Nvidia GeForce MX130 Mobile

[Esthar]

Code: Select all

λ FAHClient.exe --lspci
VendorID:DeviceID:PCI Bus:PCI Slot:PCI function:Vendor Name:Description
0x8086:0x9d1a:0:29:2:Intel Corporation:Mobile 6th/7th Generation Intel(R) Processor Family I/O PCI Express Root Port #11 - 9D1A
0x8086:0x9d23:0:31:4:Intel Corporation:Synaptics SMBus Driver
0x10de:0x174d:1:0:0:NVIDIA Corporation:NVIDIA GeForce MX130
0x8086:0x5917:0:2:0:Intel Corporation:Intel(R) UHD Graphics 620

Code: Select all

*********************** Log Started 2020-03-18T04:05:42Z ***********************
04:05:42:************************* Folding@home Client *************************
04:05:42:        Website: https://foldingathome.org/
04:05:42:      Copyright: (c) 2009-2018 foldingathome.org
04:05:42:         Author: Joseph Coffland <joseph@cauldrondevelopment.com>
04:05:42:           Args: --open-web-control
04:05:42:         Config: C:\Users\Esthar\AppData\Roaming\FAHClient\config.xml
04:05:42:******************************** Build ********************************
04:05:42:        Version: 7.5.1
04:05:42:           Date: May 11 2018
04:05:42:           Time: 13:06:32
04:05:42:     Repository: Git
04:05:42:       Revision: 4705bf53c635f88b8fe85af7675557e15d491ff0
04:05:42:         Branch: master
04:05:42:       Compiler: Visual C++ 2008
04:05:42:        Options: /TP /nologo /EHa /wd4297 /wd4103 /Ox /MT
04:05:42:       Platform: win32 10
04:05:42:           Bits: 32
04:05:42:           Mode: Release
04:05:42:******************************* System ********************************
04:05:42:            CPU: Intel(R) Core(TM) i5-8250U CPU @ 1.60GHz
04:05:42:         CPU ID: GenuineIntel Family 6 Model 142 Stepping 10
04:05:42:           CPUs: 8
04:05:42:         Memory: 19.88GiB
04:05:42:    Free Memory: 6.22GiB
04:05:42:        Threads: WINDOWS_THREADS
04:05:42:     OS Version: 6.2
04:05:42:    Has Battery: true
04:05:42:     On Battery: false
04:05:42:     UTC Offset: 7
04:05:42:            PID: 7536
04:05:42:            CWD: C:\Users\Esthar\AppData\Roaming\FAHClient
04:05:42:             OS: Windows 10 Enterprise
04:05:42:        OS Arch: AMD64
04:05:42:           GPUs: 0
04:05:42:  CUDA Device 0: Platform:0 Device:0 Bus:1 Slot:0 Compute:5.0 Driver:10.2
04:05:42:OpenCL Device 0: Platform:0 Device:0 Bus:1 Slot:0 Compute:1.2 Driver:442.59
04:05:42:OpenCL Device 1: Platform:1 Device:0 Bus:NA Slot:NA Compute:2.1 Driver:22.20
04:05:42:  Win32 Service: false
04:05:42:***********************************************************************

thanks.

[bruce]

The GeForce MX130 cannot be supported by FAH.

1) All laptop computers have difficulty dissipating the heat generated when performing heavy calculations continuously.
2) The "mobile" versions of GPUs are traditionally more limited in their ability to complete WUs within the deadlines imposed by FAH.
3) The Double Precision Floating Point set of instructions are required. I can't determine if the GeForce MX130 supports DP instructions, but even if it does, you would not be happy with the work you might be assigned based on 1 and 2, above.

[JimboPalmer]

Welcome to Folding@Home!

https://www.techpowerup.com/gpu-specs/g ... x130.c3043

This is a slow version of maxwell with 384 shaders. (also called CUDA cores) it does support both OpenCL 1.2 and Double Precision floating point math. These are the major hardware needs to be supported.

Under ideal conditions it might be as fast as Nvidia's slowest current desktop card the GT 1030

I see two complications: bruce mentioned heat, a laptop never has as good ventilation as a desktop, and has tiny fans that spin loudly. You have not mentioned what computer this is in, but frequently laptops use an Intel GPU for light loads and swap to the Nvidia for real graphics when needed (It would be hotter, louder, and more power hungry than the Intel) The buzzword is Optimus https://www.geforce.com/hardware/techno ... technology

The problem is that F@H does not support Intel GPUs, and will crash before the swap occurs. I hear rumors you can force it to always use the MX130 in the Nvidia control panel 3d settings. Even if that works, you now have a hot, noisy, power hungry laptop and it is possible those are not features you wanted in a laptop.

It is your call, but you can see why not many people will have experience Folding on your hardware. Unless you use it as a desktop away from humans, the heat and noise will make it a poor neighbor.

[Zodan]

Hi there all,

I wanted to add my MX130 to my personal GPUs.txt file, what would I need to add from the following information :

fahclient --lspci output :
0x10de:0x174d:2:0:0:NVIDIA Corporation:NVIDIA GeForce MX130

GPU-Z information :
Name : NVIDIA GeForce MX130
GPU : GM108
Revision : A2
Subvendor : Dell
Device ID : 10DE 174D - 1028 089E

Would you need more information from me ?

I understand that FAH might crash before a swap of GPU usage ( my laptop run the Intel integrated graphics for normal desktop applications), but I am able to force ( through the NVIDIA Control Panel ) that certain executables run with the MX130 directly.
I am curios which of the EXEs I would have to force though ( FAHClient.exe, FAHControl.exe or FahCore_a7.exe ) so that the detection would happen, please may you help me.

I'm more interested in seeing that FAH can detect my GPU than using it outright for folding ( at the moment my i7 CPU is doing all the work on that )

I'd appreciate any feedback you may give me.

Greets, Hendrik

[Joe_H]

First the GPU would need to be added to the list on the server, changing your local GPUs.txt file only allows you to create a GPU slot that will never get work until the entry is added on the server level. If you do create such an entry, please do not add to the request connection saturation on the servers by enabling to run until the entry is in the downloaded list from the server.

You have provided enough information for the GPU to be added to the server list.

As long as Dell did not get a custom chip that disables features standard on the regular MX130, it is listed as supporting OpenCL and DP calculations.

The folding cores are FAHCore_21 and FAHCore_22. Not sure what else you might have to specify, and it might not be enough to do it this way. You might have to leave the iGPU disabled all of the time the client is active.

[toTOW]

Added the following GPU to the list of supported GPUs :
0x174d / GM108M [GeForce MX130]
0x1789 / GM107GL [GRID M3-3020]