AMD's Future Chips & SoC's: News, Info & Rumours.

[8350rocks]

@goldstone

It looks like it's going to be Glofo's version of Samsungs faster 14LPX process...

"14+ should be 14LPX, which is Samsung's High Performance version of 14nm... Samsung's nodes always have three iterations, they go 14LPE (Low Power Early), 14LPP (Low Power Plus, what Zen & Vega & Polaris are on now), and then 14LPX (Low Power X which stands for Extra or Extreme not sure).
Think beefier transistors that should allow higher clocks without sacrificing density, so 14LPX should be closer to Intel's 14nm in power and frequency, but Intel probably has a slight density lead still."
It's discussed here:
https://www.reddit.com/r/Amd/comments/6c3567/has_anyone_elaborated_on_14nm_stated_in_the_slide/

This article reckons early next year (actually most do). It should have better clock speeds an better power efficiency... Sounds great ! A lot of sites an forums believe early next year. It's supposed to be a tock (zen+) an then a tick (zen2) die shrink... so it would have to be soon enough. Especially with mounting pressure from Intel's Coffee Lake.
We Shall see I guess:
http://digiworthy.com/2017/05/22/amd-ryzen-refresh-series-zen-2/

zen+ = Zen2

Zen2 is what AMD formely named Zen+.

The more I think about this, the more I believe this is a wise marketing move from AMD. People is now in every forum confused about Zen, Zen+, and Zen2.

Also so far as I know Samsung has not licensed any other process node to Glofo. So far as I know Glofo "14nm+" is another marketing name for the same 14LPP used in Zen.

14nm+ is 14nm HP (IBM node)

 

[juanrga]

Intel already developed an multidie technology named EMIB or something as that. This new technology doesn't have the deficiencies and limitations of the older MCM approach. Therefore Intel will be moving from a monolithic die approach when a suitable alternative was mature enough to be used commercially. Whereas AMD is using the MCM approach today because it is cheaper for AMD than designing a monolithic die.

So your now in agreement with me then ?
Being a cheaper solution is not the only reason to take this approach. As Intel is now adopting it too.. They have no choice it seems.

Forward thinking by AMD for sure.. An Intel now seem to be playing catch up on a Muti Die Approach.. I hope EMIB show's some better numbers than MESH did when they dusted it off an implemented it.

Intel are developing "Heterogeneous Compute" with this EMIB approach...

Even the term "Heterogeneous Compute" was coined by AMD.
This is something AMD has been developing for years now, the MCM approach and their Infinity Fabric are all geared towards Heterogeneous Compute... Intel are playing catch up with this also.

I suppose imitation is the highest form of flattery after all !

I disagree!

Intel is not adopting the MCM approach that AMD is using. Intel has developed a new technology without the limitations and weaknesses of the MCM approach that AMD is using. Intel will use that new technology when was ready for commercial usage.

AMD didn't develop "the MCM approach". MCM was invented by other people in the past and been used in many chips. For instance the older Power2 chips from IBM.

Heterogeneous computation was developed by academic scientists and engineers. Here you have an older article on the subject

http://ieeexplore.ieee.org/document/1541876

As you can see it was published before AMD acquired ATI.

Intel isn't imitating AMD. Intel has been working in heterogeneous computation during decades. Six years ago Intel Labs constructed three prototypes of heterogeneous processors and studied each one of them under different workloads. Much before, in 2005 Intel was giving talks about its vision of mixing different cores in heterogeneous processors

Here you have another ancient presentation

https://www.pcper.com/reviews/Processors/Intels-80-Core-Terascale-Chip-Explored-4GHz-clocks-and-more/Defining-Architectu-0

if you pay attention to the diagrams, you can see that six cores in the right hand side of the chip are different to the rest of chips. This is because engineers considered the possibility of using different cores for different tasks, aka an heterogeneous design.


I wonder who is imitating to whom.

This is something AMD has been developing for years now, the MCM approach and their Infinity Fabric are all geared towards Heterogeneous Compute... Intel are playing catch up with this also.

I suppose imitation is the highest form of flattery after all !

 

[juanrga]

@goldstone

It looks like it's going to be Glofo's version of Samsungs faster 14LPX process...

"14+ should be 14LPX, which is Samsung's High Performance version of 14nm... Samsung's nodes always have three iterations, they go 14LPE (Low Power Early), 14LPP (Low Power Plus, what Zen & Vega & Polaris are on now), and then 14LPX (Low Power X which stands for Extra or Extreme not sure).
Think beefier transistors that should allow higher clocks without sacrificing density, so 14LPX should be closer to Intel's 14nm in power and frequency, but Intel probably has a slight density lead still."
It's discussed here:
https://www.reddit.com/r/Amd/comments/6c3567/has_anyone_elaborated_on_14nm_stated_in_the_slide/

This article reckons early next year (actually most do). It should have better clock speeds an better power efficiency... Sounds great ! A lot of sites an forums believe early next year. It's supposed to be a tock (zen+) an then a tick (zen2) die shrink... so it would have to be soon enough. Especially with mounting pressure from Intel's Coffee Lake.
We Shall see I guess:
http://digiworthy.com/2017/05/22/amd-ryzen-refresh-series-zen-2/

zen+ = Zen2

Zen2 is what AMD formely named Zen+.

The more I think about this, the more I believe this is a wise marketing move from AMD. People is now in every forum confused about Zen, Zen+, and Zen2.

Also so far as I know Samsung has not licensed any other process node to Glofo. So far as I know Glofo "14nm+" is another marketing name for the same 14LPP used in Zen.

14nm+ is 14nm HP (IBM node)

Nope.

 

[goldstone77]

I have thought once we reach the barriers of shrinking Cpu's would grow in core count and in the direction of specialized cores for specific tasks.

 

[goldstone77]

Hot Chips 2017: AMD Outlines Threadripper And EPYC's MCM Advantage, Claims 41% Cost Reduction
by Paul Alcorn August 25, 2017 at 8:40 AM
"Final Thoughts
Although AMD has taken the high road and not responded directly to Intel's EPYC slide deck, its Hot Chips presentation seems to dispute a few of Intel's key arguments. First and foremost, AMD designed the die specifically for the data center. Intel's "glue" remark refers to glue logic, which is an industry term for interconnects between die (in this case AMD's Infinity Fabric). In either case, the insinuation that AMD is merely using desktop silicon for the data center certainly has a negative connotation. However, AMD's tactic is innovative and reduces cost. As we stated in our piece covering Intel's deck:

This mitigates the risk that goes into manufacturing complex, monolithic processors, potentially improving yields and keeping costs in check. It also helps the company increase volume at a time when it's going to want plenty of supply. It's a smart strategy for a fabless company that doesn't have Intel's R&D budget to throw around.

Interestingly, Intel's Programmable Solutions Group outlined the company's EMIB (Embedded Multi-die Interconnect Bridge) technology at the show, too. EMIB provides a communication pathway between separate chips to create a unified processing solution, which Intel considers a key technology for its next-gen processors. Although the approaches are different, the motivation behind Intel's EMIB and AMD's Infinity Fabric are similar, which AMD feels validates its approach.
In either case, AMD is humming right along with its EPYC processors, and a wide range of blue chip OEMs and ODMs have platforms coming to market. AMD also announced at the China EPYC technology summit that Tencent will deploy EPYC solutions by the end of the year and JD.com will deploy in the second half of the year. We expect more announcements in the future as the war for data center sockets rages on."

There is a lot more information just follow the link below.
http://www.tomshardware.com/news/amd-threadripper-epyc-mcm-cost,35306.html

 

[jdwii]

I have thought once we reach the barriers of shrinking Cpu's would grow in core count and in the direction of specialized cores for specific tasks.

X86 has hit a wall years ago not sure if its over Intel needing to make a new architecture or it really is time for X86 to die.

Single core performance will matter for ever. I actually hope we make stronger single core CPUs and SMT makes up for the unused resources.

That way we have the best of two worlds. Ryzen SMT works better then HT i believe over Zen not being able to use all the resources for one thread per core as efficiently as Kaby-Skylake so their is more resources left over for the other thread.

But until some other design can run X86 at the same speed i doubt we will ever move over to something different.

If we could get a X86 architecture to split a thread to multiple cores then i'm all for having more cores in my system. Yes their is a mock up of this and i highly doubt it's feasible but i'd love to be wrong.

Something like a VISC+Morphcore design here is a fun read even if its outdated Juan has some very good info on this subject

https://www.extremetech.com/extreme/192858-visc-cpu-virtual-core-design-emerges-could-this-be-the-conceptual-breakthrough-weve-been-waiting-for

Now i really do feel like we should move on from dual cores and heck even quad cores in 2017 for "enthusiast class" PC.

 

[goldstone77]

I have thought once we reach the barriers of shrinking Cpu's would grow in core count and in the direction of specialized cores for specific tasks.

X86 has hit a wall years ago not sure if its over Intel needing to make a new architecture or it really is time for X86 to die.

Single core performance will matter for ever. I actually hope we make stronger single core CPUs and SMT makes up for the unused resources.

That way we have the best of two worlds. Ryzen SMT works better then HT i believe over Zen not being able to use all the resources for one thread per core as efficiently as Kaby-Skylake so their is more resources left over for the other thread.

But until some other design can run X86 at the same speed i doubt we will ever move over to something different.

If we could get a X86 architecture to split a thread to multiple cores then i'm all for having more cores in my system. Yes their is a mock up of this and i highly doubt it's feasible but i'd love to be wrong.

Something like a VISC+Morphcore design here is a fun read even if its outdated Juan has some very good info on this subject

https://www.extremetech.com/extreme/192858-visc-cpu-virtual-core-design-emerges-could-this-be-the-conceptual-breakthrough-weve-been-waiting-for

Now i really do feel like we should move on from dual cores and heck even quad cores in 2017 for "enthusiast class" PC.

This only pushes my point even harder. True throughput can only happen with more cores! Eventually, processors and software will be designed to optimize for each other for specific goals. Again I say this a lot, but once A.I. takes over hardware and software design we will see optimizations like this become seamless.

 

[gamerk316]

I have thought once we reach the barriers of shrinking Cpu's would grow in core count and in the direction of specialized cores for specific tasks.

X86 has hit a wall years ago not sure if its over Intel needing to make a new architecture or it really is time for X86 to die.

Single core performance will matter for ever. I actually hope we make stronger single core CPUs and SMT makes up for the unused resources.

That way we have the best of two worlds. Ryzen SMT works better then HT i believe over Zen not being able to use all the resources for one thread per core as efficiently as Kaby-Skylake so their is more resources left over for the other thread.

But until some other design can run X86 at the same speed i doubt we will ever move over to something different.

If we could get a X86 architecture to split a thread to multiple cores then i'm all for having more cores in my system. Yes their is a mock up of this and i highly doubt it's feasible but i'd love to be wrong.

Something like a VISC+Morphcore design here is a fun read even if its outdated Juan has some very good info on this subject

https://www.extremetech.com/extreme/192858-visc-cpu-virtual-core-design-emerges-could-this-be-the-conceptual-breakthrough-weve-been-waiting-for

Now i really do feel like we should move on from dual cores and heck even quad cores in 2017 for "enthusiast class" PC.

Intel DID make a new CPU architecture; I was called Itanium. It failed over the "quick fix" that was x86-64, and now that it's entrenched as the architecture for 64-bit computing, it isn't going anywhere for several decades. We're stuck with it, like it or not.

 

[jaymc]

@goldstone

It looks like it's going to be Glofo's version of Samsungs faster 14LPX process...

"14+ should be 14LPX, which is Samsung's High Performance version of 14nm... Samsung's nodes always have three iterations, they go 14LPE (Low Power Early), 14LPP (Low Power Plus, what Zen & Vega & Polaris are on now), and then 14LPX (Low Power X which stands for Extra or Extreme not sure).
Think beefier transistors that should allow higher clocks without sacrificing density, so 14LPX should be closer to Intel's 14nm in power and frequency, but Intel probably has a slight density lead still."
It's discussed here:
https://www.reddit.com/r/Amd/comments/6c3567/has_anyone_elaborated_on_14nm_stated_in_the_slide/

This article reckons early next year (actually most do). It should have better clock speeds an better power efficiency... Sounds great ! A lot of sites an forums believe early next year. It's supposed to be a tock (zen+) an then a tick (zen2) die shrink... so it would have to be soon enough. Especially with mounting pressure from Intel's Coffee Lake.
We Shall see I guess:
http://digiworthy.com/2017/05/22/amd-ryzen-refresh-series-zen-2/

zen+ = Zen2

Zen2 is what AMD formely named Zen+.

The more I think about this, the more I believe this is a wise marketing move from AMD. People is now in every forum confused about Zen, Zen+, and Zen2.

Also so far as I know Samsung has not licensed any other process node to Glofo. So far as I know Glofo "14nm+" is another marketing name for the same 14LPP used in Zen.

I don't think so.. not according to AMD's own Roadmap..
First Zen on 14nm
Then 14nm+ or what is being referred to as Zen+ at the moment anyway (Rumours are it's a new node but may be a tweaked 14nm process).
Then Zen 2 on 7nm (the shrink)
An then we have Zen 3 which is on 7nm+
check it here:
http://cdn.wccftech.com/wp-content/uploads/2017/05/AMD-Zen_3.png

 

[-Fran-]

I wouldn't dwelve into what "Zen2" and "Zen+" did mean or actually mean for now. Marketing slides are not very good in terms on naming products with no official code-names. Is it even relevant?

At worst, ok, Zen+ is not Zen2 but Zen1 in 14nm+; at best is Zen2 = Zen+ and it will be in 7nm around Q4 2018 / Q2 2019. The important information is that Zen2, which is the real successor to Zen1, will be in 7nm and at/after H2 2018.

Cheers!

 

[jaymc]

Would be nice to have a product refresh before Zen 2 though.. With some higher clocks..
I think this is the "at best" scenerio...

 

[-Fran-]

Why would it be the best scenario? I rather GF use their resources in getting 7nm out of the factory faster and AMD focus whatever they have of resources on Zen2 while Ryzen keeps on being sold as 14nm and in rev1 flavor. I mean, they did hold for like 4 years or more with Piledriver, so what's an extra 6 months of Ryzen? XD

Cheers!

 

[juanrga]

Hot Chips 2017: AMD Outlines Threadripper And EPYC's MCM Advantage, Claims 41% Cost Reduction
by Paul Alcorn August 25, 2017 at 8:40 AM
"Final Thoughts
Although AMD has taken the high road and not responded directly to Intel's EPYC slide deck, its Hot Chips presentation seems to dispute a few of Intel's key arguments. First and foremost, AMD designed the die specifically for the data center. Intel's "glue" remark refers to glue logic, which is an industry term for interconnects between die (in this case AMD's Infinity Fabric). In either case, the insinuation that AMD is merely using desktop silicon for the data center certainly has a negative connotation. However, AMD's tactic is innovative and reduces cost. As we stated in our piece covering Intel's deck:

This mitigates the risk that goes into manufacturing complex, monolithic processors, potentially improving yields and keeping costs in check. It also helps the company increase volume at a time when it's going to want plenty of supply. It's a smart strategy for a fabless company that doesn't have Intel's R&D budget to throw around.

Interestingly, Intel's Programmable Solutions Group outlined the company's EMIB (Embedded Multi-die Interconnect Bridge) technology at the show, too. EMIB provides a communication pathway between separate chips to create a unified processing solution, which Intel considers a key technology for its next-gen processors. Although the approaches are different, the motivation behind Intel's EMIB and AMD's Infinity Fabric are similar, which AMD feels validates its approach.
In either case, AMD is humming right along with its EPYC processors, and a wide range of blue chip OEMs and ODMs have platforms coming to market. AMD also announced at the China EPYC technology summit that Tencent will deploy EPYC solutions by the end of the year and JD.com will deploy in the second half of the year. We expect more announcements in the future as the war for data center sockets rages on."

There is a lot more information just follow the link below.
http://www.tomshardware.com/news/amd-threadripper-epyc-mcm-cost,35306.html

EMIB cannot be compared with Infinity Fabric. EMIB is a new multidie packaging technology. Infinity Fabric is an interconnect.

 

[jaymc]

Why would it be the best scenario? I rather GF use their resources in getting 7nm out of the factory faster and AMD focus whatever they have of resources on Zen2 while Ryzen keeps on being sold as 14nm and in rev1 flavor. I mean, they did hold for like 4 years or more with Piledriver, so what's an extra 6 months of Ryzen? XD

Cheers!

Well I think 7nm will happen when it happens regardless. They are working on it in a separate fab completely are they not ?
There will be a 14nm+ refresh they are gonna tweak the node.. at the very least.
Any delay (hopefully no delay's but) will most likely come from problems with GF an the node shrink... although they are saying they are on schedule, there may be delays.

Also remember AMD are free to use a different foundry.. there's nothing to stop them approaching Samsung an using there faster 14nm lpx node while they wait on 7nm to be ready at GF. There are only locked in with GF for so many chips an then there free to go elsewhere. Should be quite easy for them to port it to Samsung's node as there very similar.

If there is any delays at GF I hope to see AMD doing something along these lines. It would teach GF a lesson at the same time.

The hard work is done on Zen, for Zen 2 they could shrink it as is maybe just add faster cache an we still gonna see huge benefits. Intel springs to mind few slight changes an quicker cache an that's it except plus a node shrink.. Intel would call it a new generation without the shrink.

 

[goldstone77]

I gave my reasons for why I have trouble understanding the release another CPU before 7nm.
1. Lack much tangible information on a processor to be released before 7nm.
2. It goes against their efficiency model being that 8 core parts are used to make up Server parts, and HEDT parts, and defects fall into R5 and R3 lines.
3. Their road map says 7nm Epyc 2nd half 2018!
4. They changed the GPU roadmap from Vega 20 to 7nm Navi for 2018!

 

[juanrga]

@goldstone

It looks like it's going to be Glofo's version of Samsungs faster 14LPX process...

"14+ should be 14LPX, which is Samsung's High Performance version of 14nm... Samsung's nodes always have three iterations, they go 14LPE (Low Power Early), 14LPP (Low Power Plus, what Zen & Vega & Polaris are on now), and then 14LPX (Low Power X which stands for Extra or Extreme not sure).
Think beefier transistors that should allow higher clocks without sacrificing density, so 14LPX should be closer to Intel's 14nm in power and frequency, but Intel probably has a slight density lead still."
It's discussed here:
https://www.reddit.com/r/Amd/comments/6c3567/has_anyone_elaborated_on_14nm_stated_in_the_slide/

This article reckons early next year (actually most do). It should have better clock speeds an better power efficiency... Sounds great ! A lot of sites an forums believe early next year. It's supposed to be a tock (zen+) an then a tick (zen2) die shrink... so it would have to be soon enough. Especially with mounting pressure from Intel's Coffee Lake.
We Shall see I guess:
http://digiworthy.com/2017/05/22/amd-ryzen-refresh-series-zen-2/

zen+ = Zen2

Zen2 is what AMD formely named Zen+.

The more I think about this, the more I believe this is a wise marketing move from AMD. People is now in every forum confused about Zen, Zen+, and Zen2.

Also so far as I know Samsung has not licensed any other process node to Glofo. So far as I know Glofo "14nm+" is another marketing name for the same 14LPP used in Zen.

I don't think so.. not according to AMD's own Roadmap..
First Zen on 14nm
Then 14nm+ or what is being referred to as Zen+ at the moment anyway (Rumours are it's a new node but may be a tweaked 14nm process).
Then Zen 2 on 7nm (the shrink)
An then we have Zen 3 which is on 7nm+
check it here:
http://cdn.wccftech.com/wp-content/uploads/2017/05/AMD-Zen_3.png

That 'roadmap' just proves what I have said. First "Zen" on 14nm. Then "Zen" on 14nm+. Then "Zen2" on 7nm.

Zen+ was renamed to Zen2.

I will repost this

2017 --> Zen on 14nm
2018 --> Zen on 14nm+.

"14nm+" is just a marketing term for a more mature 14nm node. If the transition Trinity --> Richland can be used as example. Then Pinnacle Ridge will have 200--300MHz higher than Summit Ridge.

 

[juanrga]

Well I think 7nm will happen when it happens regardless. They are working on it in a separate fab completely are they not ?
There will be a 14nm+ refresh they are gonna tweak the node.. at the very least.
Any delay (hopefully no delay's but) will most likely come from problems with GF an the node shrink... although they are saying they are on schedule, there may be delays.

Also remember AMD are free to use a different foundry.. there's nothing to stop them approaching Samsung an using there faster 14nm lpx node while they wait on 7nm to be ready at GF.

Except they would pay twice. One to Samsung for the fabricating chips. Other to Glofo for violating the WSA.

 

[juanrga]

I gave my reasons for why I have trouble understanding the release another CPU before 7nm.
1. Lack much tangible information on a processor to be released before 7nm.
2. It goes against their efficiency model being that 8 core parts are used to make up Server parts, and HEDT parts, and defects fall into R5 and R3 lines.
3. Their road map says 7nm Epyc 2nd half 2018!
4. They changed the GPU roadmap from Vega 20 to 7nm Navi for 2018!

And the public and private roadmaps contradict what you say!

First Zen on 14nm on 2017. Then Zen on 14nm+ on 2018. Then Zen2 on 7nm on 2019.

 

[jaymc]

I gave my reasons for why I have trouble understanding the release another CPU before 7nm.
1. Lack much tangible information on a processor to be released before 7nm.
2. It goes against their efficiency model being that 8 core parts are used to make up Server parts, and HEDT parts, and defects fall into R5 and R3 lines.
3. Their road map says 7nm Epyc 2nd half 2018!
4. They changed the GPU roadmap from Vega 20 to 7nm Navi for 2018!

I don't mean it will be another CPU release, same cpu on faster process hence the + beside 14nm.
It may be just a little faster. With the very same uArch an same name just rev2 or something in the name !
It's in the Roadmap there is something happening before Zen 2...
That's why there is a 14nm+ underneath the current process node that's being used at the mo.
It's normal to at least tweak the process node.. same thing happened with Polaris when it went from rx480 to rx580. The design of the chip doesn't have to change just the process node was tweaked.
Anyway time will tell we shall see...

 

[jdwii]

Why would it be the best scenario? I rather GF use their resources in getting 7nm out of the factory faster and AMD focus whatever they have of resources on Zen2 while Ryzen keeps on being sold as 14nm and in rev1 flavor. I mean, they did hold for like 4 years or more with Piledriver, so what's an extra 6 months of Ryzen? XD

Cheers!

https://www.globalfoundries.com/news-events/press-releases/globalfoundries-track-deliver-leading-performance-7nm-finfet-technology

first customer products based on 7LP are expected to launch in the first half of 2018, with volume production ramping in the second half of 2018

I actually think if Amd can get it together and start samples in 2018 1H they could get Ryzen 2 out the door in the 4th quarter of 2018.

The thing is man sorry after all these years i just don't trust global foundries to do it without a delay so making Ryzen+ or whatever they call it on in improved 14nm+ and releasing the CPU in 1H 2018 seems like a more reliable safe option for Amd. Plus having reviews and tech tubers review Amd products keeps them relevant.

Another thing to consider is just about all the reviews of ryzen are old. Most running on unstable memory old AGESA's, running HEDT on instead of off which was in issue, software patches on several games happen and made a big differences even if Ryzen was reviewed again today it would look better think about future games using more cores for example PUBG had an update to run on 6+ cores.

Ryzen+ or zen on 14nm+ will look better plus lets not forget its pretty typical of Amd to do much better the second round then the first.


Also come on Amd get a marketing name out for zen on 14nm+ intel has them for 5+ future products ever think this is why you do poorly in the stock market get a more detailed timeline so people know the progress that should be expected.

 

[jdwii]

I gave my reasons for why I have trouble understanding the release another CPU before 7nm.
1. Lack much tangible information on a processor to be released before 7nm.
2. It goes against their efficiency model being that 8 core parts are used to make up Server parts, and HEDT parts, and defects fall into R5 and R3 lines.
3. Their road map says 7nm Epyc 2nd half 2018!
4. They changed the GPU roadmap from Vega 20 to 7nm Navi for 2018!

Well we will see but if history has taught us anything..........i'll leave it at that my faith in global foundries is highly lacking.

Be cool if they could get it out the door though they will need too last thing i want to see is Amd's top mainstream product selling at I5 prices.

 

[-Fran-]

And the public and private roadmaps contradict what you say!

First Zen on 14nm on 2017. Then Zen on 14nm+ on 2018. Then Zen2 on 7nm on 2019.

Come on, Juan. You know AMD changes its mind from one day to the other with product releases. The change from "Zen+" to "Zen2" should make you doubt those roadmaps as things "set on stone"; do I need to remind you all of the times AMD has changed its mind on other things? Plus, in the very little tiny chance that GF is not BS'ing everyone and they do get their stuff out the factory on time, the times would align to H2 2018 full fledged release. Fat chance given track record, yes, but still a possibility.

Plus, we all win if they can get stuff sooner. Intel really needs the competition. Any "on schedule" and "expected" words are things to celebrate when it comes from AMD/GF, lol.

https://www.globalfoundries.com/news-events/press-releases/globalfoundries-track-deliver-leading-performance-7nm-finfet-technology

first customer products based on 7LP are expected to launch in the first half of 2018, with volume production ramping in the second half of 2018

I actually think if Amd can get it together and start samples in 2018 1H they could get Ryzen 2 out the door in the 4th quarter of 2018.

The thing is man sorry after all these years i just don't trust global foundries to do it without a delay so making Ryzen+ or whatever they call it on in improved 14nm+ and releasing the CPU in 1H 2018 seems like a more reliable safe option for Amd. Plus having reviews and tech tubers review Amd products keeps them relevant.

Another thing to consider is just about all the reviews of ryzen are old. Most running on unstable memory old AGESA's, running HEDT on instead of off which was in issue, software patches on several games happen and made a big differences even if Ryzen was reviewed again today it would look better think about future games using more cores for example PUBG had an update to run on 6+ cores.

Ryzen+ or zen on 14nm+ will look better plus lets not forget its pretty typical of Amd to do much better the second round then the first.

Also come on Amd get a marketing name out for zen on 14nm+ intel has them for 5+ future products ever think this is why you do poorly in the stock market get a more detailed timeline so people know the progress that should be expected.

I'm not trusting GF neither, but if they have officially said they taped out the 7nm stuff, that means they need to tweak now and prepare for ramp up. ES'es for Zen2 should be hitting people around Q1 2018 then? If we see leaks around that time, then H2 2018 is absolutely feasible. If we don't see *anything*, then we can assume there's going to be a delay. This would be a "no news is bad news" scenario, I guess... Or "no leaks is bad news", haha.

Cheers!

 

[goldstone77]

I gave my reasons for why I have trouble understanding the release another CPU before 7nm.
1. Lack much tangible information on a processor to be released before 7nm.
2. It goes against their efficiency model being that 8 core parts are used to make up Server parts, and HEDT parts, and defects fall into R5 and R3 lines.
3. Their road map says 7nm Epyc 2nd half 2018!
4. They changed the GPU roadmap from Vega 20 to 7nm Navi for 2018!

And the public and private roadmaps contradict what you say!

First Zen on 14nm on 2017. Then Zen on 14nm+ on 2018. Then Zen2 on 7nm on 2019.

Juanrga I know you like to use leaked roadmaps like the one you keep showing off, but let's just stick to the latest official releases from Analyst day! The dates on the roadmaps are not exact time tables for launches, which you have pointed out in the past. I don't give an exact time tables either. I give possible time tables not that I believe it's going to happen, but a possibility even if it is a slim one. I would assume as well that 14+ would be a filler if they couldn't get to 7nm, because it doesn't make sense with their efficiency model, which uses 8 core parts to make server CPUs. Also, TSMC could potentially be an options as well.

Zen there are no release dates on the roadmaps for 2018-2019, and for 2017 14nm and 14+.

Epyc roadmap shows the next processor is 7nm not 14+.

2018-2019 not listed on this roadmap either although it does say 14nm and 14+ over 2017.

https://www.techpowerup.com/233389/amd-to-continue-working-with-tsmc-globalfoundries-on-7-nm-ryzen

"As it is, AMD has to pay GLOBALFOUNDRIES for its wafer orders that go to other silicon producers (in this case, TSMC), in a quarterly basis since the beginning of 2017, based on the volume of certain wafers purchased from another wafer foundry. In addition, AMD has annual wafer purchase targets from 2016 through the end of 2020, fixed wafer prices for 2016, and a framework for yearly wafer pricing in this amendment, so the company is still bleeding money to GLOBALFOUNDRIES. However, AMD is making the correct decision in this instance, I'd wager, considering GLOBALFOUNDRIES' known difficulties in delivering their process nodes absent of quirks."
https://www.techpowerup.com/225471/amd-announces-amendment-to-wafer-supply-agreement-with-globalfoundries

"In addition to modifying certain terms of the WSA applicable to AMD's microprocessor, graphics processor, and semi-custom products, the amendment:
Covers a 5-year period, spanning from calendar year 2016 through 2020;
Establishes a comprehensive framework for technology collaboration between AMD and GF for the 7nm technology node, building on the success of the 14nm node;
Provides AMD with the flexibility to manufacture certain products with another wafer foundry;
Sets annual wafer purchase targets from 2016 through the end of 2020, fixed wafer prices for 2016, and a framework for yearly wafer pricing.
In partial consideration for these rights, AMD will:
Make a $100 million cash payment to GF, paid in installments beginning in Q4 2016 through Q3 2017.
Make quarterly payments to GF beginning in 2017 based on the volume of certain wafers purchased from another wafer foundry.
Grant to West Coast Hitech L.P., a wholly-owned subsidiary of the Mubadala Development Company PJSC, a warrant to purchase 75 million shares of AMD common stock at a purchase price of $5.98 per share. The warrant may be exercised in whole or in part prior to February 29, 2020. The warrant is only exercisable to the extent that Mubadala or its subsidiaries do not beneficially own, either directly or indirectly, an aggregate of more than 19.99 percent of AMD's outstanding capital stock after the exercise."
https://www.techpowerup.com/225471/amd-announces-amendment-to-wafer-supply-agreement-with-globalfoundries

 

[juanrga]

And the public and private roadmaps contradict what you say!

First Zen on 14nm on 2017. Then Zen on 14nm+ on 2018. Then Zen2 on 7nm on 2019.

Come on, Juan. You know AMD changes its mind from one day to the other with product releases. The change from "Zen+" to "Zen2" should make you doubt those roadmaps as things "set on stone"; do I need to remind you all of the times AMD has changed its mind on other things? Plus, in the very little tiny chance that GF is not BS'ing everyone and they do get their stuff out the factory on time, the times would align to H2 2018 full fledged release. Fat chance given track record, yes, but still a possibility.

Plus, we all win if they can get stuff sooner. Intel really needs the competition. Any "on schedule" and "expected" words are things to celebrate when it comes from AMD/GF, lol.

https://www.globalfoundries.com/news-events/press-releases/globalfoundries-track-deliver-leading-performance-7nm-finfet-technology

first customer products based on 7LP are expected to launch in the first half of 2018, with volume production ramping in the second half of 2018

I actually think if Amd can get it together and start samples in 2018 1H they could get Ryzen 2 out the door in the 4th quarter of 2018.

The thing is man sorry after all these years i just don't trust global foundries to do it without a delay so making Ryzen+ or whatever they call it on in improved 14nm+ and releasing the CPU in 1H 2018 seems like a more reliable safe option for Amd. Plus having reviews and tech tubers review Amd products keeps them relevant.

Another thing to consider is just about all the reviews of ryzen are old. Most running on unstable memory old AGESA's, running HEDT on instead of off which was in issue, software patches on several games happen and made a big differences even if Ryzen was reviewed again today it would look better think about future games using more cores for example PUBG had an update to run on 6+ cores.

Ryzen+ or zen on 14nm+ will look better plus lets not forget its pretty typical of Amd to do much better the second round then the first.

Also come on Amd get a marketing name out for zen on 14nm+ intel has them for 5+ future products ever think this is why you do poorly in the stock market get a more detailed timeline so people know the progress that should be expected.

I'm not trusting GF neither, but if they have officially said they taped out the 7nm stuff, that means they need to tweak now and prepare for ramp up. ES'es for Zen2 should be hitting people around Q1 2018 then? If we see leaks around that time, then H2 2018 is absolutely feasible. If we don't see *anything*, then we can assume there's going to be a delay. This would be a "no news is bad news" scenario, I guess... Or "no leaks is bad news", haha.

The only change was that "Zen+" was renamed to "Zen2". The rest is unchanged.

Yes, AMD has a long history of changing roadmaps, but in the "worse" sense, such as canceling products or delaying launches. There is absolutely no indication either direct or indirect that Zen2 comes next year. Customers would benefit if 7nm Zen2 was launched tomorrow, but don't confound reality with personal wishes.

 

[-Fran-]

The only change was that "Zen+" was renamed to "Zen2". The rest is unchanged.

Yes, AMD has a long history of changing roadmaps, but in the "worse" sense, such as canceling products or delaying launches. There is absolutely no indication either direct or indirect that Zen2 comes next year. Customers would benefit if 7nm Zen2 was launched tomorrow, but don't confound reality with personal wishes.

Since when "feasibility" is "wishing"?

Juan, you're pulling straws, stop it.

Cheers!