Re: TT-Miner 2.2.1: ProgPoW, MTP, LYRA2V3, UBQHASH, MYR-GR Win/cuda9.2/10.0/10.1

@TrailingStop Hi,

One of my friend has a problem with TT-Miner 2.2.2 beta2, he doesn't try older versions but probably problem is also on every version.

He is getting error when trying to mine MTP with 8 Nvidia cards, default -intensity 15.
We didn't find limit for number of GPUs in README guide.

OS: Win 10
Driver: 419.67
RAM: 4 GB
Virtual Mem.: 99 GB
GPUs:
6 x 1070 8GB
2 x 1070ti 8GB

Running with -a MTP-101

Quote

CudaError: 2
Not enough memory to run algo. GPU disabled.(0/1)
Plese check if your Virtual Memory settings....

https://cdn.discordapp.com/attachments/379391607406985218/566651306886955008/unknown.png

And these are settings for Virtual Memory (Initial size: 99000 MB, Maximum size 99000 MB, Currently allcated: 99000 MB
https://cdn.discordapp.com/attachments/379391607406985218/566652780631293992/unknown.png

Thx. for support.

Hi,

there is no static limitation in the miner - from that point of view it should work with 8 GPUs. Can you please try with a lower intensity? In the beta you check the 'auto optimize' function is active if you do not add -i or -gs to the command line. It might be that I use too high values for optimization in this beta. So please try with -i 14 - just to see if that works better.

Please let me know - I will take a deeper look into it.

Thanks.

Hi, I'm a friend of Hogh with 8 gpus on one rig (6x 1070 & 2x1070TI)

It does not work with -i 14 also Miner crashes with "no memory" warning

That is because you have mixed Video Cards ...

1070 is only capable of a lesser intensity than the 1070TI, whichmeans you may have to assign them individually. We used to have this same issues in our miners until we made ALL cards exactly the same. Much smoother that way, but of course not always possible.

#crysx