System Specs:
Intel Core i7-6700K
Asus Z170-E motherboard
32 GB of RAM (I forget brand and speed)
EVGA GeForce GTX Titan 6 GB (original, not X, Xp, or any of the other, newer variants of the Titan line of cards)
Windows 7 SP1
PhoenixMiner 5.5c for Windows
Nvidia GeForce driver version 457.51
So from the research that I’ve done, I am able to run the benchmark which tests it at epoch 2 I think (or something like that) and it’s able to get about 18 MH/s on said GTX Titan.
However, when I tried to connect to an actual pool, or where the epoch, I think, is currently sitting at epoch 399, right now, it crashes. (Or more specifically, the system reports that the video driver kernel crashes.)
Nvidia-SMI reports that the system is currently using 78 MiB out of 6144 MiB of FB memory and BAR1 memory is using 5 MiB out of 125 MiB.
Using the benchmarking flag in PhoenixMiner, I was able to find out that it can run upto around epoch 195, but then at around epoch 200, that’s when it will start crashing the video driver kernel.
With 6144 MiB of FB memory, there should be plenty of room to be able to fit the ~2 GiB DAG into said VRAM.
So, what I am wondering is if other people have encountered something like this before (I searched the forums, and this problems seems to be more prevalent for AMD GPUs as opposed to Nvidia GPUs). (I’m trying to make use of what I have at the moment before committing to buying new hardware.)
From Wikipedia, it says that the core configuration of the GK110-400-A1 of the GTX Titan should be 2688:224:48 (unified shaders:texture mapping units:render output units).
(Sidebar: Note, my primary video output display IS off of the integrated GPU from the 6700K, which is HD Graphics 530, so I have no monitors, etc. attached to the GTX Titan right now.)
My thought was that the DAG should have been entirely driven by VRAM.
But if that’s the case, then why would the GTX Titan be failing at around epoch 200 where the size of the DAG is only supposed to be around 2.58 GB?
Therefore; is the failure or the video driver kernel crashing also tied to the core config of the GPU processor itself?
If anybody can shed a little bit of light in regards to this, that would be greatly appreciated.
Thank you.
Also sidenote: I’m not doing any overclocking or anything like that.
The command that I am using, has no GPU control flags (e.g. -tt, -cclock, -mclock, etc.).
For benchmarking (where I can still tell that it fails starting from around epoch 200), the command that I use is literally:
PhoenixMiner.exe -bench 200
Thanks.