Speculation: Ryzen 4000 series/Zen 3

Page 207 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Veradun

Senior member
Jul 29, 2016
564
780
136
Yeah, that to me is probably the most surprising info from the event: After all these years teasing 7nm+ for Zen 3 it turned out to be some node optimizations by AMD itself, not a separate node by TSMC, and the Zen 2 XT chips are actually using those as well.
N7+ was dead the very moment N6 spawned to life.

Needing a complete redesign vs just new masks for minimal gains has been the full array of nails in the coffin.
 

moinmoin

Diamond Member
Jun 1, 2017
4,933
7,619
136
N7+ was dead the very moment N6 spawned to life.

Needing a complete redesign vs just new masks for minimal gains has been the full array of nails in the coffin.
Well, nobody ever publicly stated that 7nm+ would refer to N7+, AMD could have moved to N6 instead optimizing the currently used N7.
 

Kedas

Senior member
Dec 6, 2018
355
339
136
They said they got al the improvements from architectural design so it is 7nm.... Zen4 will be a process change. Besides 7nm will be getting cheaper and Zen 3 as last CPU with DDR4 has a very long life time to go.
AM5 DDR5 will not be for the budget market for a long time.
 

Martimus

Diamond Member
Apr 24, 2007
4,488
152
106
Yeah, that to me is probably the most surprising info from the event: After all these years teasing 7nm+ for Zen 3 it turned out to be some node optimizations by AMD itself, not a separate node by TSMC, and the Zen 2 XT chips are actually using those as well.
It never made sense for them to change the process node, even to n7+ or n7p. They had too many improvements in the base design and too little time between launches to do that. Plus the cost of designing and implementing the new tooling, especially the new EUV tooling, when they already have tooling and design processes for the n7 process they are using on all their other projects. It never really made sense, but I can understand why the thought of new processes would excite some people enough to hope for those changes.
 
  • Like
Reactions: Elfear and Tlh97

moinmoin

Diamond Member
Jun 1, 2017
4,933
7,619
136
It never made sense for them to change the process node, even to n7+ or n7p. They had too many improvements in the base design and too little time between launches to do that. Plus the cost of designing and implementing the new tooling, especially the new EUV tooling, when they already have tooling and design processes for the n7 process they are using on all their other projects. It never really made sense, but I can understand why the thought of new processes would excite some people enough to hope for those changes.
Well, with this Zen 3 is the first Zen gen not to change the node even though AMD did indicate earlier that a change would happen. Even Zen+ changed the node, even if it reused the mask unchanged and only had microcode changes on top of that.
 
  • Like
Reactions: Tlh97

ThatBuzzkiller

Golden Member
Nov 14, 2014
1,120
260
136
What is something better? No idea really; I just have heard Linus Torvalds opinion and I also think that trying to make a cpu act like a gpu is probably a bad idea. It seems like Intel's kludge before they finally gave up and decided to design a gpu. Also, if they were going to support it, I would have expected it with the new architecture (Zen 3). I guess it may be plausible that they are waiting for Zen 4 (5 nm and chip stacking to deliver more bandwidth).

So you basically just admitted to making up a soundbite and then rolled with it ?

There would be some special programming required to make use of a GPU style compute unit, but I have to wonder how much AVX512 is actually just compiled with auto-vectorization from basic C++. I would expect most AVX512 use is through hand tuned low level libraries or it is HPC code that is also hand tuned. Making use of a cache coherent gpu compute unit wouldn't be much different as long as you supply the necessary libraries. There is probably some other possibilities with a chiplet gpu compute unit that is cache coherent; there is no unnecessary copying and latency could be very low.

Libraries are fine. What's not OK is needing to rely on another compiler on a really unstable platform. At least with AVX-512, you only deal with one compiler with minor microarchitecture specific backends. With GPUs, you see new compilers sprouting up all the damn time because the new hardware generation is incompatible with the previous generation's ISA ...

Most people in the HPC sector in their right minds don't want to deal with GPUs because it's a very hopeless game of trying to micro-optimize for current hardware when new hardware will inevitably OBSOLETE all of their old code because the hardware vendor in question wants to keep chasing higher efficiency by making more incompatible ISAs ...

CUDA which is a leading edge GPU compute platform requires massively more effort to maintain compared to x86 CPUs. A GPU only has like a maximum of 7 years of support (sometimes even less) before it becomes total paperweight with future releases of a GPU compute platform like CUDA deprecating them ...

Somehow programmers are supposed to believe that the future is constantly rewriting tens of thousands of lines of code on an annual basis for optimal performance, a platform deprecating their hardware only after several years, vendors who discourage against micro-optimizing for specific architectures, and then possibly having to deal with buggy toolchains with deprecated hardware ?

Sheesh, it's no wonder why most programmers are cranky about GPUs and they'll always be a distant second when it comes to the HPC sector ...
 

DisEnchantment

Golden Member
Mar 3, 2017
1,590
5,722
136
https://www.phoronix.com/scan.php?page=news_item&px=AMD-SEV-SNP-IOMMU-Linux-5.10
SEV SNP has landed. Well executed by AMD and other contributors like SUSE etc.
Lots of new guys working on AMD Linux drivers can be seen from the commits.

On another note,
If AMD's Xilinx bid is real, expect AMD to keep announcing new products, roadmaps etc. to keep stock prices high.
Zen3 came out and then boom Xilinx rumors.
Expect Milan announcements and special HPC announcements soon to keep stock holders excited. #ExascaleWeek
Then RDNA, consoles, etc., and staggered launch of everything to be in the news continuously.
All the way to CES, Cezanne ...
 

eek2121

Platinum Member
Aug 2, 2005
2,904
3,906
136
I noticed that ARM fanboys tried to hijack this thread again. I just wanted to point out that the A14 has less ST performance (or equal, depending on the benchmark) than TGL, and Zen 3 has more, so please go back to your own thread and leave this one to talk about AMD Zen 3.
 

lobz

Platinum Member
Feb 10, 2017
2,057
2,856
136
I noticed that ARM fanboys tried to hijack this thread again. I just wanted to point out that the A14 has less ST performance (or equal, depending on the benchmark) than TGL, and Zen 3 has more, so please go back to your own thread and leave this one to talk about AMD Zen 3.
That's some real dangerous talk here. If this was Path of Exile, I'd say your comment has a 13% chance to attract monsters from beyond.
 

nicalandia

Diamond Member
Jan 10, 2019
3,330
5,281
136
@uzzi38

Not using N7+, N7, or N7P? That's . . . odd. Might explain why AMD didn't get any all-core clock boosts with Vermeer over Matisse (in fact, they have regression). Totally unexpected. Hopefully the wider core design will help mitigate that. It's also an interesting opportunity for overclockers, especially if AMD has done anything to fight hotspotting.
What Mitigation you are talking about? 20% IPC increase is enough "mitigation"
 

kurosaki

Senior member
Feb 7, 2019
258
250
86
I noticed that ARM fanboys tried to hijack this thread again. I just wanted to point out that the A14 has less ST performance (or equal, depending on the benchmark) than TGL, and Zen 3 has more, so please go back to your own thread and leave this one to talk about AMD Zen 3.
Ahhh you mean that Zen 3 would have a higher IPC?
Like this:
With that said, AMD is all we got at the moment for desktop. But team x86 is going to have to shape up soon, or else there will be competitive derivates of ARM incoming. I'm quite sure.
 

eek2121

Platinum Member
Aug 2, 2005
2,904
3,906
136
Ahhh you mean that Zen 3 would have a higher IPC?
Like this:
With that said, AMD is all we got at the moment for desktop. But team x86 is going to have to shape up soon, or else there will be competitive derivates of ARM incoming. I'm quite sure.

Again, this is an AMD thread. Feel free to create your own thread to continue spewing your bull...err I mean discussing this.
 

kurosaki

Senior member
Feb 7, 2019
258
250
86
Again, this is an AMD thread. Feel free to create your own thread to continue spewing your bull...err I mean discussing this.
I'm in this thread because the 5000-series is closest in hand right now. Then can you be a bit worried at the same time? How is the x86 going to survive this time. All I want is greater performance, and AMD/Intel is going to have to release a heck of an improvement. I believe they both can feel the heat from the blowtorch. I hope we can see a new amdx64, or Core2duo moment in the upcoming years. The arm figures aren't trolling, it's a quite interesting subject to debate as well in the x86 environment. Like C2D, halving the frequency and still perform better than the fastest Prescott's of the time. Quite revolutionary.
 

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
25,478
14,434
136
I'm in this thread because the 5000-series is closest in hand right now. Then can you be a bit worried at the same time? How is the x86 going to survive this time. All I want is greater performance, and AMD/Intel is going to have to release a heck of an improvement. I believe they both can feel the heat from the blowtorch. I hope we can see a new amdx64, or Core2duo moment in the upcoming years. The arm figures aren't trolling, it's a quite interesting subject to debate as well in the x86 environment. Like C2D, halving the frequency and still perform better than the fastest Prescott's of the time. Quite revolutionary.
Arm has its use and AMD/Intel have theirs. You need to stay on topic and NOT talk about ARM.

As was already said Make your own ARM thread.
 

Carfax83

Diamond Member
Nov 1, 2010
6,841
1,536
136
I'm in this thread because the 5000-series is closest in hand right now. Then can you be a bit worried at the same time? How is the x86 going to survive this time. All I want is greater performance, and AMD/Intel is going to have to release a heck of an improvement. I believe they both can feel the heat from the blowtorch. I hope we can see a new amdx64, or Core2duo moment in the upcoming years. The arm figures aren't trolling, it's a quite interesting subject to debate as well in the x86 environment. Like C2D, halving the frequency and still perform better than the fastest Prescott's of the time. Quite revolutionary.

We've had those discussions plenty of times on this forum. There's no need to hijack this thread if you want to talk about ARM vs x86-64. Just make a dedicated thread and people will respond.
 

amrnuke

Golden Member
Apr 24, 2019
1,181
1,772
136
As far as speculation goes for Zen 3, with the move to an 8 core CCX with shared L3$, this seems to obviate the idea of a 4 core 5300X or 5500. I do, however, think that a 3xxxXT might be a good solution for those chips to bring them up to date.

Given the 4xxx APU layouts, I do wonder if Zen 3 APUs are really going to see as much of a benefit, or if they'll essentially be 4xxxXT chips at that point?
 

kurosaki

Senior member
Feb 7, 2019
258
250
86
Arm has its use and AMD/Intel have theirs. You need to stay on topic and NOT talk about ARM.

As was already said Make your own ARM thread.
Try not to misinterpret me now.
If I start an ARM thread with this theme, my guess is that your response will be the same in that thread but in reverse short there after.
This is an AMD thread I believed, and my impression was that the area surrounding AMD was ok to discuss. But I get it, we should not talk about anything else than AMD and AMD only. Quite the narrow parameter, but I guess it's necessary if we don't want the thread to gallop away in the wrong direction.
is there some sort of sub forum for open discussions regarding more than one specific subject at a time?
 
  • Like
Reactions: Lodix

Zucker2k

Golden Member
Feb 15, 2006
1,810
1,159
136
Try not to misinterpret me now.
If I start an ARM thread with this theme, my guess is that your response will be the same in that thread but in reverse short there after.
This is an AMD thread I believed, and my impression was that the area surrounding AMD was ok to discuss. But I get it, we should not talk about anything else than AMD and AMD only. Quite the narrow parameter, but I guess it's necessary if we don't want the thread to gallop away in the wrong direction.
is there some sort of sub forum for open discussions regarding more than one specific subject at a time?
Just create a thread and title it: "ARM vs x86." Or, "ARM vs The World." Or, something along those lines. That should be all encompassing, I think?
 

ksosx86

Member
Sep 27, 2012
105
44
101
No indicators that Zen 3 mobile is at an IPC disadvantage to Zen 2 mobile and or Zen 3 desktop to Zen 3 mobile. (Referring to the APU and or SOC etc any additional embedded implementation etc) If anything the naming schema is just that, to represent parity between mobile components & 105W TDP counterparts. I'm sure whatever loss or attrition for sake of thermals will be minor.
 

soresu

Platinum Member
Dec 19, 2014
2,617
1,812
136
No indicators that Zen 3 mobile is at an IPC disadvantage to Zen 2 mobile and or Zen 3 desktop to Zen 3 mobile. (Referring to the APU and or SOC etc any additional embedded implementation etc) If anything the naming schema is just that, to represent parity between mobile components & 105W TDP counterparts. I'm sure whatever loss or attrition for sake of thermals will be minor.
We probably won't see any hide nor hair of the Cezanne APU until around new year or CES 2021 time - we already know that it will be a virtual conference so it's unlikely that the date will change.