Read more.This date was put forward at the recent Q2 2019 earnings call from Samsung.
Read more.This date was put forward at the recent Q2 2019 earnings call from Samsung.
I can not wait for this. I really do fancy a giggle. Judging by my Vega64, I'm guessing you'll have to attach the "EXTREME MOBILE GAMING BACKPACK" to the phone which pipes coolant around it and supplies the necessary battery life whilst also serving as a giant heatsink.
Basically, you're gonna look like a ghostbuster whilst gaming one the train.
I want in.
This is good news since apparently different AMD divisions don't share funding the GPU department need to raise their own capital.
Also this co-development could mean more R&D in things like power efficiency which might translate to the discrete GPU side.
Vega is actually really efficient when its not overvolted or has clocks that surpass its 'comfort zone'.
This is quite evident from 2500u, 2700u, 3500H and 3700H APU's for example.
Besides, the article says the following:
"Samsung will be paying AMD royalties and license fees for the use of Radeon RDNA tech."
So, Vega won't be used, but instead, Navi will be.
It will be interesting to see how Navi does in the APU segment.
Given the overall improvements it exhibits over the original GCN, it should do well.
Personally, I'm ok with my Vega56 in my Acer Helios 500 laptop.
Its cool, efficient and really powerful (plus the laptop is quiet even when stressed).
Considering this 'beast' has 2700 CPU (full desktop chip - socketed one) inside, I think I'm happy.
Acer would do better to release an updated BIOS for the laptop so I can eventually put in Zen 2 or even Zen 3 inside... or maybe AMD will.
But please continue to tell me I'm wrong... because it wasn't clear enough it was a joke.
Also, they won't be using Navi unless they are mental. The amount of juice this needs to sip means they'll almost certainly be designing a very low power GPU from scratch and using some AMD IP to lubricate the process. To base this at all on an existing desktop or laptop chip would be asking for trouble. You're looking at orders of magnitude different power consumption for starters, never mind the rest. It's a totally different proposition and once you start changing one thing in an existing design you end up chasing all the knock on effects which then screws with something someone else is working on and it just gets out of hand.
Nvidia merged their desktop & mobile graphics years ago, it can be done and looking at their low power 12W APUs that include 4 power hungry Zen+ cores they probably just need a lowering of clock speeds so they can drop the voltage down to be in the ballpark.
Are we done saying you are wrong now? (clearly your slapstick imagery was too subtle for the interwebz, but I think if I had to carry my Vega56 around all day I would put my back out )
What about size? Packaging? Different architecture and memory types in mobile? Desktop and laptop being similar with tweaks makes sense but mobile is a different kettle of fish, surely? Or am I overcomplicating what is actually very simple?
I think I need a rubber stamp gif with "JOKE" for the sake of my sanity....
Overcomplicating, sort of. This is selling IP, so a collection of gates to be laid down in some silicon with some help with connecting to Samsung's on chip fabric. It will need Samsung's memory controllers to feed it a certain amount of data. But hey, Samsung, AMD and ARM are all founding members of the HSA so hopefully by now this is all smooth as silk (I almost typed that with a straight face... almost). http://www.hsafoundation.com/members/
A rubber stamp gif? Nah, this is the Internet, someone will just take the gif the wrong way.
philehidiot (09-08-2019)
I am curious about the heat output on mobile departament. Because AMD has a bad reputation about heat output in GPU market.
There are currently 1 users browsing this thread. (0 members and 1 guests)