Last edited by a moderator:
AMD and Samsung Announce Strategic Partnership in Mobile IP
anexanhume
Veteran
Years ago when they were cash strapped and mobile hadn’t taken off yet.
D
Deleted member 13524
Guest
Is this putting an end to rumors of Samsung developing their own GPU?
It was certainly a surprise though. Samsung has been trying to get GPU performance parity with Qualcomm for years especially for their flagships shipping outside the US, and with ARM they've never been successful at it.
Actually in Samsung's case the flagships (Galaxy S line) are the ones that get better sales numbers by far, so it's a really big win nonetheless.
It was certainly a surprise though. Samsung has been trying to get GPU performance parity with Qualcomm for years especially for their flagships shipping outside the US, and with ARM they've never been successful at it.
Actually in Samsung's case the flagships (Galaxy S line) are the ones that get better sales numbers by far, so it's a really big win nonetheless.
Bondrewd
Veteran
Yes.
Ye, especially given that ImgTec was for sale and SS did nothing.
Nebuchadnezzar
Legend
It's not a rumour when they openly boast about it: https://samsungsemiconductor-us.com/labs/labs/austinrd.html
Anyway I'm a bit confused as where Samsung GPU ends and AMD IP starts, but it looks like something akin to what Apple is doing with IMG licenses.
Bondrewd
Veteran
Or not.
Probably not.
D
Deleted member 13524
Guest
Does it need to be the same GPU?
Couldn't they be using Radeon for the higher-end SoCs and their own architecture for lower-end ones, in order to cut away from Mali graphics for good?
AMD's IP should be a much more proven architecture than Samsung's in-house one, and in case the later underperforms it wouldn't be such a big problem to include it into a lower-end exynos for lower margin devices.
Bondrewd
Veteran
I kind of wonder why would they do that.
Both G76 and G77 doesn't seem too bad.
Nebuchadnezzar
Legend
Probably yes.
I don't think they have the man-power for two GPU designs.
It looks likely this deal happened 3 years ago and is merely just made public today.
Bondrewd
Veteran
We don't know.
But I bet on "no".
Nebuchadnezzar
Legend
I reiterate my yes.
Bondrewd
Veteran
Well, opinions.
I wonder when they're going to ship it.
D
Deleted member 90741
Guest
Samsung is pushing hard in the Automtive industry. One of the more common complaints however was the GPU on their SoCs is a bit behind when compared to the NVIDIA Xavier and Parker.
You have to consider that in Automotive the power envelope while generally not so high as desktop is much higher than smartphone and tablets.
One of the things that immediately comes to my mind is the integration of compute oriented architecture of AMD's Radeon in Samsung's Exynos Auto.
The requirement for compute power in Automotive is growing exponentially due to ADAS and Autonomous vehicles running a wide variety of inferencing and DNN algorithms and NVIDIA is quite strong in this area with their Tegra SoCs.
And it would not make sense for Samsung to license it from the competitor NVIDIA who is making Tegra SoCs for the GPU.
Couple this to the fact that AMD has a strong open source drivers and a budding compute stack it could be something to watch out.
You can guess who is the only closed source GPU vendor. Intel has massive open source components and DNN/compute stack as well.
You have to consider that in Automotive the power envelope while generally not so high as desktop is much higher than smartphone and tablets.
One of the things that immediately comes to my mind is the integration of compute oriented architecture of AMD's Radeon in Samsung's Exynos Auto.
The requirement for compute power in Automotive is growing exponentially due to ADAS and Autonomous vehicles running a wide variety of inferencing and DNN algorithms and NVIDIA is quite strong in this area with their Tegra SoCs.
And it would not make sense for Samsung to license it from the competitor NVIDIA who is making Tegra SoCs for the GPU.
Couple this to the fact that AMD has a strong open source drivers and a budding compute stack it could be something to watch out.
You can guess who is the only closed source GPU vendor. Intel has massive open source components and DNN/compute stack as well.
D
Deleted member 13524
Guest
In 2018 alone Samsung SLSI launched one SoC with a Mali G71 MP1, one with a Mali G71-MP2, one with a Mali G72 MP3, one with a Mali G72 MP18 and a supposed die-shrink with a Mali T720.
In their ARMv7 days, they'd release SoCs with Mali and PowerVR GPUs in parallel.
The team adapting AMD's Radeon IP GPU to their SoC could be a different team than the one working on their in-house GPU.
Besides, if Samsung's in-house GPU started being developed 7 years ago and so far we haven't seen in live, there's a chance the architecture is in development hell and Samsung is killing it off.
The announcement specifically states it's for the mobile market.
Bondrewd
Veteran
Why would they bother with this licensing deal assuming their in-house stuff is anywhere decent?
Valhall seems decent, too.
Or Intel. They licensed from both Nvidia and AMD. Doesn't mean much w.r.t. their internal teams. Sometimes IP is just "don't sue me" shielding.
D
Deleted member 13524
Guest
Intel licensing IP from nvidia and AMD to produce their Gen GPUs was announced as a license agreement. No architecture nor specific GPU naming was given, as it's probably just Intel buying the rights to cover general patents like "shader processors".
This announcement carries a completely different language. Specifically "Radeon" and "RDNA" are named:
AMD said:
I know @Nebuchadnezzar knows a whole lot more than us mortals regarding this market, but it really doesn't look like this is a case where Samsung negotiated with AMD a patenting deal 3 years ago and it's announcing it only now.
If this is not Samsung implementing a scaled-down Navi in their SoCs, then the language used is very deceptive, to say the least.
Funny thing is this is something nvidia set out to do 6 years ago, but has been unsuccesful so far.
Similar threads
