hashcat Forum

Cooling:
- https://www.youtube.com/watch?v=HiWRt2aL9tg

System Specs:
- Case : Lian Li PC-D8000
- Power : 2x LEPA G1600-MA-EU
- GPUS : 4x XFX HD 7990 ( FX-799A-XNF9 )
- MOBO : ASRock X79 Extreme11
- CPU : i7-3930K
- RAM : Corsair 16 GB DDR3-1600

Note: GPU/CPU are all stock clocks!

Dude, you totally rock!

Any chance of a quick recording showing the inside spec of you machine?
...drools

(09-03-2013, 06:29 PM)phillips321 Wrote: [ -> ]Any chance of a quick recording showing the inside spec of you machine?
...drools

I'm building an exact same second machine (VCL them together), once all setup will post a full vid / picture repo featuring all aspects.

For those interested:



Dual System setup: https://85.17.59.133/hashcat/IMG_7008.JPG

Close-up: https://85.17.59.133/hashcat/IMG_7019.JPG

gpufreak would this setup also function with 2* 1250-X series from seasonic?

I think that 1250-X has 3 PCI-E dedicated power outputs and one mixed CPU/PCI-E. So if you use that mixed port for PCI-E you have 4 connectors per power supply to feed the cards. The entire system usage is around 1700-1750 under full load so 2x 1250 should be enough if you distribute the load.

Be sure to pick a motherboard which you can feed additional 12V for the cards else you will draw too much power on your 24 pin connector and it will melt ;-)

That is why I personally choose the ASRock which has 2 extra 12V inputs for the PCI-E's.

It seems to have 8* PCI-E 6/8 pins on each PSU so 16 in total. On the reference card there are 2 PCI-E 6pins each so thats enough connectors.



I own a Asus maximus Extreme-z with 1*12v on mobo as you can see in the picture. Will this be enough to give the mobo that extra power?

The idea was to put 2 7990's in crossfire on 1 psu and the other 2 cards on the second one.

The numbers in your benchmark are actually a bit low. At 925 Mhz, single hash NTLM performance should be about 15700 MH/s per GPU, or 125.6 GH/s total. But actually stock clock for your 7990 should be 1000 Mhz. At 1000 Mhz, you should be getting about 17000 MH/s per GPU. Our 4x 7990 boxes pull about 136 GH/s total.

this is more like what you should be seeing for single ntlm brute force: