Ryzen benchmarking and overclocking results

[RejZoR]

My X99 is unable to wake from any kind of sleep. It goes into sleep and when it has to wake up, it just cycles the boot like a retard. From day 1. On platform that should be more than just mature at this point. Should I blame ASUS, Intel or Microsoft, I don't know... The point I'm portraying here is that no one is flawless. And having system that can't use sleep mode at all a bit more important than the fact you can't overclock (already top of the line) CPU.

 

[cadaveca]

My X99 is unable to wake from any kind of sleep. It goes into sleep and when it has to wake up, it just cycles the boot like a retard. From day 1. On platform that should be more than just mature at this point. Should I blame ASUS, Intel or Microsoft, I don't know... The point I'm portraying here is that no one is flawless. And having system that can't use sleep mode at all a bit more important than the fact you can't overclock (already top of the line) CPU.

My own OC'd 6950X/X99.3200 MHz memory sleeps fine. Sounds like an issue with your particular config. Funny, that's a similar problem as to what AMD has... bad review configs.


ROFL

So you're as incapable of building a proper system as AMD's marketing staff is! Excellent!


J/k.

 

[The Lighthouse]

People seem to completely forgot (or not old enough to remember / had no interested in PC at that time) that Intel had massive problems with HT when it was introduced. We were seeing massive up to 20% performance loss due to Intel's new arch.

As well as when Athlon came out. There were massive patches for Athlon processors because they were new.

People said Bulldozer was terrible (due to, again, new arch) and will be obsolete so fast thus better to buy i3 instead. From what I seen at AdoredTV, Bulldozer, at worst, particularly did not lag further behind i5/i7. At best, it now beats 2500k at least. People who bought i3 at that time by following that advice were probably forced to buy yet another new cpus because i3 got completely obsoleted.

No matter how hard you are prepared, there will be problems with the launch, especially when you try to push new things.


Also, I really hope people should stop using games as CPU benchmark or at least take with a grain of salt. They are just really unreliable due to the fact that:
-Processor-specific optimizations.
-Various amount of factors and quality of the testings done by reviewers.


1) Processor-specific optimizations and API issues :
-Those who have any modern CPUs realize they cannot play a lot of old dos games because the games simply ran too fast or glitches. You need either blessings from GOG.com or use programs that make CPU cores slow enough for those games.
-Bethsada's games' physics go wild as soon as you go over 60 fps, GTA 5 begins to lag after a certain fps is exceeded.
-You cannot play Mass Effect 1 well without glitches on modern AMD CPUs, because it automatically forces to use 3D Now when AMD CPU is detected, but no modern AMD CPU supports 3D Now.
-There are several games that results would flip based on what API is used (DX11 vs DX12)

2) Various amount of factors and quality of the testings done by reviewers.
-A tester usually has no clue what 'part' of the game he/she is testing in the first place. Some point it's draw call getting wrecked. Or Physics-related, or that, or that.
One of the examples would be Gamers Nexus's testing of Watch Dog 2. Basically the author tests the game by putting a player character in very empty street, barely moving back and forth for 30 seconds. Repeat that three times. Obviously this does not actually force CPU much because he never tested in-city environment where there are a lot of objects and interactive points moving around. Not to mention driving car also brings several new issues, which are not accounted for Gamers Nexus review. At such environment, it was more of renderer pipeline speed test.
-BF1 Multiplayer's environment always changes because each game is different from another. Some sites test with multiple rounds to get 'average', while others don't.


While productivity programs are not free from this blame, games in particular really suffer optimization issues. In essence, those game reviews tell us that game developers have more time working on Intel CPU rather than AMD CPU.

Now, this may be a valid argument for preferring Intel CPU because it still means that you will get more performance (or it seems) from Intel CPU. But you cannot say this AMD CPU is bad for the gaming because CPU itself is not directly related to the gaming performance to certain degree. Unless it is really lags behind like Bulldozer, it is not appropriate to say Ryzen is sucking at gaming.

I mean, it is also not like Bulldozer that was really far behind everything else. If we forget about the gaming Ryzen really takes top. And for gaming I only see maybe 10% performance loss, at max 15%, for 100+ fps average results. That's far more than acceptable. Not to mention as AdoredTV showed us that it does not mean Ryzen is going to lag further behind either (probably more likely opposite would happen.)


Cores vs single-thread: I still remember when I was getting 2600K for my PC, people told me it was the waste of money since 2500k would give same performance for less money. But these days I don't see that is being true as well. Even during Sandy days the games were slowly moving to multi-core design. And its speed is now insanely accelerated when AMD got a hold of console market and pushed for multiple cores with weak single core performance chips into the consoles.


Finally, after several years, I am looking for building a new PC. I really have no allegiance to any companies. I pretty much flipped on Nvidia and Radeon/AMD for graphics cards and never had bought any AMD CPUs for my own PC because they were not really viable when I was looking for new CPUs.

I will wait for a while. I am getting 1080 TI for my new PC (of course custom cooler design one) so I still have some time left to see how it is going. But unless Intel suddenly give me a nice octa-core processor with decent pricing during that time, it seems I am getting a AMD CPU first time ever since 1994.

 

[RejZoR]

My own OC'd 6950X/X99.3200 MHz memory sleeps fine. Sounds like an issue with your particular config. Funny, that's a similar problem as to what AMD has... bad review configs.


ROFL

So you're as incapable of building a proper system as AMD's marketing staff is! Excellent!


J/k.

So, it's "issue with particular config" and "you don't know how to build a system" when it's Intel, but when it's AMD, ALL are bad because of some people had issues. On how many levels of double standards are you on? 5, maybe 6 ?

 

[cadaveca]

So, it's "issue with particular config" and "you don't know how to build a system" when it's Intel, but when it's AMD, ALL are bad because of some people had issues. On how many levels of double standards are you on? 5, maybe 6 ?

No double standards at all. My point was made, and really, you personally have nothing to do with it.

The fact remains that AMD sent out boxes full of shit to reviewers. My own personal testing has confirmed such; AMD did not choose right memory for review kits. The Ryzen chips need specific memory sticks, single-sided sticks, in order to work properly. Board used doesn't really matter. If they are incapable of appropriately dealing with such a detail, there is NO REASON for anyone to think that they are capable of "fixing" issues they couldn't even deal with themselves for their reviews.

 

[thesmokingman]

No double standards at all. My point was made, and really, you personally have nothing to do with it.

Except you made a personal joke on him.

J/K?

 

[cadaveca]

Except you made a personal joke on him.

J/K?

Very true. That's part of it. He has an issue he cannot identify a fix for. So did AMD, clearly, by what they sent to reviewers. Yet I was able to fix it by simply changing memory kits. How he fixes his issue, by my own system, could be fixed with a "proper" config in a similar way.

So how come these two parties have similar issues? You decide. I wasn't joking at all, but I can see how it may have been perceived as such.

 

[RejZoR]

No double standards at all. My point was made, and really, you personally have nothing to do with it.

The fact remains that AMD sent out boxes full of shit to reviewers. My own personal testing has confirmed such; AMD did not choose right memory for review kits. The Ryzen chips need specific memory sticks, single-sided sticks, in order to work properly. Board used doesn't really matter. If they are incapable of appropriately dealing with such a detail, there is NO REASON for anyone to think that they are capable of "fixing" issues they couldn't even deal with themselves for their reviews.

Do you seriously believe marketing people are experts in system building? Most of them (in all companies) are tie wearing penguins with very basic knowledge of what they are actually selling. That's a harsh reality. It does look sad when any company doesn't double check things to make their product shine when it's sent out to reviewers. I'm not saying you have to cheat and rig things, but at least make sure everything is in check. I'm not going to argue with that, whoever was in charge for the review components should be slapped.

So, we're at a point where the product is not bad, it's just some people doing bad job. That's quite a difference...

 

[cadaveca]

So, we're at a point where the product is not bad, it's just some people doing bad job. That's quite a difference...

It is. To be completely honest, the caveats to getting a decent Ryzen experience are a bit heavy, but there isn't really anything truly wrong with the platform, for sure.

Do you seriously believe marketing people are experts in system building? Most of them (in all companies) are tie wearing penguins with very basic knowledge of what they are actually selling. That's a harsh reality.

That simply shows they have the wrong people doing the job. I am quite angry that AMD failed so hard on this, and I cannot blame any site for any negativity they might have towards Ryzen. I think Summit Ridge is a great platform, but it was not marketed correctly, and its pricing is out of whack due to that marketing.

But what doesn't really change is the performance. It's pretty good already, and it can be improved a bit, but anyone expecting vast differences is being misguided.

 

[TheLostSwede]

Very true. That's part of it. He has an issue he cannot identify a fix for. So did AMD, clearly, by what they sent to reviewers. Yet I was able to fix it by simply changing memory kits. How he fixes his issue, by my own system, could be fixed with a "proper" config in a similar way.

So how come these two parties have similar issues? You decide. I wasn't joking at all, but I can see how it may have been perceived as such.

You know, not everyone's a reviewer and have unlimited access to hardware. I sometimes wish I was still a reviewer, as it sure made life much easier when something didn't work. I can't afford to go out and buy a new motherboard, SSD and RAM to see which part is causing the weird issues I've been having. What is amazing though is that I've had more firmware updates for my SSD than UEFI updates for this POS Gigabyte motherboard.

Do you seriously believe marketing people are experts in system building? Most of them (in all companies) are tie wearing penguins with very basic knowledge of what they are actually selling. That's a harsh reality.

Ok, sorry, but I happen to know the marketing people at AMD personally and they're nothing like this, in fact, most of them are very competent people that have been in the business for a very long time. But yes, I've dealt with utterly useless marketing people over the years too, but AMD and Intel generally don't hire them.

 

[RejZoR]

I mean, it has always been like this. If you want perfect experience you need to STRICTLY follow the QVL list for RAM support. Be it Intel or AMD.

EDIT:
Yeah, it sucks not having the right gear around for testing. If I want to rule something out, I basically have to buy new component. Instead of just swaping with stuff you have laying around...

 

[alucasa]

All this means it's better to wait, for now, for us low Earthings who have limited fund.

 

[cadaveca]

You know, not everyone's a reviewer and have unlimited access to hardware. I sometimes wish I was still a reviewer, as it sure made life much easier when something didn't work. I can't afford to go out and buy a new motherboard, SSD and RAM to see which part is causing the weird issues I've been having. What is amazing though is that I've had more firmware updates for my SSD than UEFI updates for this POS Gigabyte motherboard.

That's what has me a bit angry about this; we could have been given clear guidelines on what hardware to buy, and there might be some users that bought stuff on launch day that are stuck with issue simply because they weren't educated on what to buy. This has been an issue at PC stores since day one; sales staff tend to not always have the right info, and it should be the onus of the hardware makers to educate these people properly. They'd sell more if they did!

Ok, sorry, but I happen to know the marketing people at AMD personally and they're nothing like this, in fact, most of them are very competent people that have been in the business for a very long time. But yes, I've dealt with utterly useless marketing people over the years too, but AMD and Intel generally don't hire them.

So what happened then? Why, when Ryzen is supposed to support 2666 MHz memory, did AMD not assure that every kit sent out was capable of doing so? Why did reviewers have problems with memory? Memory AMD sent them?

All this means it's better to wait, for now, for us low Earthings who have limited fund.

You don't need to wait; you just need to buy the right stuff the first time. I have a system sitting next to me that works great, but yes, as a reviewer, I do have access to a huge number of parts to be able to assemble such, and that's the only reason why I do. I was sent boards and a chip only, so choosing memory was up to me, and I found a kit easily out of the like 12 I have.

 

[thesmokingman]

All this means it's better to wait, for now, for us low Earthings who have limited fund.

Yeap, always been this way. Early adopters have all the fun but they have to deal with the expected teething issues. Part of the problem with Ryzen's release was that many early adopters went into it ignoring the caveats of being an early adopter. Hell, even many reviewers went full you know what and acted like they expected krabby lake levels of integration.

 

[TheLostSwede]

So what happened then? Why, when Ryzen is supposed to support 2666 MHz memory, did AMD not assure that every kit sent out was capable of doing so? Why did reviewers have problems with memory? Memory AMD sent them?

Because it was what they got from their partners for free? That's normally how it works. Both Intel and AMD tend to give out whatever crap they can blag, as they're too cheap to pay for motherboards and RAM. It's stupid, I know, but this isn't the first time review kit has been sent out that's junk. Intel used to send Intel boards (that were junk), but as they don't really make boards any more, it seems that they send out partner boards now, which they get for free.

I remember many moons ago when I worked at PCW in the UK, I got some new motherboard from VIA with support for something like DDR-266 or 333 and got a single memory module from Winchip or some other useless brand like that. It wouldn't run at 400MHz regardless of settings and VIA claimed it should work. Do you think I was happy that I wasted hours upon hours to test that crap?

Or when I got the first PC in for review with a what I think was an Ultra ATA/100 or 133 controller and nothing I did for three days made it work at the right speed. Turned out Promise has shipped it with the wrong firmware...

This really doesn't surprise me at all. At least these days, AMD doesn't have to ship out Biostar and Asus boards without the company logo to try and make sure the motherboard makers avoid the wrath of Intel. Shit, when AMD launched the slot-A Athlon it was a total nightmare. We spent two or three days in the lab with the company that provided the system to get it to run stable, as the BIOS was so damn buggy and the AMD chipset had several issues. I think we had pretty much the same issues that AMD is having now. I had kind of forgotten about that. It was a truly awful launch and I guess the slotted processors were never that popular for AMD. It's really history repeating itself some 18 years later...

"MSI tried this strategy and ended up having to recall a portion of their 6167 (Athlon) motherboards due to stability issues associated with their "modified" design."

"Until the Athlon truly makes its introduction into the server market, we probably won't see more than four DIMM slots on a Slot-A motherboard simply because of DRAM integrity issues, as well as cost. The Fester, like the other two boards tested, had problems with 256MB or registered DIMMs, regardless of size, which limits the maximum memory to 384MB for now"

"The phrase that comes to mind when looking at the BIOS setup of the 7IX is "it works and that's it" which is the closest you can get to completely explaining it's BIOS setup in five words."

"What most people took this as meaning that the SD11 would be the ideal overclocker's board but, it turns out that the "feature" wasn't implemented in the BIOS at the time of production."

"Power supply compatibility issues will be a big factor with the first batch of Athlon motherboards, which is why the boards that are available now are basically engineering samples released to the public to find out what works and what doesn't. "

And so on...

Source: http://www.anandtech.com/show/384/

How quickly we forget... Or maybe I'm just an old fart

 

[the54thvoid]

You don't need to wait; you just need to buy the right stuff the first time. I have a system sitting next to me that works great, but yes, as a reviewer, I do have access to a huge number of parts to be able to assemble such, and that's the only reason why I do. I was sent boards and a chip only, so choosing memory was up to me, and I found a kit easily out of the like 12 I have.

Hi @cadaveca - I'm an anorexic ball hair away from buying a Ryzen chip (1800X). Can you advise me on memory to buy? I know Ryzen prefers single sided DIMM's but obviously e-tailers tend not to say what is what. I'll be using a Thermalright Le Grand Macho (so need to get an AM4 mount) unless I go with the Asus Crosshair board. What would you recommend to me for the easiest Ryzen early adopter, russian roulette experience?

BTW - I'm coming to Canada on vacation in Summer (Vancouver with BC road tour) so be nice or I'll punch your country in the asphalt.

 

[RejZoR]

Hi @cadaveca - I'm an anorexic ball hair away from buying a Ryzen chip (1800X). Can you advise me on memory to buy? I know Ryzen prefers single sided DIMM's but obviously e-tailers tend not to say what is what. I'll be using a Thermalright Le Grand Macho (so need to get an AM4 mount) unless I go with the Asus Crosshair board. What would you recommend to me for the easiest Ryzen early adopter, russian roulette experience?

BTW - I'm coming to Canada on vacation in Summer (Vancouver with BC road tour) so be nice or I'll punch your country in the asphalt.

What motherboard are you planning to buy for it?

 

[the54thvoid]

What motherboard are you planning to buy for it?

Ahem...cough.. ahem - the Asus Crosshair (it is the only AM4 with AM3 compatible mountings). Otherwise I need to contact Thermalright to get the adapter kit. I've heard rumours though that board is terrible....

 

[RejZoR]

Refer to this list and pick the exact same memory kit as listed:
http://dlcdnet.asus.com/pub/ASUS/mb...IR-VI-HERO_DRAM_QVL_forAMDRyzenProcessors.pdf

This is imo the only way to be 100% sure it'll work.

 

[TheLostSwede]

Ahem...cough.. ahem - the Asus Crosshair (it is the only AM4 with AM3 compatible mountings). Otherwise I need to contact Thermalright to get the adapter kit. I've heard rumours though that board is terrible....

Asus has a list of QVL modules here http://dlcdnet.asus.com/pub/ASUS/mb...sors.pdf?_ga=1.75806933.1372906845.1486212314

G.Skill and some Corsair modules seem like a good bet, but I can't find most of the G.Skill modules listed.

 

[cadaveca]

Ahem...cough.. ahem - the Asus Crosshair (it is the only AM4 with AM3 compatible mountings). Otherwise I need to contact Thermalright to get the adapter kit. I've heard rumours though that board is terrible....

Well, what board do we have in this thread? What memory did cdawall use? Each board's BIOS, at this point, will be tuned for specific modules. So what works on one board won't work on another, most likely. QVL suggestions are good, but at the same time, I'd talk to people that have the exact same hardware you plan to buy.


I am currently using ASRock Taichi and two modules form this kit : https://www.techpowerup.com/reviews/GSkill/F4-3200C14Q-32GTZSW/. I was able to simply enable XMP and have 3200 MHz work fully, but XFR is disabled (which is no problem since CPU OC and XFR should not be used together IMHO).

 

[Kanan]

Haven't tested it to be honest.



Intel basically writes the bios for their stuff all the manufacturers do is add a skin, AMD? well they basically tell everyone figure it out. That is why Intel doesn't have issues and amd does.

As someone who has actually had these products (unlike you). They are an unfinished product. This shouldn't be on the market yet. No one is exaggerating when they say they aren't a smart buy for most of the market right now.

Well, I think they will sort out the problems, I'm positive. So Ryzen is rightfully at the market now and can be recommended as well, but always with a grain of salt too.

 

[Grings]

Crucials website already lists some 3466 2x8gb sets as compatible with the crosshair 6 and taichi/fatality pro gaming

 

[notb]

And again, blaming underutilization of cores on AMD, instead on game studios, not making scalable games because they are so used to work with 4 cores thanks to Intel's long domination...

So what? You think all game creators will jump to optimizing for 16 threads, because that's what AMD offers for high-end gaming desktops? Get real.
Game studios have to earn money by selling copies, not race for the best fps possible on high-end desktops.
Most people will still game on 4 cores - either from Intel or AMD (Ryzen 3/5 will be mainstream, Ryzen 7 is for enthusiasts).
And keep in mind we're still talking about desktops, while most PCs are laptops (also among frequent gamers).

Intel manages to squeeze a 65W i7 down to 45W laptop-sensible version (e.g. 6700 -> 6700HQ).

Ryzen 7 has a nominal TDP of 95W, but it's been shown that it can suck 120W in max load.
This means that - to keep 8 cores - AMD would have to cut power draw by half and that would simply obliterate single-thread performane (already not mind-blowing).
So yes, AMD managed to make an excellent flagship 8C/16T CPU for the Ryzen line.
But in the long run, when market becomes saturated with Ryzen CPUs, it's highly probable that most of them (by far) will be 4 core units.

Above all, a CPU design has to be very flexible to work well in many different tasks, in many different types of PCs.
However, a GPU doesn't have to be flexible. It can concentrate on multi-thread performance above anything else.
Why change this status quo? Why force game studios to spend huge money on optimizing games for the minority of >4C owners, when it really doesn't affect the image quality nor FPS?

Sure, if you're very brave, you can blame all the game studios that they don't utilize more cores than 99% of their clients have.
But we have (at least) equally good reasons to blame AMD for not making a CPU that matches the current state gaming market.

 

[Aenra]

I'm almost certain my next CPU will be AMD. Even if they are slightly worse, they are a viable option now. With Bulldozer, even though I wanted AMD, I just couldn't bring myself buying a sub par CPU. But Ryzen is none of that and if they continue optimizing it well in the future, they'll be doing just fine. I also hope AM4 will live as long as AM3+ has.

Funny.. Despite aaaaall my complaints (valid ones mind you), my fondest memories are with AMD chips, lol. There is an element of perversion in there somehwere, granted; but nonetheless.

You know what my number one complaint is? Not my FX9370 and that insane heat, not my elec. bills, not my old Atis, none of the usual. The goddamm drivers, that's what. How many times i updated only to have infinite reboots, bugs, crashes and what not, i don't even remember. If only they'd sort that out, can be really a problem.

Anyway, lol, likewise. While my current rig is faster than i thought possible (straight upgrade from the Fx mentioned above), i already know i won't be going the Intel way again. For a number of reasons unrelated to 4% or 9% performance differences