ZEN ES Benchmark from french hardware Magazine

[bjt2]

There are no Ax steppings, just A0 at different node versions.
Of course it doesn't imply that A0 would or won't be the stepping Ryzen is going to launch with

It seems that there are at least A1, A2 and maybe A3 steppings... Charlie of semiaccurate said that on realworldtech forum...

 

[The Stilt]

It seems that there are at least A1, A2 and maybe A3 steppings... Charlie of semiaccurate said that on realworldtech forum...

If Charlie from Semiaccurate says so, then that must be the case then

 

[krumme]

There are no Ax steppings, just A0 at different node versions.
Of course it doesn't imply that A0 would or won't be the stepping Ryzen is going to launch with

You just tell us if you make a beefed up mb with vrm fit for a tesla s, then we take care of the rest

 

[The Stilt]

There should be plenty of motherboards equipped with beefy VRMs

 

[raghu78]

If Charlie from Semiaccurate says so, then that must be the case then

dude you have been saying so many things over the past year which are being proven wrong one by one. Its really absurd of you to keep making statements as if you know something because you definitely do not have a clue about Zen. Whats even more hilarious is you are ridiculing others who may or may not have factual information. You know what they say about people living in glass houses not throwing stones at others. btw we are talking about a completely new micro-architecture on a brand new process. Its almost assured that there are bugs in the early steppings which are fixed in later revisions. Anyway a word of advice. Drop the condescending attitude and stop pretending that you know something about Zen.

 

[mikk]

- 5GHz air OC. Confirmed.

Lol confirmed. Wake me up if you have a screenshot running Prime95 AVX2 30 min+ on air with all cores enabled from a trustable guy.

 

[Dresdenboy]

http://www.cpchardware.com/cpc-hardware-n31-precisions-elucubrations/

Yeah, for anybody who thinks that it's impossible for AMD to pull the same crap with SMT,it's just a matter of how they state it in the patents.

I think they'll go with the terms used at Hot Chips. What's your point? That a kitchen knife could be used to kill someone?

 

[Dresdenboy]

Lol confirmed. Wake me up if you have a screenshot running Prime95 AVX2 30 min+ on air with all cores enabled from a trustable guy.

That's not what they "promised".

 

[IEC]

From buildzoid on reddit:

I'll just use this comment to say this since it's relevant to me.

I don't have a Zen but I know people who do.

I continue to stand by my comment about 5G not happening on air without too much voltage. This doesn't disprove the statement of ZenOC@Air=5G however I wouldn't get your hopes up because we don't know how many cores or volts actually managed that clock. (my sources never gave me absolute max clock just that it does about X.X for daily use)

For all those wondering about X.X. Well lets just say that I am pleased with X.X.

Source: https://www.reddit.com/r/Amd/commen..._canard_pc_magazine_about_zen_it_has/dbqaxai/

This is the guy who has a 1580/2310 overclocked Radeon 480 (XFX GTR). I anticipate this means that mid 4's are possible with good air cooling. But we will have to bait for wenchmarks™

 

[The Stilt]

dude you have been saying so many things over the past year which are being proven wrong one by one. Its really absurd of you to keep making statements as if you know something because you definitely do not have a clue about Zen. Whats even more hilarious is you are ridiculing others who may or may not have factual information. You know what they say about people living in glass houses not throwing stones at others. btw we are talking about a completely new micro-architecture on a brand new process. Its almost assured that there are bugs in the early steppings which are fixed in later revisions. Anyway a word of advice. Drop the condescending attitude and stop pretending that you know something about Zen.

Was that a question?

EDIT: Let's revisit your statement after the product has been launched

Also could you provide few examples about the things I was wrong about?

 

[moonbogg]

This 5ghz talk sounds kind of ridiculous to me. I expect to see people here with around 4.2-4.3ghz listed in their sigs. I will use two water blocks and stack one on the other to increase cooling. You'll see 4.8ghz in my sig.

 

[F-Rex]

Was that a question?

EDIT: Let's revisit your statement after the product has been launched

Do we also need to wait ryzen launch to revisit your estimated 2.8/3.2 GHz shipping frequencies? LOL

 

[raghu78]

Do we also need to wait ryzen launch to revisit your estimated 2.8/3.2 GHz shipping frequencies? LOL

I am waiting for AMD to demo Turbo Boost at CES 2017 to really rub it in to the people who were making authoritative statements about Zen being a low frequency design.

 

[The Stilt]

Do we also need to wait ryzen launch to revisit your estimated 2.8/3.2 GHz shipping frequencies? LOL

We definitely do, as frequency is subjective

Also if 8C/16T Ryzen ships with 3.4GHz base & 3.7GHz boost, then my worst estimation was off by ~21.5 / 15.6% and my best estimation by 6% / 0%. Considering that these were made >= 9 months prior the launch of the product and there was no reliable information on how the process improved, I cannot say I'm too sorry about the estimations I made. Especially when I haven't seen a more accurate predictions from you or anyone else for that matter.

https://forums.anandtech.com/threads/new-zen-microarchitecture-details.2465645/page-92#post-38366324 - =< 3.6GHz
https://forums.anandtech.com/threads/new-zen-microarchitecture-details.2465645/page-70#post-38269041 - 2.6GHz - 3.0GHz (base) / 3.2GHz - 3.6GHz (boost)
https://forums.anandtech.com/threads/new-zen-microarchitecture-details.2465645/page-51#post-38238198

As I said, frequency is subjective and that's why we need to wait for the proper reviews until the final accuracy of my estimations can be judged

 

[The Stilt]

I am waiting for AMD to demo Turbo Boost at CES 2017 to really rub it in to the people who were making authoritative statements about Zen being a low frequency design.

And what would the "non low-frequency" boost range be in your opinion?

 

[iBoMbY]

I would even go as far an say there will be at least one 8c/16t SKU with 3.45 GHz base frequency.

What interests me most is, how the new dynamic turbo is going to work, and how well it will work with non-linear thread load. In gaming for example, when you could need a much higher frequency at one, or maybe two, cores, and a lower frequency on the others. It would be really nice, if it could clock single cores much higher, while it keeps the frequency reduced at others, even if there is some load.

 

[raghu78]

And what would the "non low-frequency" boost range be in your opinion?

I am waiting for AMD to reveal turbo Boost speeds at CES though I don't think they will want to give Intel too much information well before launch which I think is going to be late March. 4 - 4.2 Ghz max turbo boost is what I would call "non low-frequency". There are indications from multiple sources that Zen Summit Rudge is not having issues clocking above 4 Ghz. I would say 4.5 Ghz average OC on air for 24x7 is now realistic. Anything higher and close to 5 Ghz is bonus and will be disruptive to Intel's current CPU stack.

 

[The Stilt]

I would even go as far an say there will be at least one 8c/16t SKU with 3.45 GHz base frequency.

What interests me most is, how the new dynamic turbo is going to work, and how well it will work with non-linear thread load. In gaming for example, when you could need a much higher frequency at one, or maybe two, cores, and a lower frequency on the others. It would be really nice, if it could clock single cores much higher, while it keeps the frequency reduced at others, even if there is some load.

Shouldn't you be under NDA?

Or perhaps I'm confusing you with another person?

 

[TheELF]

I think they'll go with the terms used at Hot Chips. What's your point? That a kitchen knife could be used to kill someone?

The point is that they have access to a ryzen sample and still try to extrapolate IPC by comparing numbers from MT results.
But then again at least they are saying it outright:

One point that comes up regularly in the comments regarding the performance gain noted " about 35% at equal frequency " against a FX-8370 (as a reminder, we still consider as a Quad Core).

 

[iBoMbY]

Shouldn't you be under NDA?

Or perhaps I'm confusing you with another person?

Maybe I should be under NDA, and I would happily sign one, if AMD would send me one (including their prototypes of course).

Otherwise I'm just good at discovering certain things which should be a secret.

Edit: The last time I had, at least limited, access to prototype PC hardware and software, under some kind of NDA, was over 8 years ago, unfortunately.

 

[.vodka]

Shouldn't you be under NDA?

Risking his life to the NDA police and the green/red vans. I like it.

-----------------------

Well, Zen pretty much has the raw performance required and can clock over 4GHz without much of a problem. I'd love to see what the final picture will be for properly beefy motherboards with VRMs up to the task!

Buildzoid hinted at X.X clocks for the 8c16t part, saying he's comfortable with that. X=X? 4.4GHz? Seems reasonable since every extra 100Mhz in that area gets exponentially more difficult to achieve, seeing what a single core can do on a flaky motherboard.

Now, pricing is next. I like the part about an overclocking demo at CES if everything goes well.

 

[CatMerc]

https://twitter.com/BitsAndChipsEng/status/814837806933692416

 

[raghu78]

Risking his life to the NDA police and the green/red vans. I like it.

-----------------------

Well, Zen pretty much has the raw performance required and can clock over 4GHz without much of a problem. I'd love to see what the final picture will be for properly beefy motherboards with VRMs up to the task!

Buildzoid hinted at X.X clocks for the 8c16t part, saying he's comfortable with that. X=X? 4.4GHz? Seems reasonable since every extra 100Mhz in that area gets exponentially more difficult to achieve, seeing what a single core can do on a flaky motherboard.

Now, pricing is next. I like the part about an overclocking demo at CES if everything goes well.

You have to remember nobody outside of AMD has the final production stepping which is rumoured to be A3. Fottemberg's teasers on clocks being better than expected on final revision make it quite difficult to say anything right now. Its all down to final retail chips and their reviews.

https://twitter.com/BitsAndChipsEng/with_replies

2017 is going to be a good year for the desktop CPU market which will see a return to competition unseen in a decade.

 

[.vodka]

You have to remember nobody outside of AMD has the final production stepping which is rumoured to be A3. Fottemberg's teasers on clocks being better than expected on final revision make it quite difficult to say anything right now. Its all down to final retail chips and their reviews.

https://twitter.com/BitsAndChipsEng/with_replies

2017 is going to be a good year for the desktop CPU market which will see a return to competition unseen in a decade.

Sure. Only AMD knows what's up as they continue ironing out the last remaining bugs and issues. Fottemberg has been so far right on his comments on Zen, his sources are good.

It's certainly looking good gathering all we know so far... which has been done on early samples. It's bound to get better. I'm beginning to like 2017. We consumers finally get a win!

 

[The Stilt]

You have to remember nobody outside of AMD has the final production stepping which is rumoured to be A3. Fottemberg's teasers on clocks being better than expected on final revision make it quite difficult to say anything right now. Its all down to final retail chips and their reviews.

https://twitter.com/BitsAndChipsEng/with_replies

2017 is going to be a good year for the desktop CPU market which will see a return to competition unseen in a decade.

As I said before, there are no Ax parts. Just A0x. Therefore you should advance in the alphabet
If someone says there is, he is either lying or being lied to.