Rebranded GCN 1.0 AMD Radeon 520 DGP in my HP laptop supported in "amdgpu"?

[agd5f]

With 4.15 kernel and above, I'm getting an error about firmware with amdgpu driver. I have the firmware available in /lib/firmware/ .

drm:amdgpu_ucode_get_load_type [amdgpu]] ERROR Unknown firmware load type

I have no clue, where to look.

Message is harmless, but fixed in this patch:
https://patchwork.freedesktop.org/patch/246240/

[deepclutch]

With 4.15 kernel and above, I'm getting an error about firmware with amdgpu driver. I have the firmware available in /lib/firmware/ .

drm:amdgpu_ucode_get_load_type [amdgpu]] ERROR Unknown firmware load type

I have no clue, where to look.

[agd5f]

Any idea if features like freesync is supported in this AMD 520 card? Also, is this card supported in AMDGPU-PRO drivers? I read that many features are added and most of these are not available with early cards like these with GCN1.0/1.1.

Your chip has no display hardware. All displays are connected to the integrated GPU. The dGPU in your system is just for offscreen rendering so display features like freesync are not possible on your platform.

[deepclutch]

Any idea if features like freesync is supported in this AMD 520 card? Also, is this card supported in AMDGPU-PRO drivers? I read that many features are added and most of these are not available with early cards like these with GCN1.0/1.1.

[agd5f]

Sun/Hainan are SI (GCN 1.0) asics with no display or multi-media engines, the rest is the same as other SI parts. As for hybrid graphics, the dGPU is completely powered down (i.e., disappears from the bus) when not in use. The driver stays loaded however because it controls when it gets powered up/down.

[deepclutch]

For dynamic switchable graphics, do we still need vga-switcheroo, especially if used in with wayland (X on wayland in Ubuntu 17.10)?
https://www.kernel.org/doc/html/v4.1...witcheroo.html .

I read the below and it seems AMD dynamic switchable graphics also supports manual mode? possible in Linux?

The details of AMD’s Dynamic Switchable Graphics are similar in practice to Optimus, but with a few differences. First, AMD always has both the IGP and GPU driver loaded, with a proxy driver funneling commands to the appropriate GPU. Where NVIDIA is able to completely power off the GPU under Optimus, AMD has modified their GPUs so that the PCI-E bus is isolated from the rest of the chip. Now when the GPU isn’t needed, everything powers down except for that PCI-E connection, so Windows doesn’t try to load/unload the GPU driver. The PCI-E link state gets retained, and a small amount (around 50mW) is needed to keep the PCI-E state active, but as far as Windows knows the GPU is still ready and waiting for input. AMD also informed us that their new GPUs use link adapter mode instead of multi adapter mode, and that this plays a role in their dynamic switchable graphics, but we didn’t receive any additional details on this subject.

As far as getting content from the dGPU to the display, the IGP always maintains a connection to the display ports, and it appears AMD’s drivers copy data over the PCI-E bus to the IGP framebuffer, similar to Optimus. Where things get interesting is that there are no muxes in AMD’s dynamic switchable graphics implementations, but there is still an option to fall back to manual switching. For this mode, AMD is able to use the display output ports of the Intel IGP, so their GPU doesn’t need separate output ports. With the VAIO C, both dynamic and manual switching are supported, and you can set the mode as appropriate.

https://www.anandtech.com/show/4839/...s-technology/2

[bridgman]

BTW, it is pathetic to see major laptop brands like HP, Dell pushing old, outdated and obsolete mobile GPU's even in their newest laptops. I checked and both Dell 5567 or HP models usually carry 4GB R7 445 or 2GB AMD Radeon 520. This is how it is. May be these laptops are not for gaming; But, they are pushing really old tech.

In fairness, if we made a newer chip in that size range I'm sure the OEMs would use it instead. The problem is that we all have to make best estimates of what the market will be like in a few years and design chips for that forecast... and we thought the very low end dGPU market would have gone away by now since we have been offering APUs with more performance for a while.

In practice, the OEMs who did pick up our APUs frequently configured them in what I would call a worst-case configuration - single channel memory and sharing the power budget with a low end dGPU - so the market didn't shift as we expected and low-end dGPUs remained in demand. The smallest dGPU we make today is Lexa, which is ~2x the size AND >>2X the performance.

A smart OEM would have configured one of our APUs running at full power (rather than reserving ~20W for a low end dGPU and forcing the APU to lower clocks) with fast dual channel memory and picked up a big chunk of the market even with APUs from a couple of years ago. I don't really know why that hasn't happened yet.

[deepclutch]

Looks like the 520 is a 5-CU configuration while 530 is 6-CU:

http://www.amd.com/en/products/graphics/radeon-520

The shader core would be GCN 1.0 while some of the other blocks would be newer and inline with the larger CI parts IIRC.

I'm not sure where marketing draws the line between R3/R5/R7 categories - although it appears they may be moving away from that naming convention. I believe the R3/R5/R7 rating was a performance measurement rather than an indicator of technology generation, so based on CU count and memory complement I would expect R5 rather than R7.

You are correct. I don't know if this info is absolute, but when I installed AMD 520 drivers to autodetect GPU and install on Windows 10, the GPU is shown as R5 M330. (I uninstalled the driver from AMD). When I installed the AMD/Intel Switchable driver from HP, the GPU is correctly shown as AMD R520. What this means, perhaps is R5 M330 may be very similar to this AMD R520..may be?

Regarding Vulkan support in Linux, is there something to be installed? I saw there is something called "radv" https://github.com/airlied/mesa/tree...src/amd/vulkan

OR is there official vulkan provided for Linux by AMD. Thanks again.

[deepclutch]

Looks like the 520 is a 5-CU configuration while 530 is 6-CU:

http://www.amd.com/en/products/graphics/radeon-520

The shader core would be GCN 1.0 while some of the other blocks would be newer and inline with the larger CI parts IIRC.

I'm not sure where marketing draws the line between R3/R5/R7 categories - although it appears they may be moving away from that naming convention. I believe the R3/R5/R7 rating was a performance measurement rather than an indicator of technology generation, so based on CU count and memory complement I would expect R5 rather than R7.

Definitely, the gaming performance as seen by the reviews is below R7 445. The doubt was whether this card gets into the list of supported GPU's in the newer "amdgpu" opensource driver. Otherwise, I would've to satiate with the old radeon. Now that it is clear that the card is supported, although the performance benchmarks are not available.Thanks for the replies. debianxfce Definitely, I think I should try the PPA's and also give bug reports to the amdgpu BTS hopefully, to get some attention for this model.

BTW, it is pathetic to see major laptop brands like HP, Dell pushing old, outdated and obsolete mobile GPU's even in their newest laptops. I checked and both Dell 5567 or HP models usually carry 4GB R7 445 or 2GB AMD Radeon 520. This is how it is. May be these laptops are not for gaming; But, they are pushing really old tech.

[bridgman]

Looks like the 520 is a 5-CU configuration while 530 is 6-CU:

http://www.amd.com/en/products/graphics/radeon-520

The shader core would be GCN 1.0 while some of the other blocks would be newer and inline with the larger CI parts IIRC.

I'm not sure where marketing draws the line between R3/R5/R7 categories - although it appears they may be moving away from that naming convention. I believe the R3/R5/R7 rating was a performance measurement rather than an indicator of technology generation, so based on CU count and memory complement I would expect R5 rather than R7.