test

neoscrypt_vliw.cl v3

HD6970: 19.5KH/s to 50.5KH/s and no HW errors!

Thanks to a much better global memory management with burst writes. Also tried a trick with copying a workgroup buffer to local memory and writing back asynchronously to global memory through async_work_group_copy(), but it didn’t work out (36KH/s and 15% of HW errors). A 2.5x speed increase is something to celebrate anyway. Anyone cares to donate towards drinks, whores and blackjack? ;)

How it use? Don’t work for me with cgminer 3.7.8. Cards in rig: 6950,6950,6870,5870,5870. Drivers 13.12, Windows 7 x64 sp1

Post more information.

For example:

Windows version and if it is 64 or 32-bit

AMD Catalyst drivers

what worksize are you using?

have you set the following environmental vars?

GPU_MAX_ALLOC_PERCENT=100

GPU_USE_SYNC_OBJECTS=1

also, upgrade to 3.7.7c or use sgminer

Yep, i forgot this settings: GPU_MAX_ALLOC_PERCENT=100 , GPU_USE_SYNC_OBJECTS=1… on sgminer5 now works fine, thanks! 24 kh/s on radeon 6870 :))

neoscrypt_vliw.cl v2

It’s 19.5KH/s now on a HD6970. FastKDF and BLAKE2s have been cleaned up and optimised, memory requirements reduced.

Yeah, I’ve mentioned this in my white paper. Not sure if it’s of any use for mining.

It is, but that’s not what concerns me now. With FastKDF removed, the kernel gets reduced in size by ~60% and outputs 30KH/s.That’s a big overhead, but not critical and I’ve expected more out of ChaCha + Salsa. With ChaCha only enabled, it’s 58KH/s and with Salsa only = 56KH/s. Scalar Salsa isn’t supposed to be about as fast as vectorised ChaCha. It’s clearly scalar because the AMD compiler isn’t really smart and the kernel size is about double of ChaCha only size. Anyway, there is a huge bottleneck somewhere and it needs to be identified.

Don’t work for me now. Cards in rig: 6950,6870,5870, miner 3.7.7b. Screen:

For those with old Radeon cards. This is my current OpenCL kernel: neoscrypt_vliw.cl

It is optimised to some extent for VLIW4/VLIW5. I get 17.5KH/s with it on a HD6970. That’s not much, but still better than 6KH/s with the default kernel.

How it use? Which miner?

and how about old cards 5800 6800 6900 series?

Please add support for old cards, 5800 6800 6900 series. Please!

test