New Zen microarchitecture details

Page 200 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Lovec1990

Member
Feb 6, 2017
88
17
51
On the bright side for consumers, this will force AMD to price 4C/8T lower than most of us expected. While overall performance might be lower, with games targeting 8 threads (thanks to consoles), there are occasions where 4C/4T would do worse in minimum framerates than 4C/8T, even if total throughput is lower, due to various timing issues.

I personally would take a 4C/8T at 3.5GHz over a 4C/4T at 4.5GHz.

But if there are frequency issues with the chip you will not be able to overclock it.
 
  • Like
Reactions: INTEL_SHILL

inf64

Diamond Member
Mar 11, 2011
3,685
3,957
136
On the bright side for consumers, this will force AMD to price 4C/8T lower than most of us expected. While overall performance might be lower, with games targeting 8 threads (thanks to consoles), there are occasions where 4C/4T would do worse in minimum framerates than 4C/8T, even if total throughput is lower, due to various timing issues.

I personally would take a 4C/8T at 3.5GHz over a 4C/4T at 4.5GHz.

Don't forget two important things:
1) 8T Ryzen supposedly has max TDP of 65W
2)8T Ryzen supposedly can clock to 4.2Ghz on solid cooling without any issues. Hopefully even at stock config XFR will take it to similar boost clocks but that remains to be seen.
 
  • Like
Reactions: Drazick

Atari2600

Golden Member
Nov 22, 2016
1,409
1,655
136
From the CanardPC article, it seems AMD is having trouble reaching 4GHz base clocks on the 4C parts - which means that they are failed 8C parts

A flaw does not affect the ability of a part to hit high frequencies.

They are two distinctly different phenomena and should have no link. I've read about possible architectural idiosyncrasies limiting clocks on a cut down chip due to the CCX layout - but am not paying a great deal of heed to them.
 

Cartel

Junior Member
Jun 29, 2012
12
1
71
Had Zero problems with the Asus A7N8X deluxe Rev 2.0 ( yes i remembered that off the top of my head :p )


edit: oh crap was that a dash -E.... i dont remember...lol

but i remember 22x 100 barton 2500XP
ati 9500pro
512 mb ddr400

ahh the good old days :)
my old sig :rolleyes:
ASUS A7N8X2.0-Deluxe 1008mod3-4283Sata AMD Athlon XP 3200+ Barton @2.20 GHz OCZ PC3200 EL Platinum Edition 2GB 2.5-3-2-9-1T@400DDR Dual Channel Enabled ATI Radeon X1650pro 512MB SoundBlaster Live! 5.1


I'll probably wait for the Zen II
The first phenom sucked till the II came out
 
Last edited:
  • Like
Reactions: INTEL_SHILL

bjt2

Senior member
Sep 11, 2016
784
180
86
From the CanardPC article, it seems AMD is having trouble reaching 4GHz base clocks on the 4C parts - which means that they are failed 8C parts (since there is only one die, which is as big as 200mm2, can someone confirm its size?). If so then AMD can't price them too low or else they'll lose money. AMD will have lesser room to maneuver in the sub-250$ range if that is the case, provided Ryzen is competitive enough with 4C Intel parts.

Selling for a low price is better than throw away... You can compensate with higher SKU prices...
 

bjt2

Senior member
Sep 11, 2016
784
180
86
A flaw does not affect the ability of a part to hit high frequencies.

They are two distinctly different phenomena and should have no link. I've read about possible architectural idiosyncrasies limiting clocks on a cut down chip due to the CCX layout - but am not paying a great deal of heed to them.

My hypothesis is that the ES were purposely low clocked, because the ES are used only for testing. They does not need to have final clocks. Only final consumption (at last), and this can be simulated with proper overvolt. I don't think that from a 3.6GHz base 8c they can't produce a 4-4.2GHz base 4c...
 
Last edited:

KTE

Senior member
May 26, 2016
478
130
76
I think they might increase the TDP to allow more silicon to "pass" as the 7740K. If they restricted to 91W, they could only sell chips that hit the higher speeds at the same power consumption, and those are probably rare birds at the moment. Better to up the TDP a little more (they don't even include a HSF in the box so who cares) so that a reasonable number of chips yield than to risk a supply shortage.

It's all about marketing, if they can squeeze another few % and extend the ST performance lead it would paint better picture when reviews are published. Desktop space does not have that many workloads that scale that well with more than 4 cores so having such a highly clocked part would be a plus for intel.
Precisely!

Phenom...

Targeted for QX6700, then Intel launched QX9650.

Game Over.
 

tamz_msc

Diamond Member
Jan 5, 2017
3,708
3,554
136
I'm just making a wild guess here, but it might be because of the way you choose the working cores out of the CCXs. If you are choosing two cores each from both the CCXs, then there is a possibility that the overall distance that a signal has to travel affects the propagation time, which in turn will affect what clock speeds are reachable.

Of course, it could all be false and Ryzen clocks very well even with the 4C chips.
 

Magic Hate Ball

Senior member
Feb 2, 2017
290
250
96
I've heard of this as well. Why do you think this is not the case?

I would say because we haven't had real confirmation on this speculation.

We've heard both ways... that 6c definitely exists, and that 6c is impossible due to the CCX layout.

It's all speculation, as the only parts we've seen are the full 8c/16t demonstrated.

EDIT: I see you were discussing the lower clocked cut-down chips. I have no answer from that, but my point stands regarding all of this potentially being FUD.
 

Vesku

Diamond Member
Aug 25, 2005
3,743
28
86

CatMerc

Golden Member
Jul 16, 2016
1,114
1,149
136
I would say because we haven't had real confirmation on this speculation.

We've heard both ways... that 6c definitely exists, and that 6c is impossible due to the CCX layout.

It's all speculation, as the only parts we've seen are the full 8c/16t demonstrated.

EDIT: I see you were discussing the lower clocked cut-down chips. I have no answer from that, but my point stands regarding all of this potentially being FUD.
Actually the reliable sources that have been leaking Zen information for years have all agreed that 6C is possible.
It's just the tech rumormill websites that spew whatever the other site farts in a circular fashion that claimed 6C is impossible.
 

raghu78

Diamond Member
Aug 23, 2012
4,093
1,475
136
It would be nice if AMD get aggressive on 4C/8T pricing. I think a price of USD 180 would have a disruptive effect. It will make the 7350k and most of the non k core i5 irrelevant. AMD need to hit atleast 4 Ghz turbo for their 4C. They do not want to fall too behind in stock benchmarks even though most people who buy it are going to slap a good air or hybrid cooler and max out the clocks.
 
  • Like
Reactions: inf64

Asterox

Golden Member
May 15, 2012
1,026
1,775
136
Actually the reliable sources that have been leaking Zen information for years have all agreed that 6C is possible.
It's just the tech rumormill websites that spew whatever the other site farts in a circular fashion that claimed 6C is impossible.

Of course it is impossible,:D about as Trump=USA president was impossible six months ago.

As usual the Internet is full of "very smart unemployed CPU designers". :cool:
 

Insert_Nickname

Diamond Member
May 6, 2012
4,971
1,691
136
It would be nice if AMD get aggressive on 4C/8T pricing. I think a price of USD 180 would have a disruptive effect. It will make the 7350k and most of the non k core i5 irrelevant. AMD need to hit atleast 4 Ghz turbo for their 4C. They do not want to fall too behind in stock benchmarks even though most people who buy it are going to slap a good air or hybrid cooler and max out the clocks.

$180-199 seems quite reasonable for a 65W 4GHz 4C/8T Ryzen. Properbly enough to explain the Intel "panic attack"...

I loved my SiS735-based K7S5A, but a lot of people had problems with those thanks to ECS cheaping out on components. Some revisions were apparently better than others. I could never tell . . . mine just worked.

The legendary K7S5A. Those were the days. :)

By the time nForce2 was out, things were getting a lot better in the AMD universe. But that was getting on towards the end of K7's run.

Of course. But Barton was a still a nice chip, especially coupled with one of those chipsets mentioned.
 

KTE

Senior member
May 26, 2016
478
130
76
Yep. He expects Sandy Bridge level int performance and not much on fp.

And the 40% referred to int_rate... A Throughput bench. No different than what I said they might have done, looking at historical data.

"AMD has disclosed little useful performance data, but it provided guidance that Zen’s IPC is 40% better on SPECint_rate2006 than Excavator’s,"

Sent from HTC 10
(Opinions are own)
 

inf64

Diamond Member
Mar 11, 2011
3,685
3,957
136
How can the referred number be from a rate benchmark when chief architect at Hot Chips stated that 40% IPC is single thread improvement and SMT comes on top of that?
Looks like D. Kanter was lazy when he wrote the article. Not to mention that 40% figure is old and will be updated to something else at launch day :)
 
  • Like
Reactions: Drazick

CHADBOGA

Platinum Member
Mar 31, 2009
2,135
832
136
How can the referred number be from a rate benchmark when chief architect at Hot Chips stated that 40% IPC is single thread improvement and SMT comes on top of that?
Looks like D. Kanter was lazy when he wrote the article. Not to mention that 40% figure is old and will be updated to something else at launch day :)
The article was from August last year. o_O
 

Sven_eng

Member
Nov 1, 2016
110
57
61
Yep. He expects Sandy Bridge level int performance and not much on fp.

And the 40% referred to int_rate... A Throughput bench. No different than what I said they might have done, looking at historical data.

"AMD has disclosed little useful performance data, but it provided guidance that Zen’s IPC is 40% better on SPECint_rate2006 than Excavator’s,"

Sent from HTC 10
(Opinions are own)

He also expects 3.2GHz which is at least 12.5% lower than what we know is true and Zen's IPC has been upgraded to 55%, which is 15% more than expected. You can do the maths on that. :)