complaining here about the dialogue on Neogaf. I like to read Neogaf. I read a lot of different forums to try to get the vibe of the tech community. I know the names of Juanrga, Shintai, Aeassa, Kyle Bennet, etc. The problem I have isn't that it isn't "the best for gaming specifically," but the tone of this thread has gone way off base; second best here is pretty damn good. When you factor in price, this chip is absolutely the bees knees.

As I've been following this thread and want it to continue without the likes of Juanrga, Shintai, and similar ilk, I thought I'd try to steer it back to reality a bit. This is a gaming forum, but it doesn't have to be blind the whole picture. This thread is not only a gaming thread, but a Neogaf thread on a new CPU. People don't just discuss games here. Where is the discussion of what Ryzen does well to balance out the dicussion of it's "complete fail" at gaming? There isn't. There is no discussion here about how Ryzen is currently the only way to go if you might want to stream in the near future.

I'm not sure there aren't paid guerilla posters in Neogaf trying to steer this conversation a certain way. If you've been around for a long time on a bunch of different popular forums, you really get the vibe that those people exist and are ruining open dialogue in the tech world.

Also, I bought a bloody Netburst CPU back in the day instead of an Athlon because I was a much younger media indoctrinated idiot. Trying to make sure that others don't fall prey to the same carnivore.
 

smisk

Member
Still rocking a Phenom II so I'm definitely gonna pick up one of these at some point. Unsure if I should go for a 1700 or wait and see what the six-core CPUs are like in a few months.
 
Even after the OS updates/BIOS updates (mainly the scheduler fix) it's going to close the gap a good amount imo but in games that prefer single threaded performance the 7700k will still come out on top more due to the higher clock speed both out of the box and OC'd.

At the end of the day it all depends on you and if you care about chasing the highest frames or if you willing to give some up for use of more cores/threads.
 

JohnnyFootball

GerAlt-Right. Ciriously.
When reviewing a GPU when it comes to games, strong CPUs should be used to eliminate a possible CPU bottleneck. The goal is to test the GPUs performance, so you can't have the CPU being a limiting factor.

When reviewing a CPU, strong GPUs and low enough resolution/settings should be used to eliminate a possible GPU bottleneck. The goal is to test the CPUs performance, so you can't have the GPU being a limiting factor. Such reviews sometimes include higher resolution tests as well just for the sake of showing it in expected usage scenarios, but extreme resolutions are avoided since there's no point in showing a situation where the majority of the tested CPUs perform the same, due to a GPU bottleneck.

That's all there is to it. Ryzen, as it currently stands, is not delivering on gaming, even in games that use many cores/threads. Sure, in situations where the GPU is a bottleneck, such differences are not noticeable, but people upgrade GPUs more often than they do CPUs, what is currently a GPU bottleneck will eventually shift to the CPU.

It's a new architecture though, if AMD can deliver on their promises of patches for games and/or windows to better support the new CPUs, the situation could change, but there's no guarantee on what kind of results they'll achieve. All the reviewers can do is judge the situation as it is at the moment.

AMD has the chance to change these first impressions with their future releases though, especially the 1600X, since that one is targetted at a more gaming oriented price.

Oh my god.....can there be a worse statement made on Neogaf?

Not delivering on gaming. Are you serious?
 

ezodagrom

Member
I've been following CPUs and graphics cards since, well a decade before GPUs were called GPUs. I understand this testing methodology is old and more or less the media standard. Most reviews come out strongly in favor of what Ryzen actually is. I'm complaining here about the dialogue on Neogaf. I like to read Neogaf. I read a lot of different forums to try to get the vibe of the tech community. I know the names of Juanrga, Shintai, Aeassa, Kyle Bennet, etc. The problem I have isn't that it isn't "the best for gaming specifically," but the tone of this thread has gone way off base; second best here is pretty damn good. When you factor in price, this chip is absolutely the bees knees.

As I've been following this thread and want it to continue without the likes of Juanrga, Shintai, and similar ilk, I thought I'd try to steer it back to reality a bit. This is a gaming forum, but it doesn't have to be blind the whole picture. This thread is not only a gaming thread, but a Neogaf thread on a new CPU. People don't just discuss games here. Where is the discussion of what Ryzen does well to balance out the dicussion of it's "complete fail" at gaming? There isn't. There is no discussion here about how Ryzen is currently the only way to go if you might want to stream in the near future.

I'm not sure there aren't paid guerilla posters in Neogaf trying to steer this conversation a certain way. If you've been around for a long time on a bunch of different popular forums, you really get the vibe that those people exist and are ruining open dialogue in the tech world.
What is there to be said? People have often said that it's a great CPU for productivity.
This topic is in the gaming side of neogaf though, so obviously the focus is...well...gaming.
 
I actually think most of the review sites have been very positive about Ryzen, with 'Must Have' awards and high scores generally.

The problem is the loud majority that own 4-core CPUs that are resistant to the idea of 8-core CPUs being better value. You know in the last 2 weeks I've found out that nearly everyone just games and nothing else at all on their PC, so Ryzen is pointless for them. Funny that...

The problem is not that Ryzen is being tested with set-ups that form a tiny minority of the market (720p/Titan X/1080) to expose perf disadvantages in gaming, it's the forming of a wholesale judgement of Ryzen because of those benches only, which I've seen far too much.

The 1700 is a HEDT competitor that costs the same as a 7700K here in the UK. It's tremendous value for money, and this will be the sentiment when that gets more widely reviewed this week I'm betting.
 

JohnnyFootball

GerAlt-Right. Ciriously.
Even after the OS updates/BIOS updates (mainly the scheduler fix) it's going to close the gap a good amount imo but in games that prefer single threaded performance the 7700k will still come out on top more due to the higher clock speed both out of the box and OC'd.

At the end of the day it all depends on you and if you care about chasing the highest frames or if you willing to give some up for use of more cores/threads.

There is also hope that Ryzen being a success will persuade more developers to push for more core utilization.
 

ethomaz

Banned
If you want to actually test CPUs, why use GPUs at all? That's my point.
Because you need a lot of components to test a CPU... it won't run alone.

Said that the only way to real test a CPU is making sure any of the others components are bootleneck to the CPU... so using the less of the GPU, using the highest RAM clock, fastest SSd, etc are all thinks you need to have a better CPU benchmark.

And to be fair benchmarks needs to be focused in what you will use the CPU too... we are in a gaming forum talking about games on PC... it is expected the most GAFers cares most or only about game performing... so you will see way more GAFers discussing the results for games than everything else.

For games Intel has better and cheaper options than Ryzen... so what advantage Ryzen has for games? It is not a good CPU for games.
 
Oh my god.....can there be a worse statement made on Neogaf?

Not delivering on gaming. Are you serious?
Yeah that's a overstatement by a lot. Ryzen in it's current state is perfectly fine for gaming and will only get better. It just looks worse against Intel chips.

There is also hope that Ryzen being a success will persuade more developers to push for more core utilization.
Well we're already heading towards that path will games like Watch Dogs 2 actually utilizing more than 4 cores if you have more and such. The thing is how long it takes for that to happen. It could be years before we finally see the benefits in terms of optimization in games on a bigger level.
 
"Ryzen, as it currently stands, is not delivering on gaming, even in games that use many cores/threads."

This sort of thing is exactly why I'm agitated.
 
You know in the last 2 weeks I've found out that nearly everyone just games and nothing else at all on their PC, so Ryzen is pointless for them.

Not that my opinion matters at all, but my PC is connected to my TV, and I literally only ever turn it on to play a game. I don't stream, edit videos, code software or anything of the sort. General internet tasks are usually just done on my phone. Believe it or not, we do exist.
 
Because you need a lot of components to test a CPU... it won't run alone.

Said that the only way to real test a CPU is making sure any of the others components are bootleneck to the CPU... so using the less of the GPU, using the highest RAM clock, fastest SSd, etc are all thinks you need to have a better CPU benchmark.

And to be fair benchmarks needs to be focused in what you will use the CPU too... we are in a gaming forum talking about games on PC... it is expected the most GAFers cares most or only about game performing... so you will see way more GAFers discussing the results for games than everything else.

My statement above is that you're not only testing a CPU, you're testing a GAME as well. A game that was never CPU constrained in the first place nor ever designed around a CPU constraint when using modern GPUs. Hell, I benchmarked the hell out of Quake 2 back in the day using a software renderer. It once made sense to bench the way you're describing, but I'm not so certain it applies so well to the modern world anymore and the way hardware has evolved. Go look at the benches for the new DOOM on Ryzen and get back to me on why that game doesn't have the same issues as others. When you can change the software stack and the downside of the CPU changes all over, is it really just a CPU benchmark?
 

ezodagrom

Member
Oh my god.....can there be a worse statement made on Neogaf?

Not delivering on gaming. Are you serious?
Yeah that's a overstatement by a lot. Ryzen in it's current state is perfectly fine for gaming and will only get better. It just looks worse against Intel chips.
It's not all about just performance, don't forget about price, so, yeah, the currently released Ryzen CPUs are not delivering on gaming, since Intel's Kaby Lake CPUs are better and cheaper...when it comes to gaming (I feel that I need to emphasize this, I'm talking from a focus on gaming perspective which doesn't need the best when it comes to productivity).

Will the 1600X do better? Maybe, especially since it'll likely be competing with the 7600K. But it could depend if AMD can deliver on those updates they talk about before then.

Well we're already heading towards that path will games like Watch Dogs 2 actually utilizing more than 4 cores if you have more and such. The thing is how long it takes for that to happen. It could be years before we finally see the benefits in terms of optimization in games on a bigger level.
The problem is when the 7700K matches or surpasses the 1800X on Watch Dogs 2, that's especially why I say it's not delivering on gaming at least as of now.
 

JohnnyFootball

GerAlt-Right. Ciriously.
It's not all about just performance, don't forget about price, so, yeah, the currently released Ryzen CPUs are not delivering on gaming, since Intel's Kaby Lake CPUs are better and cheaper...when it comes to gaming (I feel that I need to emphasize this, I'm talking from a focus on gaming perspective which doesn't need the best when it comes to productivity).

Will the 1600X do better? Maybe, especially since it'll likely be competing with the 7600K. But it could depend if AMD can deliver on those updates they talk about before then.

You see. This is a bit better of a statement.

And yes, if you need a gaming system and need the best of the best, there is no question that the 7700K is the absolute best. And yes, it is cheaper than the 1700X and 1800X.

But lets not pretend that those who got a 1700, 1700X and an 1800X are getting poor gaming.
 

ethomaz

Banned
My statement above is that you're not only testing a CPU, you're testing a GAME as well. A game that was never CPU constrained in the first place nor ever designed around a CPU constraint when using modern GPUs. Hell, I benchmarked the hell out of Quake 2 back in the day using a software renderer. It once made sense to bench the way you're describing, but I'm not so certain it applies so well to the modern world anymore and the way hardware has evolved. Go look at the benches for the new DOOM on Ryzen and get back to me on why that game doesn't have the same issues as others. When you can change the software stack and the downside of the CPU changes all over, is it really just a CPU benchmark?
While the tests are in lower resolutions to see which CPU is better for games even when you are at high resolutions framerate and framepacing can be affect by CPUs... the avg. can be 60fps but the gameplay experience will be better on a stronger CPU.

Said that you are buying a CPU for games... you buy it to hold for at least 4-5 years while you change only the GPU... so you will obvious choose what give more for the money you have in hands.

Actually Ryzen is not a good option for gamers because it delivery less costing more than Intel options.

About Doom most sites didn't tested it but the ones I read the Intel 7700K was a better and cheaper option.
 

ethomaz

Banned
I liked their conclusion aiming for specif market.

The AMD R7 1700 is priced in a way that makes it worthy of consideration at the cost, but with some of the same caveats as the R7 1800X. In the case of the R7 1700, we see mixed workload use cases shine better at the $330 price point, where the CPU deftly outperforms the 7700K in Premiere and Blender rendering tasks when CPU accelerated. If you're pushing renders to the CPU and doing some gaming, it's not a bad buy &#8211; it just depends on how much production you do versus how much gaming, or how much you care about hitting high refresh rates in games. For folks favoring high refresh rate (e.g. 120-144Hz), the 7700K is the best option &#8211; overclocking recommended. Hands down. If you're only gaming, but don't necessarily care about high refresh, we'd still point you toward a more cost effective i5 (or even i7, regardless of -K tag). For a mix of gaming + video editing or animation, the 1700 is actually defensible where the 1800X's price disallowed the same defense. Again, we're considering our core audience, here. If you're not part of that, perhaps consider other reviews.
It is like if you are not core audience (gaming) then it is better to look other reviews.

For gaming they give indeed the best options.
 
What is there to be said? People have often said that it's a great CPU for productivity.
This topic is in the gaming side of neogaf though, so obviously the focus is...well...gaming.

There appears to be software issues which are impacting the performance of the Ryzen CPUs, I strongly believe reviewers should re-test Ryzen when these issues get sorted out.

The 6 and 8 core Ryzen CPUs have a lot of power and will eventually outpace the i7 7700K in games.
More and more of the latest games are taking advantage of processors with 4+ threads and 6+ core processors will soon show sizable advantages over 4 core CPUs, some games already do show this, such as Battlefield 1 and Watch Dogs 2.

Source - PC Games Hardware
 
While the tests are in lower resolutions to see which CPU is better for games even when you are at high resolutions framerate and framepacing can be affect by CPUs... the avg. can be 60fps but the gameplay experience will be better on a stronger CPU.

Said that you are buying a CPU for games... you buy it to hold for at least 4-5 years while you change only the GPU... so you will obvious choose what give more for the money you have in hands.

Actually Ryzen is not a good option for gamers because it delivery less costing more than Intel options.

About Doom most sites didn't tested it but the ones I read the Intel 7700K was a better and cheaper option.


This is not true. It delivers more, but not in pure FPS. Look at minimum frametimes as well as watching netflix while gaming. Or streaming. It delivers more outside the bounds of the review tests. You're not painting a true picture with your statement. How many places tested ten open tabs in chrome on Neogaf while running their benchmarks? I play that way, nobody tested it for me.

Why don't most sites test one of the most acclaimed graphically modern games of 2016? Is there a reason? Hell, ID Software used to be THE software to benchmark with. And all the reviews I read showed Doom performing wonderfully on Ryzen compared to a 7700K.

"Actually Ryzen is not a good option for gamers because it delivery less costing more than Intel options." Consoles my friend, you are correct when talking about consoles. If all you do is game, the PC doesn't have a hand in the game.
 

Papacheeks

Banned
While the tests are in lower resolutions to see which CPU is better for games even when you are at high resolutions framerate and framepacing can be affect by CPUs... the avg. can be 60fps but the gameplay experience will be better on a stronger CPU.

Said that you are buying a CPU for games... you buy it to hold for at least 4-5 years while you change only the GPU... so you will obvious choose what give more for the money you have in hands.

Actually Ryzen is not a good option for gamers because it delivery less costing more than Intel options.

About Doom most sites didn't tested it but the ones I read the Intel 7700K was a better and cheaper option.

It is good for games in the future that will utilize it. As it is right now no one even Microsoft has their software up to date to accommodate the new Architecture and take advantage of it.

As an all around performer for both gaming and production it's a steal.
 
Top Bottom