Official AMD Ryzen Benchmarks, Reviews, Prices, and Discussion

[unseenmorbidity]

Look, if there is some bug causing huge performance problems AMD whould have delayed the launch, they dint. So there may be issues, that does not automatically means that some bios update gonna come and bring some awesome perf increase. AMD provided guidelines to work around the issues and make sure their CPUs performed as best as possible.

Is this your first rodeo or something?

Bugs are a guarantee with the launch of any new architecture. Intel's no different in that regard either. X99 was just as bad, or even worse.

If amd waited until everything was perfect, then we would be waiting until christmas.

 

[Shivansps]

Is this your first rodeo or something?

Bugs are a guarantee with the launch of any new architecture. Intel's no different in that regard either. X99 was just as bad, or even worse.

yes they are, that still does not mean they will cause a performance issue, and this will be a severe one.


it is know that AMD has a history of having mutiple performance problems with their first iteration of a new uarch, fixing it in the 2nd iteration providing for great perf increase, like Phenom 1 to Phenom 2? Bulldozer to Vishera? Llano to Richland? and so on. I tend to belive this is again one of those cases.

 

[tamz_msc]

Look, if there is some bug causing huge performance problems AMD whould have delayed the launch, they dint. So there may be issues, that does not automatically means that some bios update gonna come and bring some awesome perf increase. AMD provided guidelines to work around the issues and make sure their CPUs performed as best as possible.

AMD provided workarounds, not guidelines. There is a difference. One example being using High Performance mode in Power Options in Windows. These, and a lot many other things will be addressed with Windows and BIOS updates.

 

[zinfamous]

What? you are crazy? since when bios update provided big ass cpu performance increase on a regular basics?

Unless there is some big bug over there screwing something really important, no such thing will happen. This is what Ryzen performs in games, just accept it already, R7 Ryzens are not intended for gaming only.

If you want more gaming oriented CPU wait for R3 and R5 Ryzens.

what "this" are we supposed to accept now? Tests are all over the place, with differences in performance of 10% or more in the same game from reviewer to reviewer. Which one of those tells you how Ryzen is performing?

LoL, even in the worst case today, Ryzen still well outperforms expectations of less than a year ago. How soon the cynical forget. Face it: This is what Ryzen already is: the price/performance champion, possibly for the next 4 years while Zen steps into better chips, and Intel is stuck on the same 4 year road that they invested in years ago.

 

[tamz_msc]

yes they are, that still does not mean they will cause a performance issue, and this will be a severe one.


it is know that AMD has a history of having mutiple performance problems with their first iteration of a new uarch, fixing it in the 2nd iteration providing for great perf increase, like Phenom 1 to Phenom 2? Bulldozer to Vishera? Llano to Richland? and so on. I tend to belive this is again one of those cases.

Seems to me that you conveniently forget the same issues Intel faced in the past.

 

[richierich1212]

A video about the Ryzen core-parking issues was just uploaded:

 

[unseenmorbidity]

The JayzTwoCents live linked here: https://www.youtube.com/watch?v=P1dhYDm7SLw

Jay talks about several interesting things in his video:

• For one, AMD screwed up their release and rushed everything. Jay called it a "cluster****" release (46:36) and mobo manufactures were scrambling to get their BIOS ready, not to mention the reviews only had several days to test
• That lead to issues and many reviewers discovered these problems leading to AMD email bast reviewers to type this in the command prompt and redoing all the benchmarks because there might be an 8% performance increase
• AMD also asked reviewers to enable certain things on the AMD side and disable things on the Intel side which (of course) they refused to do
• Microsoft gives priority to Intel and NVidia (edit: in terms of bug fixing in Windows 10 - this is from Jerry on barnacules nerdgasm who seemed to have worked at Microsoft before), not AMD so low level OS optimizations might take a while
• Be prepared for more performance increase as BIOS, OS and game updates comes out!

 

[SketchMaster]

Seems to me that you conveniently forget the same issues Intel faced in the past.

I remember when Intel first came out with HT and it caused worse performance, people were mocking that new tech at the time also.

Once again, I'm typing this on an i7 5820k system. The Ryzen 7 is still a MASSIVE success in my book for the HEDT market. I can buy a Ryzen 1700 for $329, a board for < $150, 32GB of DDR4 RAM and with a little OC can have a system that meets/beats an i7 6900K for fraction of the CPU cost alone.

Why is this a dud?

 

[unseenmorbidity]

BIts And Chips: The benchmarks in all Ryzen reviews shuld be redone. A lot fo new improved BIOS, for all AM4 mainboards, were released just yesterday.

 

[richierich1212]

I remember when Intel first came out with HT and it caused worse performance, people were mocking that new tech at the time also.

Once again, I'm typing this on an i7 5820k system. The Ryzen 7 is still a MASSIVE success in my book for the HEDT market. I can buy a Ryzen 1700 for $329, a board for < $150, 32GB of DDR4 RAM and with a little OC can have a system that meets/beats an i7 6900K for fraction of the CPU cost alone.

Why is this a dud?

It's not. This is a big win for AMD.

 

[.vodka]

Why is this a dud?

Gotta spin it, you know.

The JayzTwoCents live linked here: https://www.youtube.com/watch?v=P1dhYDm7SLw

Jay talks about several interesting things in his video:

• For one, AMD screwed up their release and rushed everything. Jay called it a "cluster****" release (46:36) and mobo manufactures were scrambling to get their BIOS ready, not to mention the reviews only had several days to test
• That lead to issues and many reviewers discovered these problems leading to AMD email bast reviewers to type this in the command prompt and redoing all the benchmarks because there might be an 8% performance increase
• AMD also asked reviewers to enable certain things on the AMD side and disable things on the Intel side which (of course) they refused to do
• Microsoft gives priority to Intel and NVidia (edit: in terms of bug fixing in Windows 10 - this is from Jerry on barnacules nerdgasm who seemed to have worked at Microsoft before), not AMD so low level OS optimizations might take a while
• Be prepared for more performance increase as BIOS, OS and game updates comes out!

BIts And Chips: The benchmarks in all Ryzen reviews shuld be redone. A lot fo new improved BIOS, for all AM4 mainboards, were released just yesterday.

Great to see the reviewers saying how it is, it's been a crappy week for them with daily BIOSes changing data points, everybody reporting different results... Funny, is it a coincidence that the guys with hands on experience with the hardware and direct communication with AMD are reaching the same conclusion we have for good part of this launch day?

Platform is unpolished, green, fresh. Gotta give it time, some people are too impatient these days. AMD doesn't have $13B to throw at a problem annually like Intel... so if you want it good, then it'll take time. I'm fine with that, since the product is solid this time around. If this were another Bulldozer or Phenom I I'd be rooting for people to go and burn down their HQ... thankfully it isn't.

That, and software has to catch up. This isn't another Intel core, it works different, things have to get optimized for it. All in all it's doing an admirable job running code tuned for Intel architectures for the past five years.

AMD decided it was good enough to be launched, and it would seem to be so far. The kinks will be worked out.

 

[ZGR]

4c8t is doomed for gaming in the next few years. Unless you know you will be playing single threaded bound titles (old AI heavy RTS games), you should get the R7 1700. The 1700X and 1800X don't seem worth it in my perspective.

Buying an i7-7700k over an R7 1700 is asking for trouble for high refresh rate monitors in games that support multi threading.

Look at Watch_Dogs 2 and Ghost Recon Wildlands. An i7 4c8t CPU struggles to hit above 120fps where as an i7 6900k does it with ease.

Nearly every new open world title heavily utilizes all 8 threads on my i7.

If you are well aware of the limited lifespan of 4c8t in gaming, but need higher single threaded performance, than Intel is still the way to go; just expect a shorter overall system lifespan with z270.

As far as I can see there are only 2 gaming CPU's that matter right now: The i7 7700k and R7 1700. Both have their own advantages and disadvantages. The R7 1700 is the safer choice financially and long term whereas the i7-7700k is superior in older titles and single threaded scenarios.

If you are a serious gamer, the obvious choice is to combine both builds in an extra large case. Use the 7700k build for older games that stress 1 thread and use the 1700 for newer games that support many threads. That is what I would call an enthusiast build for cheap.

 

[majord]

Did I read somewhere that Samsung modules were best for high memory clocks? Y

 

[airfathaaaaa]

The problem is if SMT in gaming does not get fixed the Ryzen 5 quad cores with SMT will get beaten by a Core i5 7500 in gaming. It also makes the R5 1600X more vulnerable as for optimal gaming you need to switch off SMT and if Intel drop the price of the Core i7 7700K,it will be 4 faster cores with SMT against 6 slower cores.

Intel must be somewhat relieved AMD managed to so monumentally fuck up SMT in games.

Now they are relying on games developers to fix performance - how is that going to work for all games??

Why do AMD like releasing half finished products to the market??

is it amd's fault that the games arent using smt since its NEW? i keep saying this all day long its a 100% software problem that wont get fixed anytime soon if the devs dont actually do something about it..

 

[tential]

You mean 3930K vs 1800x? I'd love to, but I'm not sure they will be worthwhile because I won't build the Ryzen rig until the 1080ti comes in. It would be 1080ti vs 2x 980ti. I can do it that way but it won't be apples to apples. I guess I could do testing at 1080p for you and it won't matter GPU wise.
Another reason I'm keeping this chip is because I seriously had to sit through HELL'S worst traffic to get to microcenter. I'm not giving up so easily after all that. I'll start compiling some benchmarks.

Oh, you meant 1800x 6 vs 8 core. Yeah sure when I get it I'll do that.

Why aren't these benches easy to find? I give you an 8 core cpu right? You want to know how the whole lineup performs and amd had conveniently provided you with tools to easily disable cores.

This is ridiculous.

Anyway these benches all show for the most part that many games don't scale with cores obviously if an i7 7700k is adequate for gaming then why would you try to match it against an 8 core variant from amd? Why not see how a similar processor from amd performs?

Hence why I'm interested in the 6 core and 4 core simulations. If no reviewer has done these then son I am disappointed.

I pray I'm wrong though but if I'm not, moonbogg, you could be the first darling.

 

[R0H1T]

Did I read somewhere that Samsung modules were best for high memory clocks? Y

That was the case with their ULP DDR3, there's not enough data on DDR4 as yet.

 

[moonbogg]

can we get some kind of pet food-eating video from you (wearing a diaper, obviously) if AMD manages to surprise you and fix all of these issues?

Zin, I swear upon all things both holy and not, that if AMD fixes these performance issues I will **eat my Ryzen box** and post proof right here in this thread.

 

[AtenRa]

That's is the bread and butter for AMD. The 4/8? channel 32 threaded Naples will be a high-end beast at a much cheaper price than Intel.

Its not only the cost of the CPU,
lower TDP means lower cooling cost for your servers due to less cooling demand both in cooling infrastructure and electricity = Lower Total Cost Of Ownership
lower Energy Consumption means lower cost for your servers due to lower electricity demand = Lower Total Cost Of Ownership

Server operating energy usage and Cooling infrastructure cost + Cooling electricity cost can be close to 30% or more of the Total Cost Of Ownership. Spending less in those two will decrease the TCO without loosing performance/Throughput vs Intel Servers. Add the lower cost of materials (CPU cost) and ZEN Server SKUs will land as a bomb on the server market in Q2 2017.

 

[coercitiv]

Hence why I'm interested in the 6 core and 4 core simulations. If no reviewer has done these then son I am disappointed.

Give them a chance at least, some barely managed to get the systems working "properly" and had to redo benches multiple times. Let them sleep, feed and poop, they get back on the job and fill the Internets with Zen tests all over again.

That having been said, they better test with 2/4 cores disabled or I'm bringing a torch.

 

[PhonakV30]

How long can Z270 stay ? I heard AM4 Socket can have Zen 2/ Zen 3 ( refer to Dr Lisa on Reddit )

 

[unseenmorbidity]

Ryzen cannot do high fps gaming!?

 

[french toast]

http://www.pcworld.com/article/3176...-or-why-you-should-never-preorder.html?page=2

Ryzen 1700 gets owned by a 3570k in gaming. 5 year old i5 beats it. I have come to the firm conclusion that AMD went for a deceitful preorder cash grab. Lots of disappointed people right now. No two ways about it. I don't think I can live with myself if I install this thing. I might have to take it back.

Arr poor you, Ryzen is being so unfair to little old moonbog, if you are searching for anti ryzen sob story gaming comparisons that highlight the buggy nature of this release , whilst ignoring what has been said about bios,SMT bug, game optimization etc then we cant help you.
Joker productions using a different gigabyte mobo on a better bios somewhat proves this over 10 games, its not like it was a 1or 2 two game cherry picked fluke at 4k?
Why dont you swap your mobo for the gigabyte and get the best of both worlds?
Failing either that or being patient why dont take it back and get a 4 core intel, or pay 1000$ for 6900k, that way you can put this terrible injustice behind you and we can get some peace.

 

[french toast]

First my apologies to the forums and mods for my crying like a baby. I'm done crying (promise) and I'm keeping the CPU. Why? I already know what its like to have an Intel CPU, and to be honest, I kind of appreciate the drama of all this. So, I will look forward to BIOS updates, clever ways to tweak the OS for more performance etc. It might be a slower CPU, but there is this challenge/fun/tweaking factor to this build that the very process of it all is kind of exciting still, so I'm sticking with it. After a year or two if I don't want it, well, then I'll swap it for something else.

Ok i redact my last comment LOL.
Good for you

 

[AtenRa]

Thats the same thing AMD fans said when FX8150 launched, and then PS4 launched, and here we are...

Yeap, FX8150 today is as good or better vs Core i5 2500K in games and still faster in MT workloads

http://www.portvapes.co.uk/?id=Latest-exam-1Z0-876-Dumps&exid=thread...-energy-consumption-in-gaming-atenra.2459963/

 

[looncraz]

So, is there no chance Zen's SMT will be an asset for gaming?

I was thinking of getting the six core,and then upgrading to an 8 core a few years from now.

SMT issues will be mitigated to the point that they are only a 2~5% penalty, instead of 10~20% as they can be now. I think the SMT penalty is being made to appear larger than it is by poor scheduling across the CCXes. Fixing that, by itself, should help reduce the deficit.

I intend to use Process Lasso to force any poorly threaded (and problematic) games/apps onto just one CCX. I suspect that VMWare Player will be just such an app, where I will need to force it to the second CCX. That, though, would also allow me to assign games to the first CCX, so I can then play games while running calculations in VMWare Player (usually in Linux Mint these days, but also Haiku OS).

That's a capability I will lack with any other reasonably priced CPU.