3.14.by forum

Also I've heard IvanG's say his ighashgpu only supports 8 cards as well. It's not true, actually. I haven't heard about systems with >8 GPUs so I have no idea will ighashgpu works with them or not. But there no hard limit of 8 GPUs in source code, AFAIR I used value like 32, so if CUDA layer will r...

I guess cost of this device (for +4 PCI-E) will be higher than cost of motherboard + case + PSU. As it always happens with non-standard solutions.

Yeah, it finally works.


Now there is an other subproblem -- as system config is GTX260+GTS250 it'll nice to have thread_n = 192 for first GPU and thread_n = 128 for second one .

Could you confirm if it works with -gpu_mask 1 and -gpu_mask 2 ? Yes, it works separately. Could you send me barswf.log when it crashes? It doesn't crash in fact, it's locks up with heart beating. Pressing Esc leading to "Stopping threads..." message but nothing happens. Log file isn't very interes...

Still problems with hybrid systems (i.e. GT200 + G92 GPUs at the same time).

If ighashgpu reports "Found N CAL device(s)" then CAL .dlls presents somewhere. For 64-bit system check out the /windows/syswow64 path as well or make whole disk search for "aticalrt.dll".

I guess it's only possible with nVidia's GPUs as they in fact can work without monitor attached if PhysX mode selected. While ATI GPUs always requires monitor (and I'm trying to use 4770+4850 setup with Win7). I doubt it'll possible to use 8400 as main GPU and extend desktop to 4870.

panda wrote:I tested BarsWF and it did work, however it uses 100% cpu aswell and that is not something I want it todo. What am I doing wrong Ivan? :/

Fully uninstall ATI driver. Particularly you don't wanna any atical*.dll's at system path.

Actually it's better to ask nVidia directly or atleast use their CUDA forum because it's not just password cracking related problem, it's CUDA layer itself cannot recognize 2nd GPU. I think there won't be any problems with Windows XP (because of different driver model) but if installing XP isn't an ...

Yes, it's very unlikely that OpenCL code will be better than CAL/IL code. There was Brook+ built on CAL/IL layer, now Brook+ replaced by OpenCL but it's still built on CAL/IL layer. It's like for CPUs -- we can write in asm (CAL/IL) or use C (OpenCL). If we're good in asm then asm code will always b...

ATI finally released OpenCL for GPUs .

drewy wrote:

IvanG wrote:Also you can try to add /cpudontcare switch

made no noticable difference.

I guess I need to increase block size for 5870. Unfortunately algorithm used makes it not as easy thing as it looks.
Actually I'd prefer to use bigger blocks even for 4850 but with 5870 it became mandatory.

5770 will be exactly the same (for hash cracking) as 4890 -- 800SP @ 850Mhz. 5750 as 720SP @ 700Mhz = 4850. But it's RV870, so better architecture anyway .

D3ad0ne wrote:Well that explains why the first four bytes are the same when running ighashgpu. Also explains why you can't set the first four characters.

Yeah, that's the main reason.

drewy, thanks for the tests. Results very close to expected though a bit lower. 5870 must be (0.85*1600 / 0.75*640) = 2.83(3) times faster than 4770, so it's should be 2680M for single MD5 & 821M for SHA1. Your SHA1 test runs for less than a second, so result isn't that accurate (it actually differs...

The only report I've got so far is for HD5850. It isn't direct information from HD5850 owner though, so errors possible.

For SHA1 5850 shows 690M (for comparison, it's 290M on 4770), so it's even better than I was expected. However system is pretty unstable, drivers aren't ready at all.

BarsMonster wrote:Here are the numbers:
CUDA: 26.73 MHash/sec
OpenCL: 25.38 MHash/sec (most of stuff is hardcoded, so actually it should be a bit slower)

Nice. I was more pessimistic about OpenCL .

Now only need to wait for ATI move.

Looks like the most cost-efficient system will be still the one based on good old HD4850's. Recently Tom's Hardware published 3 gamer configs, one of them with 4x4850 -- http://www.tomshardware.com/reviews/amd-enthusiast-crossfire,2421.html . Gamers wasn't too happy to see 4xGPUs in system but for G...

So which one is better? Don't get me wrong :). If you want to do hash cracking only , ATI Radeons, no doubts, best choice. But if you don't wanna stuck with hash cracking only then you should consider nVidia's GPU's even if their peak performance lower than ATI's ones. That's my point. There is nea...

Anyway Radeon HD 5870s are way better than water cooled GeForce GTX 295s. It's theoretically (no GPGPU tests with 5870s were done yet) better than GTX 295 for hash cracking if you know how to program ATI hardware . I'd prefer to correct sentence in such way :). If you want professional system for d...

@IvanG:obviously, you had written your stuff in .ptx and maybe compiled with ptxas, but where's 'module' aka .cubin in raw??? No point to use .ptx, compiler is good enough to use plain C. Cubins compressed and stored inside .exe. Actually if you can't find them by you own then you don't need them a...

Haven't tried nVidia's OpenCL by my own but have read that it's slower than CUDA. Not a surprise actually, compatibility costs some resources. As for ATI's OpenCL I have serious doubts that it'll available any time soon. Also chances quite high than only OpenCL compatible GPUs will be HD5800s. Short...

fotbollsplan wrote:ASUS P6T7 WS SuperComputer
7xPCI-Ex(2.0)x16

7 x GTX295?

4xGTX295 is a maximum. BIOS/drivers limits GPUs number to 8, also don't forget about cooling system which takes 2 slots (actually there is water cooling for GTX295 which 1 slot only but I never seen such systems even on inet photos).

Queuing hashes, advanced statistics/options, auto-save/resuming, etc -- everything planned for GUI version. I guess it'll commercial though as I really bored to fight with ATI SDK/crappy written software and PR their GPUs for free :lol:. About password generation -- it done that way to get optimal s...