New Zen microarchitecture details

Page 174 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

coercitiv

Diamond Member
Jan 24, 2014
7,133
16,549
136
Nowhere it is said that 95W are exceeded easily, the term "bien" mean "for good" or "effectively", but is in no way related to a quantification of the excess.
"easily surpass" does not imply "exceedingly surpass". You could always brush up on that English, it may help us communicate better.
 
Reactions: Phynaz

Nothingness

Diamond Member
Jul 3, 2013
3,292
2,357
136
With 128bit fp workload is there any difference in efficiency between a 128 or 256 capable fpu? And for 256bit code?
For a workload running 256-bit vector code, there would be a difference between 128- and 256-bit units as it would require splitting the 256-bit instructions into at least two 128-bit uops; this would consume rename bandwidth and uop space in various places, thus potentially impacting performance. The alternative to split only after issue would require both 128-bit units to be available at the same time which would impact OoO execution.

I hope I made it clear... If not, ask again
 

dacostafilipe

Senior member
Oct 10, 2013
792
274
136
You could always brush up on that English, it may help us communicate better.

Your translation is wrong. "easily surpass" would translate to "dépassent facilement" and not to "dépassent bien".

So, to help improve communication based on non-english information, please use the right translations ...
 

Abwx

Lifer
Apr 2, 2011
11,786
4,695
136
"easily surpass" does not imply "exceedingly surpass". You could always brush up on that English, it may help us communicate better.

The term "easily" would have been "aisement" or "facilement" if translated litteraly, or eventually "significativement" , "largement", the latters needing no translator..

When the term "bien" is used in the middle of the sentence it s to say "for good", if he had placed the term at the end of the sentence he would had said "pour de bon" , wich can be translated liitteraly, and rightly, by "for good", the term is qualifying the statement itself not what is said by the statement, that s a way to say that he has good confirmation that the TDP exceed, but not that it exceed it by a good margin.
 
Last edited:

coercitiv

Diamond Member
Jan 24, 2014
7,133
16,549
136
So, to help improve communication based on non-english information, please use the right translations ...
Well then, can you please help with the right translation and explain the effective difference between yours and mine?

he has good confirmation that the TDP exceed, but not that it exceed it by a good margin.
And I'm telling you again that "easily" does not quantify the margin. If you wanted to correct my translation you could have easily done so without repeatedly implying something I did not write in any form, since nowhere do I state a quantification for the TDP overshoot.
 
Reactions: Phynaz

krumme

Diamond Member
Oct 9, 2009
5,956
1,595
136
For a workload running 256-bit vector code, there would be a difference between 128- and 256-bit units as it would require splitting the 256-bit instructions into at least two 128-bit uops; this would consume rename bandwidth and uop space in various places, thus potentially impacting performance. The alternative to split only after issue would require both 128-bit units to be available at the same time which would impact OoO execution.

I hope I made it clear... If not, ask again

Thanx. I think i understand that
But then what about the power gating of parts of the 256b fpu that dresdenboy mentioned when running 128b code? I mean how do you do that? It seems to me there must be some slack here?
 

Abwx

Lifer
Apr 2, 2011
11,786
4,695
136
Well then, can you please help with the right translation and explain the effective difference between yours and mine?


And I'm telling you again that "easily" does not quantify the margin. If you wanted to correct my translation you could have easily done so without repeatedly implying something I did not write in any form, since nowhere do I state a quantification for the TDP overshoot.

You said that it easily surpass the 95W TDP, as such you did quantify the excess itself and imply that FI it should be closer to say 10W excess than 1W, i dont think that if you were implying 1W you would have used the term "easily", "barely" would come instantly to mind instead...

What he said is that the 95W TDP are surpassed for good, notice that in this translation it can be 1W or 10W excess, absolutely nothing point to a quantification of said excess, only that there is one.
 

LTC8K6

Lifer
Mar 10, 2004
28,520
1,575
126
If we say that a car easily surpasses 150mph, do we generally mean that it barely made it to 151?
 
Reactions: Phynaz

imported_jjj

Senior member
Feb 14, 2009
660
430
136
Easily is quantitative. In this context it implies a substantial margin, the same as "well over" and that was the way some translated the CPC tweet instead of "indeed".
Not that it matters much if it's extrapolated from their results with and ES, an early mobo and BIOS.

I'm more curious about idle power if lowest clocks are 400-550MHz.
 

dahorns

Senior member
Sep 13, 2013
550
83
91
You said that it easily surpass the 95W TDP, as such you did quantify the excess itself and imply that FI it should be closer to say 10W excess than 1W, i dont think that if you were implying 1W you would have used the term "easily", "barely" would come instantly to mind instead...

What he said is that the 95W TDP are surpassed for good, notice that in this translation it can be 1W or 10W excess, absolutely nothing point to a quantification of said excess, only that there is one.

As a native English speaker, I agree that "easily surpass" does imply that the excess was above a de minimis amount. But I also agree with coercitiv that "easily surpass" and "exceedingly surpass" have very different meanings. At any rate, it sounds like the most accurate translation would be that the sample definitely surpassed 95 W.

Y'all can now continue to argue for no reason if you'd like.
 
Reactions: Phynaz

KTE

Senior member
May 26, 2016
478
130
76
The rated TDP is 125W but at that times It was INTEL that was famous to be generous for the TDP and AMD was conservative, if you remember. Anyway that CPUs didn't have turbo, so the TDP was calculated with a power virus, with Tj=125C and Vcore +10% nominal.
If you read what you posted, in a game the CPU peaked at 269W, over 20W more than you posted. And a Game is not a 100% load, since we know that all modern CPUs will turbo in this case. And 2008 games were not famous to be multithreading... So we have 80+35=115. In line with 125W. But ok, link me a review of an 8088 to prove your point...

That's not CPU power, it's system power (read carefully, it's the graphics card).

So your figures are still unexplainable and very wrong.

Sent from HTC 10
(Opinions are own)
 

Atari2600

Golden Member
Nov 22, 2016
1,409
1,655
136
What he said is that the 95W TDP are surpassed for good, notice that in this translation it can be 1W or 10W excess, absolutely nothing point to a quantification of said excess, only that there is one.

While my engineering French is now somewhat rusty (and to be honest, was never that great even when I was out there! ), I'd read it as translating to:

"clearly exceeds 95W TDP"

That is, not by 1 or 2W (which may be due to disputable measurement techniques), but clear enough for it not to be an anomaly or not representative.


However, what revision CPC had for this is not so clear, was it their A0 as used in their preview? Or was it a later ES?
 

coercitiv

Diamond Member
Jan 24, 2014
7,133
16,549
136
You said that it easily surpass the 95W TDP, as such you did quantify the excess itself and imply that FI it should be closer to say 10W excess than 1W, i dont think that if you were implying 1W you would have used the term "easily", "barely" would come instantly to mind instead...
So you would rather have me quantify the value by using "barely" than live with the impression that I might have thought 10W+ instead of 1W+.

Well, I guess it's grammar nazi time:
easily
adverb
  1. without difficulty or effort.
    "he climbed the mountain easily"
    synonyms: effortlessly, comfortably, simply, straightforwardly; More
    with ease, without effort, with no trouble, with no bother, without difficulty, without a hitch, smoothly, on cruise control;
    skilfully, deftly, nimbly, smartly, very well;
    informal no sweat
    "I overcame this problem quite easily in the end"
    antonyms: laboriously
    • more quickly or frequently than is usual.
      "they get bored easily"
  2. without doubt; by far.
    "he was easily the bravest man I've ever met"
    synonyms: undoubtedly, doubtlessly, without doubt, without question, indubitably, indisputably, undeniably, definitely, certainly, assuredly, positively, absolutely, clearly, obviously, patently, simply, surely, by far, far and away, by a mile, beyond the shadow of a doubt;
    informal as sure as eggs is eggs
    "he's easily the best military brain in the country"
    • very probably.
      "the body could easily be that of an actress"

Examples of usage:
  • Zen can easily overclock - means it takes little or no effort to overclock Zen - no quantification
  • Zen can easily overclock past 4Ghz - means Zen will undoubtedly overclock past 4Ghz - no quantification, it can be 4100Mhz or 4900Mhz
  • Zen is easily the best overclocker in PC history - means the overclock margin is so great that no other CPU can match it - quantification time: value is by far greater than any other
Zen can easily exceed 95W TDP means Zen undoubtedly exceeds 95W TDP. It does not mean TDP is 105W, 125W or whatever. It may mean that the margin is great enough to be measured beyond doubt, but that's it.

Again, you could have easily proposed a better translation and that would have been fine, a win for all. But no, you had to come in and claim "Translation is 100% wrong" then proceed with the supposition that I wouldn't have used "easily" to describe low margins.

In other words you based your remarks not on facts, but on impressions. You can easily do better than that, hopefully by a large margin.
 

KTE

Senior member
May 26, 2016
478
130
76
I don't think you know what you are talking about. POWER and CLOCKS are directly linked. They always have been and unless something dramatic changes, as far as silicon is concerned, always will be.

Relax your power requirements and you will obtain better clocks for 99+% of silicon. (Unless you live in a land where this doesn't apply, neither does gravity and HTC10s do indeed have their own opinion on internet forums?)


CPC have indicated they see AMD getting their target clocks. They have said, based on extrapolation from their early (A0?) ES, they didn't see this happening at <95W.

But, AMD have subsequently demo'd on later respin supposedly overvolted parts @3.4 GHz at 95W. Even later steppings are said to have further improved on the clock/power ratios.


edit: For clarity, I don't give 2D for the discussion over relative power consumption of Zen vs. BWE, its too finely balanced to know now for sure how it will go, and it will probably change depending on the instructions requested by each software pack. But I do note, and don't care for, the strawman argument which attempts to say that methods that could be valid now are invalidated due to the results of older generations of CPU - which have very different Cstate(ULT) values.
You replied to a post talking about clocks whereas I was talking about calculating accurate power from the AC.

Not sure what you're going on about after that.

Sent from HTC 10
(Opinions are own)
 

Doom2pro

Senior member
Apr 2, 2016
587
619
106
Somebody leak something already lol... *Bangs head on desk*

Only tidbit today I can find is a new article about Vega on SemiAccurate... I feel like I'm at sea, in the middle of the ocean on a long voyage with nothing to do but listen to shipmates debate about ratios of barnacle size and wave heights.

Desperate times.
 

Atari2600

Golden Member
Nov 22, 2016
1,409
1,655
136
You replied to a post talking about clocks whereas I was talking about calculating accurate power from the AC.

No. You*, were talking about the power envelope of Zen in that post, then drew parallels with AMD's 95W demo at 3.4 GHz with the 3.0 GHz Barcelona rumour from back in the day.

Go re-read your own posts.


*Unless it was actually the HTC10 that posted it - but it did include the disclaimer that the opinion was your own... Maybe it copied it to trick us. Damn!
 

dacostafilipe

Senior member
Oct 10, 2013
792
274
136
Examples of usage:
  • Zen can easily overclock - means it takes little or no effort to overclock Zen - no quantification
  • Zen can easily overclock past 4Ghz - means Zen will undoubtedly overclock past 4Ghz - no quantification, it can be 4100Mhz or 4900Mhz
  • Zen is easily the best overclocker in PC history - means the overclock margin is so great that no other CPU can match it - quantification time: value is by far greater than any other

Ex: "Zen can overclock past 4Ghz" and "Zen can easily overclock past 4Ghz"

For me (!) the first sentence explains a possibility to overclock past 4Ghz where the second is more of the impossibility NOT to reach those 4Ghz.
 

Abwx

Lifer
Apr 2, 2011
11,786
4,695
136
While my engineering French is now somewhat rusty (and to be honest, was never that great even when I was out there! ), I'd read it as translating to:

"clearly exceeds 95W TDP"

I wont go at much further length about this issue, you are qualifying the excess as well, the term is used to confirm an info, not the matter of the info itself, and is generally used when answering to a question or when asked for more precision.

FI if one ask you if a price is more than 150$ whatever the excess the answer will be that "le prix depasse bien 150$" but there s no indication of the amplitude of said excess, it amount to say " i know for good/i m sure that the price exceed 150"...
 

bjt2

Senior member
Sep 11, 2016
784
180
86
That's not CPU power, it's system power (read carefully, it's the graphics card).

So your figures are still unexplainable and very wrong.

Sent from HTC 10
(Opinions are own)

20W for the graphic card? Anyway my considerations still stands: the CPU did not have turbo boost and the AMD's TDP in these days was calculated in worst conditions (Vcore=+10%, Tj=125C, power virus software).

Prove it that they are very wrong. And anyway it draws less than the 6900K so Zen TDP<6900K TDP
 

bjt2

Senior member
Sep 11, 2016
784
180
86
I wont go at much further length about this issue, you are qualifying the excess as well, the term is used to confirm an info, not the matter of the info itself, and is generally used when answering to a question or when asked for more precision.

FI if one ask you if a price is more than 150$ whatever the excess the answer will be that "le prix depasse bien 150$" but there s no indication of the amplitude of said excess, it amount to say " i know for good/i m sure that the price exceed 150"...
Why not ask CanardPC on twitter? I think they will answer...
 

Abwx

Lifer
Apr 2, 2011
11,786
4,695
136
Zen can easily exceed 95W TDP means Zen undoubtedly exceeds 95W TDP. It does not mean TDP is 105W, 125W or whatever. It may mean that the margin is great enough to be measured beyond doubt, but that's it.

Again, you could have easily proposed a better translation and that would have been fine, a win for all. But no, you had to come in and claim "Translation is 100% wrong" then proceed with the supposition that I wouldn't have used "easily" to describe low margins.

In other words you based your remarks not on facts, but on impressions. You can easily do better than that, hopefully by a large margin.

The margin he talk about is the one about the sureness of his affirmation, not the one of the TDP, what he said is that he s 100% sure that the TDP exceed 95W, that s what the term "bien" is related to, that is, the credibility of his saying, that s what is actually qualified by this word.

As to interpretation you are the only one doing one here since i can read french in the text being a french myself, see my explanation above about the use of this word in a close context...

Why not ask CanardPC on twitter? I think they will answer...

No need to for the reasons explained above, i have no problem understanding what he says without any translator, he could even use slangs if he wanted...
 

coercitiv

Diamond Member
Jan 24, 2014
7,133
16,549
136
As to interpretation you are the only one doing one here since i can read french in the text being a french myself
It's not French you're having trouble with.

Lookup the adverb "easily" and see how it's being used - in the context of my translation it means "undoubtedly, doubtlessly, without doubt, without question, indubitably, indisputably, undeniably, definitely, certainly, assuredly, positively, absolutely, clearly, obviously, patently, simply, surely".

The only time "easily" is synonym with "by far, far and away, by a mile, beyond the shadow of a doubt" and is when superlatives are involved: "This spaceship is easily the fastest around. / This is easily the best thread I ever read." Even here the adverb is used to express certainty, but also implies overwhelming proof, exceedingly so.

Then again, by now we're probably gone so far that you'll never admit you were wrong to begin with.
 

Atari2600

Golden Member
Nov 22, 2016
1,409
1,655
136
I wont go at much further length about this issue, you are qualifying the excess as well, the term is used to confirm an info, not the matter of the info itself, and is generally used when answering to a question or when asked for more precision.



Thats what I said.

It clearly exceeds 95W, that is, by a margin and/or consistency that the author is comfortable with stating it clearly exceeds 95W.

Therefore, by simple deduction, that means it exceeds by more than 1-2 W, as such a close margin would leave the author pondering did they obtain a relatively poor quality sample.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |