[Dev] NeoScrypt GPU Miner - Public Beta Test
-
There is no point of using HD5000 and HD6000 series with v14.x drivers. They are all optimised for the GCN cards. I’m fine with v13.1 and SDK v2.6. v13.4 also seems good. v13.8beta was crap. Didn’t even bother with v13.12 and newer.
-
There is no point of using HD5000 and HD6000 series with v14.x drivers. They are all optimised for the GCN cards. I’m fine with v13.1 and SDK v2.6. v13.4 also seems good. v13.8beta was crap. Didn’t even bother with v13.12 and newer.
I’m pretty sure I have 13.12 on my rig with the 69xx cards and maybe this is why I only saw a 1.25kh/s increase in hash with the v2 over v1?
-
There is no point of using HD5000 and HD6000 series with v14.x drivers. They are all optimised for the GCN cards. I’m fine with v13.1 and SDK v2.6. v13.4 also seems good. v13.8beta was crap. Didn’t even bother with v13.12 and newer.
What if we have a mult-card setup with both GCN and pre-GCN cards?
-
Here you guys go: https://ottrbutt.com/tmp/neoscrypt.cl
-
[2014-11-10 22:25:50] Started cgminer 3.7.8
[2014-11-10 22:25:50] Network diff set to 551K
[2014-11-10 22:25:50] Stratum from pool 0 detected new block
[2014-11-10 22:25:50] Stratum from pool 0 requested work restart
[2014-11-10 22:25:52] Error -11: Building Program (clBuildProgram)
[2014-11-10 22:25:52] :486:16: error: attribute requires 1 argument(s)
__attribute__((reqd_work_group_size(WORKSIZE, 1, 1)))
^
:512:11: warning: incompatible pointer types passing ‘ulong16 __attribute__((add
ress_space(1))) *’ to parameter of type 'uint16 __attri
[2014-11-10 22:25:52] Failed to init GPU thread 0, disabling device 0
[2014-11-10 22:25:52] Restarting the GPU from the menu will not fix this.
[2014-11-10 22:25:52] Try restarting cgminer.
Press enter to continue: -
Here you guys go: https://ottrbutt.com/tmp/neoscrypt.cl
Thank you again Wolf! My hash rates are so much higher now!
-
Here you guys go: https://ottrbutt.com/tmp/neoscrypt.cl
Wolf0, that one didn’t help on pre-GCN cards, at least on mine. Hashrate went from 16.5 Kh/s to 13 Kh/s. Not that you said it would help old cards, but for those of us with older hardware, this appeared to be a step in the wrong direction.
-
What if we have a mult-card setup with both GCN and pre-GCN cards?
insanid, at the moment, optimize for the GCN card and whatever you get with the pre-GCN card is gravy. Or mine something else with the pre-GCN cards. DMD is good to mine. You won’t get a lot of coins, but at least it’ll cover the electric bill. At the moment, mining neoscrypt with pre-GCN cards is a losing prospect. Per card, per day, you’re looking at maybe 1 FTC, 200 PXC, or maybe 800 UFO. You’ll be lucky to hit 10,000 satoshi in value (per card). Mine Diamond. You’ll only get about 1 a day, but it’s worth 40,000+.
I’m not trying to bash neoscrypt, and I’m certainly not trying to bash the developers who have made some really positive strides over the last week or so. I’m a long time miner of PXC, and I’m really hoping to continue mining it, and now FTC and UFO also. But right now, I’m not buying new graphics cards, and just doing the math, neoscrypt on pre-GCN hardware doesn’t make any sense. Hopefully that will change.
-
I am getting a lot of hardware errors with the new kernel by wolf0 on all my miners. Still trying to find optimal settings for it.
-
Drops my GT 630M down to 2Kh/s from 3.
So yeah. For your older cards or cutdown cards (like mine), the first kernal is better it seems.
Let’s see what everyone can get on newer gear. I have nothing to test on currently as I have the gpu’s but no mobo or risers.
-
BTW, wolf’s kernels are only optimized for AMD cards. No nvidia.
Using Wolf0’s newest kernel:
On my AMD 290x (Hynix) I have tried w32, w48, w64, w128, w256 and I12-15 and I get nothing but hardware errors.
Same thing with AMD HD7950, AMD HD7850, AMD HD7870
Using sgminer-dev binaries wolf compiled on 11-02
Edit: Nothing I have tried can get the new wolf0 kernel to work with sgminer. Too many hardware errors. I am going back to previous kernel for the time being until this new kernel is sorted out.
-
insanid, at the moment, optimize for the GCN card and whatever you get with the pre-GCN card is gravy. Or mine something else with the pre-GCN cards. DMD is good to mine. You won’t get a lot of coins, but at least it’ll cover the electric bill. At the moment, mining neoscrypt with pre-GCN cards is a losing prospect. Per card, per day, you’re looking at maybe 1 FTC, 200 PXC, or maybe 800 UFO. You’ll be lucky to hit 10,000 satoshi in value (per card). Mine Diamond. You’ll only get about 1 a day, but it’s worth 40,000+.
I’m not trying to bash neoscrypt, and I’m certainly not trying to bash the developers who have made some really positive strides over the last week or so. I’m a long time miner of PXC, and I’m really hoping to continue mining it, and now FTC and UFO also. But right now, I’m not buying new graphics cards, and just doing the math, neoscrypt on pre-GCN hardware doesn’t make any sense. Hopefully that will change.
Yeah I agree with you on pre-GCN cards (even with the faster pre-GCN optimized kernel from ghostlander); however, Wolf said he would be willing to try his hand at optimizing the kernel for pre-GCN AMD cards. I have been trying to get a test machine up and running with a 6950 for him to use, but I’ve been busy with other shit IRL, so it’s been hard for me to dedicate time to getting it up and running.
-
BTW, wolf’s kernels are only optimized for AMD cards. No nvidia.
Using Wolf0’s newest kernel:
On my AMD 290x (Hynix) I have tried w32, w48, w64, w128, w256 and I12-15 and I get nothing but hardware errors.
Same thing with AMD HD7950, AMD HD7850, AMD HD7870
Using sgminer-dev binaries wolf compiled on 11-02
Edit: Nothing I have tried can get the new wolf0 kernel to work with sgminer. Too many hardware errors. I am going back to previous kernel for the time being until this new kernel is sorted out.
same for me 280x 14.9beta2 nothing but HW errors! tried w48-512 I 13-15
-
Did you delete the .bin file before starting up sgminer ?
I found the kernal wouldn’t work with either my versions of cg, but it did work on sg for me.
Swapped out the .cl and deleted the bin files. Started it up, sgminer put together the new bin file, and now it starts like normal every time.
-
Did you delete the .bin file before starting up sgminer ?
I found the kernal wouldn’t work with either my versions of cg, but it did work on sg for me.
Swapped out the .cl and deleted the bin files. Started it up, sgminer put together the new bin file, and now it starts like normal every time.
Yes, all pre-existing bins were deleted beforehand. I use cgwatcher, and I have vbs + bat scripts that are scheduled to auto delete all the bins from the cg/sgminer folder(s) when I stop or restart the miner.
-
same for me 280x 14.9beta2 nothing but HW errors! tried w48-512 I 13-15
I am afraid we will have to wait for Wolf0, or someone else, to debug the code in his kernel to find out what’s going on. I couldn’t get a single worksize, thread-concurrency, thread number, or intensity setting to work on any of the AMD GCN cards I use.
Also, if you haven’t done so already, consider sending a small donation to Wolf0 for his work. He doesn’t have any obligation to release his personal kernels to us. His FTC address is posted a few pages back. He also accepts donations in LTC https://litecointalk.org/index.php?topic=17533.0
-
Ah ok…
It works on my nVidia, but slower.
This is all somewhat worrying.
Let’s wait on wolf I suppose.
-
Yeah, issues here too…
SGminer 4.2.2/5.0 (wolf0 build)
Run with the same bins, runs the exact same speed.
Delete the bins, 50% speed increase, but 100% hw errors, no accepted.
-
13-xx driver gives accepted share but speed equal to 14.9 drivers with previous kernel
-
Yeah, issues here too…
SGminer 4.2.2/5.0 (wolf0 build)
Run with the same bins, runs the exact same speed.
Delete the bins, 50% speed increase, but 100% hw errors, no accepted.
Running same bins will always give you same speed. As long as settings stay the same, cg/sgminer detects the existing bins and will skip using the AMD opencl compiler to compile an optimized bin from the new kernel.
