A personal update, I'll be traveling for a longer while (weeks, 1month), starting tomorrow. I won't be online easily or often. My computers & GPUs will be temporarily off, and no active GpuOwl development for a bit, unfortunately. So if it seems I don't answer etc, that traveling may be the reason.

[QUOTE=SELROC;496752]My 332M exponent was > 60 days of work, now at 40 days of work, I am keeping ver. 3.5 for speed. I am going to test the new ver. with some smaller exponent.[/QUOTE]

I correctly understand that ETA for 332M is 40 days on v.4 on RX 580?

[QUOTE=Lorenzo;497280]I correctly understand that ETA for 332M is 40 days on v.4 on RX 580?[/QUOTE]




Ciao.
The initial ETA for 332M exponent was > 60 days with version 3.5, so I am continuing with ver. 3.5

[QUOTE=SELROC;497281]Ciao.
The initial ETA for 332M exponent was > 60 days with version 3.5, so I am continuing with ver. 3.5[/QUOTE]

Hi, Valerio!)
ahhh, Ok) Thank you for the clarifying :smile:

[QUOTE=kriesel;494281]Please provide any references (links or tools) for AMD DP/SP ratios. I've found that GPU-Z gives values for NVIDIA but not for AMD.[/QUOTE]

Sorry I'm so late on this, but I just found the source.
Google 'board name/number and techpowerup' as in:
vega 64 techpowerup
Look for the entry that has techpowerup in the url and the word specs in the title.
The SP/DP is toward the middle right of the page and marked 'Theoreticsl Performance'. You should ignore the nice graph titled 'Relative Performance' because IMHO the data there seems misleading.
Techpowerup are the guys who make GPU-Z, BTW, and their DB seems quite complete.

[QUOTE=tServo;497363]Sorry I'm so late on this, but I just found the source.
Google 'board name/number and techpowerup' as in:
vega 64 techpowerup
Look for the entry that has techpowerup in the url and the word specs in the title.
The SP/DP is toward the middle right of the page and marked 'Theoreticsl Performance'. You should ignore the nice graph titled 'Relative Performance' because IMHO the data there seems misleading.
Techpowerup are the guys who make GPU-Z, BTW, and their DB seems quite complete.[/QUOTE]
Thanks!

RX-480 1/16 [URL]https://www.techpowerup.com/gpu-specs/radeon-rx-480.c2848[/URL]
RX-550 1/16 [URL]https://www.techpowerup.com/gpu-specs/radeon-rx-550.c2947[/URL]

I've been a fan of GPU-Z for quite a while, except for the part about it failing on AMD sensors when Windows Remote Desktop is used; frequencies and temperature specifically. I've reported it repeatedly. I see 2.11 is now out. Yup, still broken in 2.11.0.

[QUOTE=Lorenzo;497280]I correctly understand that ETA for 332M is 40 days on v.4 on RX 580?[/QUOTE]


gpuOwl performance: the check takes 5 sec. for a 300M exponent, and it takes 9 sec. for a 332M exponent.


BTW, I am now running a newer kernel, 4.19-rc7 which seems a bit faster.

[QUOTE=SELROC;498119]
BTW, I am now running a newer kernel, 4.19-rc7 which seems a bit faster.[/QUOTE]

With amdgpu-pro? what version?

[QUOTE=preda;498125]With amdgpu-pro? what version?[/QUOTE]


kernel 4.19-rc7
amdgpu-pro 18.20

openowl version 3.5

[QUOTE=preda;498125]With amdgpu-pro? what version?[/QUOTE]

[QUOTE=SELROC;498119]gpuOwl performance: the check takes 5 sec. for a 300M exponent, and it takes 9 sec. for a 332M exponent.


BTW, I am now running a newer kernel, 4.19-rc7 which seems a bit faster.[/QUOTE]




Update: by putting the GPU in a Gen3 16x slot, the timing is the same, but the check becomes faster by 1-2 sec.


I think the gpu-cpu communication is faster, this affects GEC speed.