• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

AMD Zen Performance Preview

oh a half quote from some random dude on reddit, ok... heres the rest of the quote too:

"Not too surprising since its $1000+ cpu and is also already out on the market. If the leaked pricing is correct then the ryzen seems like it will be an amazing processor ready to take on Intel and knock them off their enthusiast cpu throne"

I guess you didn't like that part or something...
 
This site is awsome. Been a reader for a few years now (not active in forums). I really appreciate the scrutiny and the willingnes to dig and get the truth out there. I'm glad AMD wasn't full if shit (yet?). Thanks for working on this Kyle!
 
pendragon1 said:
oh a half quote from some random dude on reddit, ok... heres the rest of the quote too:

"Not too surprising since its $1000+ cpu and is also already out on the market. If the leaked pricing is correct then the ryzen seems like it will be an amazing processor ready to take on Intel and knock them off their enthusiast cpu throne"

I guess you didn't like that part or something...
Click to expand...

Because its irrelevant to the compare done. Just as I doubt you wish to talk about the power deltas when doing the Blender test.

What did AMD do with the FX launch? Take a wild guess what they compared to. Or the Polaris idiocy with CF vs a single card.

You should ask why Ryzen couldn't do better in BF1. Because relatively cheap CPUs already do.

Perhaps you will find out that AMDs PR haven't changed the last 10 years. :)
 
pendragon1 said:
quote box but no source? youre one of the one always yellin about "wheres your source/proof?" so...
Click to expand...


There is just no way anyone can take AMD's word on just one benchmark man, too many times they have screwed us, I bought an AM3 motherboard, prior to BD launches thinking it was going to killer based on what that AMD employee stated. Yep 300 bucks down the toilet, sold it on ebay.....

And BF1 was a GPU bound benchmark, don't even need anyone telling anyone anything, it is common understanding, the only single card that can get over 60 fps on BF1 is Titan X p (4k)
 
I really hope we have something good. Tired of 6 core boringness from Intel to be honest. I'm too cheap to go 8-10 core Intel. Almost nabbed a 5960x earlier this year though.

This 6800k was meh for the $430 I paid for it. I should have kept my 5820k.
 
This event was a soft launch of its upcoming desktop CPU that carries the architecture codename "Zen
Click to expand...

Kyle, did AMD make the CPU available to anyone outside their offices yet?

I walked away that this event was more a sneak peak of 'almost final' silicon. In order to keep AMD in peoples minds moving into the next year. I do not think anyone in a select audience has had the privilege to get their hands on one yet benchmark. If they did, well then I say we go get some hard hitting blokes, and go to town on that somebody.
 
It sure would be nice to see competition in the CPU market again. I'm a little nervous that AMD hasn't shown us any single core benchmarks (They did the same thing with Bulldozer, we found out because it's IPC was garbage.). The other thing that bothers me is this revelation that the Intel system was capable of quad-channel but was only running in dual-channel. I understand it doesn't affect that benchmark, but if Ryzen only has Dual-Channel; I'm concerned about feeding the beast in many real life scenarios. 8 cores and 16 threads is a lot of compute power seems like it would need a lot of memory bandwidth to make use of all those cores and threads. I suppose if it is clock for clock competitive with Intel and serves up at least 4 cores 8 threads, it only needs to be priced properly to shake things up a bit in the mainstream to low-end power user market. I'm thinking AMD serves up a 4C/8T with a price perhaps closer to 4C/4T i5-6600($200-$250) at the same clocks? Then the 8C/16T at the typical $300-350 price point of the 4C/8T i7-7700/6700 again similar clocks, maybe slightly lower.
 
iQuasarLV said:
Kyle, did AMD make the CPU available to anyone outside their offices yet?

I walked away that this event was more a sneak peak of 'almost final' silicon. In order to keep AMD in peoples minds moving into the next year. I do not think anyone in a select audience has had the privilege to get their hands on one yet benchmark. If they did, well then I say we go get some hard hitting blokes, and go to town on that somebody.
Click to expand...


Rumors, without even screenshots, and I'm questioning why a french retailer would be in charge of installing a rig for CES which is in the U.S (unless LDLC is showing up to exhibit?) but if retailers are intended to present rigs at CES, it would be a decent indicator for a CES launch.

Edit: I can't find anything on LDLC showing up at CES.
 
Last edited:
OrangeKhrush said:
As an AMD fan I am somewhat hesittent to outright committ to IPC as Blender tests SMT more than IPC
Click to expand...

I'm getting confused in my old age. I thought SMT meant simultaneous multi-threading and IPC meant instructions per clock. SMT's sole purpose is to increase IPC by utilizing dormant hardware for another process's thread.
 
You guys are great for holding their benches to the test.

So to sum it up.... things are going exactly as I expect they will.

The chip will perfrom on par with a year old Intel chip... which is fine because Intel isn't pushing performance anymore clearly unless they have something sitting in a back room waiting on AMD. If they do screw them all the more, that would be a good enough reason to support the AMD engineers for me.

As for the AMD marketing and general salesmanship... ya they suck as they have for a long time. We all know that.

As a AMD booster I was happy with Zen. I thought cool a name that isn't completely stupid. All well trust their marketing geniuses to screw that up to. I hope the AMD marketing dept makes it possible for us to buy "Xtreme Ryzen 9000s" when it launches.

Looking forward to being able to ignore all the AMD fud and instead read some proper real world benchmarks on sites like this one.
 
Seriously good reporting on all this. No other tech sites are even close to getting any real data up from the test that I've seen, so this is huge. Glad the test wasn't all fluff, too. Wonder if we're going to get back to the early 2000s when AMD and Intel really traded blows and there was legitimate difference enough between the two designs that it made a realistic difference in your end results depending what you were trying to do with the chip.

Haven't bought a top-line AMD CPU since Phenom II so can't wait to get back into team Red.
 
ChadD said:
You guys are great for holding their benches to the test.

So to sum it up.... things are going exactly as I expect they will.

The chip will perfrom on par with a year old Intel chip... which is fine because Intel isn't pushing performance anymore clearly unless they have something sitting in a back room waiting on AMD. If they do screw them all the more, that would be a good enough reason to support the AMD engineers for me.

As for the AMD marketing and general salesmanship... ya they suck as they have for a long time. We all know that.

As a AMD booster I was happy with Zen. I thought cool a name that isn't completely stupid. All well trust their marketing geniuses to screw that up to. I hope the AMD marketing dept makes it possible for us to buy "Xtreme Ryzen 9000s" when it launches.

Looking forward to being able to ignore all the AMD fud and instead read some proper real world benchmarks on sites like this one.
Click to expand...
AMD - A1337 Ryzen w/Radeon Instinct & TruAudio technology...run m*thaf*cka run the AI take over has begun!
 
wouldn't the real test be something like prime95 32M to test IPC, wonder why AMD didn't do something like that.
 
@Kyle - +++ props for putting in the work to get this info to us, the thoroughness of the work you guys do is what has kept me coming here for so many years.

Even if Ryzen isn't competitive on high-end desktop, if it hits a good price/performance point and makes AMD some money then that's a win for all of us. Intel has been pretty much running their own program for years with no real motivation to progress or drop prices, a healthy AMD corp can push both those buttons and that just makes things more interesting. I personally haven't had a reason to upgrade in several years and realistically I still don't, but that doesn't mean I won't spend $$ on something if it grabs my attention enough.
 
Awesome news if it all pans out. Consumers (or at least enthusiasts) will have a choice and if Intel really has been holding back (as opposed to hitting a wall), then they'll up their game too. Either way, this sounds like a win for consumers. It's been a long time (10 years?).
 
Dannotech said:
Here is my deep dive into the IPC of Blender and extrapolation of Ryzen IPC for anyone who is curious.

Click to expand...



Interesting sounds like a pretty good way to analyze IPC of Zen on a per application basis, lets see if it pans out.
 
THUMPer said:
Nice, but I can't take anyone seriously who still has the Cortana search box enabled. :dead::dead:
Click to expand...

I used to feel the same way.. but once I gave it a chance, it works great for quick finding stuff that is used often that I don't have and don't want a shortcut to.
 
razor1 said:
Interesting sounds like a pretty good way to analyze IPC of Zen on a per application basis, lets see if it pans out.
Click to expand...

Yes, this is the intent. I'm also going to do a video comparing the IPC in several applications on multiple generatoins of Intel Core i7's and may even include a Core 2 Quad in the demonstration =)

The only core I don't have is a Broadwell-E but I might be able to get access to one at the office. We'll see.
 
well if it works out between the other cores over multiple programs, I think you will be on to something :)
 
THUMPer said:
Nice, but I can't take anyone seriously who still has the Cortana search box enabled. :dead::dead:
Click to expand...

Yea I get the same kind of treatment at work because I carry around a Windows phone. The bullying is pretty funny, actually, its one of the reasons why I keep it. :)

But on a serious note, you should take me seriously. I've been programming x86 in C++ and assembly for 20 years and my optimizations are in at least one of the game you've played. :)
 
The important question is...how does *anyone* at AMD think "Ryzen" is a better name than Zen? Or was this a case of being unable to trademark "Zen" for some reason?
 
Dannotech said:
Yes, this is the intent. I'm also going to do a video comparing the IPC in several applications on multiple generatoins of Intel Core i7's and may even include a Core 2 Quad in the demonstration =)

The only core I don't have is a Broadwell-E but I might be able to get access to one at the office. We'll see.
Click to expand...

What's the IPC if you disable hyperthreading? My thought is Blender is ideally suited to hyper threading/SMT so could see a large boost in that. Ergo, Ryzen could be really good in multithreaded environments and still be lackluster in single thread even though they technically look to be even with 8C/16T vs 8C/16T Intel. If the Intel HT doesn't give a true 2x boost in Blender but AMD can get close to that, it'd look even in something like this bench but still be a fair bit behind in single threaded apps.

I want AMD to compete fully (see sig) as that's what I want to upgrade to, but the nature of this benchmark makes me a little skeptical that they can in all conditions. Granted, single threaded performance isn't as important as it was, but a lot of games are still not very multithreaded (and certain functions can't be) so it could hurt it in certain circumstances if it's not also close in single threaded conditions.
 
bobzdar said:
What's the IPC if you disable hyperthreading? My thought is Blender is ideally suited to hyper threading/SMT so could see a large boost in that. Ergo, Ryzen could be really good in multithreaded environments and still be lackluster in single thread even though they technically look to be even with 8C/16T vs 8C/16T Intel. If the Intel HT doesn't give a true 2x boost in Blender but AMD can get close to that, it'd look even in something like this bench but still be a fair bit behind in single threaded apps.

I want AMD to compete fully (see sig) as that's what I want to upgrade to, but the nature of this benchmark makes me a little skeptical that they can in all conditions. Granted, single threaded performance isn't as important as it was, but a lot of games are still not very multithreaded (and certain functions can't be) so it could hurt it in certain circumstances if it's not also close in single threaded conditions.
Click to expand...

Hmm I'm interested. I'll run the benchmark again with hyperthreading disabled and post an update to this thread after I get home from work.
 
I think my rig can hold out until these get out into the wild... I'm also curious about this vega stuff when planning my next rig...
 
Palladium@SG said:
Matching Intel HEDT per core is already destroying Intel. AMD's current CPUs are in such shambles that they have nothing to lose by previewing samples.
Click to expand...
Agreed. I am pretty sure Intel can pull off faster CPUs than they have on the market, similarly to what they did with Coppermine and were holding well till Prescott fiasco.

As a very basic thing, about half of 6700K die is GPU, so Intel certainly can release Kaby Lake 8-core for 170/270 platform with same die size = cost. Or release effectively OC'ed models. But such a move will severly eat into their margins as they won't be able to milk the market as much. Not a good news for a company that is struggling to keep its profits level, nor the shareholders.
 
THUMPer said:
wyCOMN5.jpg


Kyle didn't you mention AMD should spell check their slides?
Click to expand...

lets be honest here, kyle was just pointing out that having the Z in the name was stupid and that it should actually be Rysen. ;)
 
THUMPer said:
wyCOMN5.jpg


Kyle didn't you mention AMD should spell check their slides?
Click to expand...
Twitter was pulling from a cached page that was not published for some reason. Have not figured that out yet. We have moved our entire infrastructure in the last couple months and still working out bugs.
 
Dannotech said:
Here is my deep dive into the IPC of Blender and extrapolation of Ryzen IPC for anyone who is curious.

Click to expand...


Out of curiosity how does is your program getting MIPS and IPC from the processor? Is there some standard API call to get this somewhere?
 
There's one rather massive and important question that is unanswered: Performance when application makes use of AVX/AVX2 instructions. It should be noticeably slower than on Intel (according to The Stilt). For games that doesn't matter at the moment though.

On the other hand that's why I think AMD will price the 8 core version around 350 €.
 
Dannotech said:
Here is my deep dive into the IPC of Blender and extrapolation of Ryzen IPC for anyone who is curious.

Click to expand...

Correct me if I am wrong, but doesn't the CPU you used in your test have 12 cores, 24 threads? I am asking because in your analysis you used Ryzen's number of threads as 8, while it should have probably been 16.

Also the fact that Ryzen 16 thread CPU matches a similarly clocked 6900k which also has 16 threads would mean that at least in this benchmark they have similar IPC per core.
 
You must log in or register to reply here.
Back
Top