Friday, February 8th 2019

No AMD Radeon "Navi" Before October: Report

AMD "Navi" is the company's next-generation graphics architecture succeeding "Vega" and will leverage the 7 nm silicon fabrication process. It was originally slated to launch mid-2019, with probable unveiling on the sidelines of Computex (early-June). Cowcotland reports that AMD has delayed its plans to launch "Navi" all the way to October (Q4-2019). The delay probably has something to do with AMD's 7 nm foundry allocation for the year.

AMD is now fully reliant on TSMC to execute its 7 nm product roadmap, which includes its entire 2nd generation EPYC and 3rd generation Ryzen processors based on the "Zen 2" architecture, and to a smaller extent, GPUs based on its 2nd generation "Vega" architecture, such as the recently launched Radeon VII. We expect the first "Navi" discrete GPU to be a lean, fast-moving product that succeeds "Polaris 30." In addition to 7 nm, it could incorporate faster SIMD units, higher clock-speeds, and a relatively cost-effective memory solution, such as GDDR6.
Source: Cowcotland
Add your own comment

135 Comments on No AMD Radeon "Navi" Before October: Report

#126
eidairaman1
The Exiled Airman
tfdsaf
This is the new "AMD drivers suck" myth, propagated by ignoramus or actual Nvidia shillboys.

FIRST AND FOREMOST, GCN 1 to GCN 4 is a COMPLETELY different architecture. Vega is a completely new architecture to top it off as well, internally they never even called Vega GCN, they always called in Vega, as a codename and later as official monkier.

Nvidia's latest Turning architecture is basically their old GTX 400 architecture on steroids. Even their GTX 400 architecture was an evolution from their GTX 200 architecture, which was Nvidia's foray into unified shaders. What has happened is that Nvidia has been continually been improving their GTX 200 architecture and over the last 12 years reached the current architecture called Turing, which is in essence their 12 years old architecture from their GTX 200 series.

Just because Nvidia have been giving different code names to their architectures, doesn't mean its a new architecture. And alternatively AMD giving the same GCN name doesn't mean its the same architecture.

So this garbage that has been propagated how GCN is the same architecture, while Nvidia magically creates brand new 100% different architecture every two years is a mega ton of horse manure. Nvidia's current architecture dates back to their GTX 200 series. That is 12 years! Is it fair to call it the same architecture? Well according to Nvidishillboys then yes, Turing the GTX 200 arch is exactly the same.

If anything AMD have been reinventing their architectures a lot more, which is part of the problem. They've never stuck long term like Nvidia to an arch and improved it over time, reiterative process over 12 years. They've done a lot more clean states.
GCN is the only 1 so far with revisions.
Posted on Reply
#128
notb
Super XP
And lastly, how accurate is this? Found it on Wccftech by a poster lol,
Possibly not very much given the weird naming and too small differences between models.
Remember Navi was meant to be a mainstream gaming platform, not high-end.

And another question one could ask: how interesting it this? Not much.
On the graph you've shown "Navi 12" performs like a Vega 64, while "Navi 10" would be a Radeon VII. "Navi 8" could be a polished 7nm GCN, but it's also now far from successful Radeon VII overclocks.

Performance is not a problem. AMD has that.

What would actually be interesting is these 3 bars being not FPS but peak Watts.
Nvidia's bars are actually pretty close to TPU results...
Posted on Reply
#129
FordGT90Concept
"I go fast!1!11!1!"
notb
What would actually be interesting is these 3 bars being not FPS but peak Watts.
Nvidia's bars are actually pretty close to TPU results...
That makes a lot more sense.
Posted on Reply
#130
notb
FordGT90Concept
That makes a lot more sense.
Assuming it is [W], it still makes no sense why "Navi 12" would draw less than Navi 8 - unless AMD changed naming (bigger die, smaller number - like Nvidia does).

200W is more or less the limit Sony is going to accept for "big" PS5 (PS5 Pro?).
Assuming they'll dump optical drive, PS5 could be half the size of PS4 Pro. I'm sure they'll go this route - users are complaining about size of current consoles.
ASRock DeskMini GTX is 2.7l (compared to PS4's 5.3l) and can come with a GTX 1080.
Posted on Reply
#131
FordGT90Concept
"I go fast!1!11!1!"
Polaris 20 = RX 580/RX 570
Polaris 21 = RX 560

AMD tends to start with the biggest chip first so, yeah, the numbering makes sense in that context (lower number = bigger chip = more power).

I don't put much faith in the Navi # they provide though.
Posted on Reply
#132
Midland Dog
Navi gonna give you up
Navi gonna let you down
Navi gonna run around and desert you
Navi gonna make you cry
Navi gonna say goodbye
Navi gonna tell a lie and hurt you
Posted on Reply
#133
Totally
Just realized this article was wrong by nearly 2Qs. Who crystal balled this one? So I could relegate them to trash tier.
Posted on Reply
#134
Super XP
Totally
Just realized this article was wrong by nearly 2Qs. Who crystal balled this one? So I could relegate them to trash tier.
Well we all now know Navi is designed for PC Gaming. AMD did a brilliant job with it. Just don't like the Blower Style Coolers.
Posted on Reply
Add your own comment