Overclock.net banner

[TweakTown] AMD's next-gen Navi GPU launching in August 2018

9K views 60 replies 42 participants last post by  Majin SSJ Eric 
#1 ·
Quote:
Once again I have an exclusive story that AMD will have Navi ready to go sometime in July-August 2018, with a Navi-based professional card being launched at SIGGRAPH 2018. We're still waiting for AMD to launch Radeon Pro SSG, something they unveiled during SIGGRAPH 2017 that hasn't yet materialized. In the meantime, Radeon Technologies Group boss Raja Koduri has taken a sabbatical from the company until early-2018.

Navi will be made on the 7nm process, but other than that we don't know much.
Read more: https://www.tweaktown.com/news/59428/amds-next-gen-navi-gpu-launching-august-2018/index.html?utm_source=dlvr.it&utm_medium=twitter&utm_campaign=tweaktown

 
See less See more
1
#2 ·
So... they talked about Vega refresh next year so they will end up as the Mid Range and low end?

We still haven't seen Vega 11, 12 and 20, unless of course its been scraped.
 
#4 ·
This seems early for Navi to me but I have no idea how possible such a thing actually is at this point. I'm sure AMD is itching to get competitive with Nvidia again ASAP but there would seem to be logistical hurdles to bringing out Navi so soon, at least to me. And we all know how very behind AMD is currently in the GPU sector.
 
#5 ·
Quote:
Originally Posted by Majin SSJ Eric View Post

This seems early for Navi to me but I have no idea how possible such a thing actually is at this point. I'm sure AMD is itching to get competitive with Nvidia again ASAP but there would seem to be logistical hurdles to bringing out Navi so soon, at least to me. And we all know how very behind AMD is currently in the GPU sector.
Neither Nvidia nor AMD will have 7nm chips out next year.

All we will be getting is 12nm.

All 7nm stuff coming out next year will be lower power devices like phones SOCs from ultra rich companies like qualcomm, samsung and apple. Considering we are still on 10nm designs this year, 7nm supply is going to get eaten up by all the telecom companies.

Neither Nvidia nor AMD or have enough at stake to get a 7nm chip out the door next year. It's too expensive. 7nm designs are super resource intensive which means AMD doesn't have the cashflow to focus on too many 7nm designs at once. It's going to be a struggle for Nvidia even which is why they are putting likely a few 12nm designs for volta.

AMD is going to put it's capital into zen 2 which means this is likely going to get a 7nm design first.

Considering this headline says navi is a professional product and Vega 20 is also one, it conflict on what we know and have comfirmed already. AMD wouldn't release Vega 20 and have it get obsoleted by Navi shortly after.
 
#6 ·
Quote:
Originally Posted by tajoh111 View Post

Neither Nvidia nor AMD will have 7nm chips out next year.

All we will be getting is 12nm.

All 7nm stuff coming out next year will be lower power devices like phones SOCs from ultra rich companies like qualcomm, samsung and apple. Considering we are still on 10nm designs this year, 7nm supply is going to get eaten up by all the telecom companies.

Neither Nvidia nor AMD or have enough at stake to get a 7nm chip out the door next year. It's too expensive. 7nm designs are super resource intensive which means AMD doesn't have the cashflow to focus on too many 7nm designs at once. It's going to be a struggle for Nvidia even which is why they are putting likely a few 12nm designs for volta.

AMD is going to put it's capital into zen 2 which means this is likely going to get a 7nm design first.

Considering this headline says navi is a professional product and Vega 20 is also one, it conflict on what we know and have comfirmed already. AMD wouldn't release Vega 20 and have it get obsoleted by Navi shortly after.
GF's IBM engineered 7LP isn't meant for phone designs. It was designed for IBM's large mainframe chips until they sold it to GF (or rather they paid GF to buy it IIRC). The frequency target was like 5GHz or something so we'll probably see it with Zen first.
 
#8 ·
So, February 2019
 
  • Rep+
Reactions: awdrifter
#9 ·
Quote:
Originally Posted by tajoh111 View Post

Neither Nvidia nor AMD will have 7nm chips out next year.

All we will be getting is 12nm.

All 7nm stuff coming out next year will be lower power devices like phones SOCs from ultra rich companies like qualcomm, samsung and apple. Considering we are still on 10nm designs this year, 7nm supply is going to get eaten up by all the telecom companies.

Neither Nvidia nor AMD or have enough at stake to get a 7nm chip out the door next year. It's too expensive. 7nm designs are super resource intensive which means AMD doesn't have the cashflow to focus on too many 7nm designs at once. It's going to be a struggle for Nvidia even which is why they are putting likely a few 12nm designs for volta.

AMD is going to put it's capital into zen 2 which means this is likely going to get a 7nm design first.

Considering this headline says navi is a professional product and Vega 20 is also one, it conflict on what we know and have comfirmed already. AMD wouldn't release Vega 20 and have it get obsoleted by Navi shortly after.
We agree about something AMD-related!!!
tongue.gif
 
#10 ·
#12 ·
Quote:
Originally Posted by tajoh111 View Post

Neither Nvidia nor AMD will have 7nm chips out next year.

All we will be getting is 12nm.

All 7nm stuff coming out next year will be lower power devices like phones SOCs from ultra rich companies like qualcomm, samsung and apple. Considering we are still on 10nm designs this year, 7nm supply is going to get eaten up by all the telecom companies.

Neither Nvidia nor AMD or have enough at stake to get a 7nm chip out the door next year. It's too expensive. 7nm designs are super resource intensive which means AMD doesn't have the cashflow to focus on too many 7nm designs at once. It's going to be a struggle for Nvidia even which is why they are putting likely a few 12nm designs for volta.

AMD is going to put it's capital into zen 2 which means this is likely going to get a 7nm design first.

Considering this headline says navi is a professional product and Vega 20 is also one, it conflict on what we know and have comfirmed already. AMD wouldn't release Vega 20 and have it get obsoleted by Navi shortly after.
GF announced that its 7nm LP node (LP = Leading Performance) is ready for partners to begin planning their designs. The first customer launches on 7nm LP are expected 12-18 months from now, and GF is promising that it can deliver up to 40 percent improved performance compared with 14nm. The company claims that its 7nm work is exceeding its performance and power targets, and is on track to deliver up to 2x area scaling compared with previous 14nm technology. The company has also been hard at work on 5nm in partnership with IBM.

"Our 7nm FinFET technology development is on track and we are seeing strong customer traction, with multiple product tapeouts planned in 2018," said Gregg Bartlett, senior vice president of the CMOS Business Unit at GF. "And, while driving to commercialize 7nm, we are actively developing next-generation technologies at 5nm and beyond to ensure our customers have access to a world-class roadmap at the leading edge."




GlobalFoundries Announces Early 7nm Availability, Huge Gains Over 14nm FinFET

Not saying you're right or wrong... Just passing along this message that has already been released before, in June of this year. Just notice what is written on the bottom right corner of the image.

Additionally, if AMD moving to TSMC is true for their graphics cards, there's still this;

TSMC's first-generation CLN7FF will enter risk production in Q2 2017 and will be used for over a dozen of tape outs this year. It is expected that high-volume manufacturing (HVM) using the CLN7FF will commence in ~Q2 2018, so, the first "7-nm" ICs will show up in commercial products in the second half of next year. When compared to the CLN16FF+, the CLN7FF will enable chip developers to shrink their die sizes by 70% (at the same transistor count), drop power consumption by 60% or increase frequency by 30% (at the same complexity).

https://www.anandtech.com/show/11337/samsung-and-tsmc-roadmaps-12-nm-8-nm-and-6-nm-added/2
 
#13 ·
Infinity fabric is coming to GPUs?
thumb.gif


Now only if they can improve performance per watt to be in line with Nvidia.
 
#14 ·
Quote:
Originally Posted by Slomo4shO View Post

Infinity fabric is coming to GPUs?
thumb.gif


Now only if they can improve performance per watt to be in line with Nvidia.
That will only happen if they get rid of their hardware scheduler. That comes at the cost of lesser DX12 performance, and a lot more driver work. I doubt they'll do that.
Or another option is to revamp GCN completely to not let so many stream processors idle all the time, particularly because developers are not optimizing for their hardware.
 
#17 ·
Quote:
Originally Posted by Pyrotagonist View Post

Quote:
Originally Posted by NightAntilli View Post

If anyone doesn't know, this image is actually an NVIDIA creation:

https://www.tweaktown.com/news/58295/nvidia-shift-multiple-gpus-future-geforce-cards/index.html

Of course, it's very likely that Navi will use a similar design.
AMD announced before both Intel and nVidia that they would also employ the CCX design in their GPU's. Intel had already followed suit after Zen's release and nVidia announced not long after that they would also take the same approach. Making enormous monolithic dies is simply not feasible any longer.

We can already see that Intel no longer disclose turbo speeds simply because they can't guarantee their chips will run at a certain frequency due to defects and heat problems, and yet we see that AMD's design allows the production of cheap many cores by 'gluing' smaller cores together.

The question is not what approach they take but which glue they use. Intel so far have applied pigeon, chicken and other bird poop. I hope nVidia don't learn from Intel and use camel dung as that is reserved for Intel only.
 
#18 ·
Quote:
This rumor unfortunately confronts the hard reality of the foundry since TSMC is the first to have started production "at risk" on its 7nm process and intends to launch by the beginning of 2018 the classic production, with an increase in volume all the way it is unlikely that it will be accessible to AMD or Nvidia for a "big" GPU at least before the end of 2018. The 7nm of GlobalFoundries or Samsung will arrive even after, with volume production not beginning before the second half of 2018, with EUV from the start at Samsung.
http://www.hardware.fr/news/15264/navi-7nm-mi-2018-peu-probable.html

Like Marc from hardware.fr, I don't believe we will 7nm GPU products since 2019. So yeah, looks like this rumor is false.
 
#21 ·
Talk about getting reference Vega 10 cards out in reasonable supply, custom vega cards(It's already mid october and no news), Vega Nano Vega 11, 20 out first before you talk about Navi, AMD.

Seems like they're always talking and bigging themselves one step ahead of what they can actually deliver. That was what happened for Vega. Poor Volta= Late pascal tier Vega.

You know a product isn't good when MSI and Gigabyte are not gonna make custom cards, yet come up with the 101st variant of the 1080Ti which is a year old but still predicted to deliver more and sell more than Vega will ever sell.

Vega is a money sink and just like the original fury, very little improvements are going to be seen. Heck, they didn't get much improvements over a 1080 for LC reference and we're talking air here. Overpriced overhyped, delayed and that's what will happened for Navi.
 
#22 ·
Quote:
Originally Posted by vida5z View Post

Talk about getting reference Vega 10 cards out in reasonable supply, custom vega cards(It's already mid october and no news), Vega Nano Vega 11, 20 out first before you talk about Navi, AMD.

Seems like they're always talking and bigging themselves one step ahead of what they can actually deliver. That was what happened for Vega. Poor Volta= Late pascal tier Vega.

You know a product isn't good when MSI and Gigabyte are not gonna make custom cards, yet come up with the 101st variant of the 1080Ti which is a year old but still predicted to deliver more and sell more than Vega will ever sell.

Vega is a money sink and just like the original fury, very little improvements are going to be seen. Heck, they didn't get much improvements over a 1080 for LC reference and we're talking air here. Overpriced overhyped, delayed and that's what will happened for Navi.
Gz on your first post. Vega might be delayed and undersupplied, but we're already seeing significant driver improvements even as loads of features aren't being utilized. I'd ask why you're assigning false flaws to a product that has plenty of real ones, but I already know the answer.
 
#24 ·
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top