Usage Monitor Plug-in and Roundlines/Dedicated GPU Process Memory Usage/Having Trouble With the Values.

Post by **Yincognito** » July 11th, 2024, 9:29 am

Thinkr8 wrote: ↑July 11th, 2024, 8:05 am Well, I slept on that question and I'm not sure what return in mGPU0MU I want, lol. So I have to look at the meter name the MU in mGPU0MU stands for Memory Usage so I was thinking "Dedicated Usage" is the choice. I'll be working on it on and off today and I'm not sure how much I'll get done, TY

What is the most popular choice the end user would like to have? Do you have any idea or is that a far reaching question?

Also after a night of sleep I realized I don't know very much about the inner working's of a GPU and thought tells me it would help to know.

One last question; after all these some odd years of Rainmeter why hasn't a GPU plug-in been created so we don't have to use third party software to gather the info we want? I'm not intelligent enough to build one, if I was, I'd do it. Lol.

Yep, knowing what you want is indeed a question that should have a precise answer.

I think that based on your OP screenshot, setting colors should be based on what the 3rd roundline and text from the 1st "row" are about, i.e. GPU0MemoryUsage taken from MSI AB. In other words, you could very well move the IfConditions in GPU0MemUsage and make them about GPU0MemoryUsage instead of mGPU0MU (no scaling needed either, if you do that), then get rid of the remaining mGPU0MU altogether. After all, all your roundlines are about general GPU attributes like usage, temp, memory usage, memory clock, fps and core clock (with only the "Rainmeter ... Top VRAM Process" line dedicated to individual process stats regarding GPU), so it would make sense that setting colors of that 3rd roundline to be based on the same value as the roundline is about (which, as I said, is not the case in the current code).

Then, if you want to have other individual process stats in your skin along the "Rainmeter ... Top VRAM Process" line, just use the typical aliases for the top processes, just like in that line. There is no point in complicating things too much, as the user will almost always be interested in the "total" stats your roundlines / texts already cover, and in the top N "individual" stats for processes.

As for your last question, there are actually some GPU plugins posted here on the forum, but then this is far from being an easy task and having a specialized 3rd party software providing such data is a much more feasible choice, since the developers of such software have the knowledge, resources and hardware required to provide such info on a much wider range of devices / systems / OSs that any Rainmeter user or developer has. Rainmeter by itself is not a hardware info tool like those are, so no point in reinventing the wheel to provide what other tools already do much better.

Thinkr8 · Post by **Thinkr8** » July 11th, 2024, 10:30 am

... In other words, you could very well move the IfConditions in GPU0MemUsage and make them about GPU0MemoryUsage instead of mGPU0MU (no scaling needed either, if you do that), then get rid of the remaining mGPU0MU altogether.

What? lol.

Here's what I did and now have better results (I think). What I have toggled off gave the same measurement as "Dedicated Usage." I thought that was weird, so I chose what isn't toggled off. Total dedicated Usage of all process's, that's the most common that should be used.

Code: Select all

[GPU0MemoryUsage]
Measure=Plugin
Plugin=MSIAfterburner.dll
GPU=0
SourceId=49
MinValue=0
MaxValue=#MaxV#
UpdateDivider=2
OnChangeAction=[!UpdateMeter MemUsed GPU0MemUsage]

[GPU0Roundline3]
Meter=ROUNDLINE
MeasureName=MeasureRing
LineColor=255,216,0,100
LineWidth=6.0
LineLength=55
LineStart=50
X=395r
Y=0r
AntiAlias=1
Solid=1
Group=Details

[GPU0MemUsage]
Meter=ROUNDLINE
MeasureName=GPU0MemoryUsage
; LineColor=184,32,15
LineWidth=6.0
LineLength=55
LineStart=50
StartAngle=(Rad (270))
RotationAngle=(Rad (360))
X=0r
Y=0r
AntiAlias=1
Solid=1
Group=Details

[mGPU0MU]
Measure=Plugin
Plugin=UsageMonitor
Category=GPU Process Memory
Counter=Dedicated Usage
; Category=GPU Process Memory
; Counter=Local Usage
; Category=GPU Adapter Memory
; Counter=Dedicated Usage
; Category=GPU Local Adapter Memory
; Counter=Local Usage
Index=0
IfCondition=(mGPU0MU < 250)
IfTrueAction=[!SetOption GPU0MemUsage LineColor "75,216,19"] 
IfCondition2=(mGPU0MU/1048576 >= 250) && (mGPU0MU/1048576 < 500)
IfTrueAction2=[!SetOption GPU0MemUsage LineColor "247,138,5"]
IfCondition3=(mGPU0MU/1048576 >= 500) && (mGPU0MU/1048576 < 1024)
IfTrueAction3=[!SetOption GPU0MemUsage LineColor "247,180,5"]
IfCondition4=(mGPU0MU/1048576 >= 1024) && (mGPU0MU/1048576 < #MaxV#)
IfTrueAction4=[!SetOption GPU0MemUsage LineColor "247,17,5"]

Image 005.jpg

If I have the Memory Usage roundline correct now and all the memory usage measurements match then I would like to move on to Memory Clock. and Core Clock (the IfConditions there don't produce the desired results either.) Is that Ok with you? But something tells me the Memory Usage meter isn't done.

EDITed: I attempted to fix Mem Clock. Needless to say, I got nothing. I don't know what math conversion steps I should take with the IfConditions if that is what's supposed to be done. It takes me a lot of slow thought to get the jest of it and I still doubt myself even when I get it right.

Post by **Yincognito** » July 11th, 2024, 1:30 pm

Thinkr8 wrote: ↑July 11th, 2024, 10:30 am What? lol.
[...]
But something tells me the Memory Usage meter isn't done.

As you wish, but you're still displaying the memory usage text and roundline based on a different value than the value you use to color the roundline - just look at the value in the log for the [GPU0MemoryUsage] measure (which you're displaying in the [GPU0MemUsage] and [mMemoryUsage] meters), compared to the one for the [mGPU0MU] measure (which controls the color of the [GPU0MemUsage] meter), this is how it looks for me after replicating the relevant parts of your code:

GPU Memory Usage.jpg

Notice how the 183 MB value for GPU0MemoryUsage is an entirely different value than the roughly 378 MB value for mGPU0MU. I don't know how it looks on your system, maybe you have a single GPU and these values are the same, but unless that's the case, it will display something and color it based on something else. This is the 3rd time I'm trying to convey this thing until now, lol.

In case it helps, this is how I'd approach things (now the roundline is colored according to the value both it and the text illustrate):

By the way, you forgot to scale the 1st IfCondition, so, if you insist on keeping the same approach (I mean, it's your skin, neither me or someone else has to be ok with it if you want to do it in a certain way), that should be:

Code: Select all

IfCondition=(mGPU0MU/1048576 < 250)

Thinkr8 wrote: ↑July 11th, 2024, 10:30 am EDITed: I attempted to fix Mem Clock. Needless to say, I got nothing. I don't know what math conversion steps I should take with the IfConditions if that is what's supposed to be done. It takes me a lot of slow thought to get the jest of it and I still doubt myself even when I get it right.

It's the same there too. First, you don't scale mGPU0MC by 1048576 like in mGPU0MU, secondly you display a value in the text and the roundline (GPU0MemoryClock), but you color the roundline based on another value (mGPU0MC, which isn't even related to the memory clock in the first place, in this case).

Thinkr8 · Post by **Thinkr8** » July 11th, 2024, 2:13 pm

Notice how the 183 MB value for GPU0MemoryUsage is an entirely different value than the roughly 378 MB value for mGPU0MU.

I took a look in About Rainmeter and you were right, the values were different. I thought they were and would have re-approached it, again. I won't blame you if you give up on me, my teacher's did, lol. I'll be disappointed, but yeah, what could I do.

Anyway, I used your approach and see it puts out the same readings as the other approach, (but now match) so that's the problem, too many different way's of getting the same result's. Sticking with one approach should eliminate any further confusion.

Since your approach works now can I ask you what do I do with Memory Clock and Core Clock because they don't produce the right results either and those equations are in megahertz which I have no idea how to approach.

EDITED: Today 11:23PM

I fix Memory Clock with the same approach, thank you.

Post by **Yincognito** » July 11th, 2024, 5:19 pm

Thinkr8 wrote: ↑July 11th, 2024, 2:13 pm I took a look in About Rainmeter and you were right, the values were different. I thought they were and would have re-approached it, again. I won't blame you if you give up on me, my teacher's did, lol. I'll be disappointed, but yeah, what could I do.

Anyway, I used your approach and see it puts out the same readings as the other approach, (but now match) so that's the problem, too many different way's of getting the same result's. Sticking with one approach should eliminate any further confusion.

Since your approach works now can I ask you what do I do with Memory Clock and Core Clock because they don't produce the right results either and those equations are in megahertz which I have no idea how to approach.

EDITED: Today 11:23PM

I fix Memory Clock with the same approach, thank you.

Yeah, I very rarely give up on something I start and teachers can make mistakes too, so you can relax, lol. I was just watching some Euro 2024 analysis on YT, hence the break in responding. Yes, the rest of the cases where you have both a MSI AB measure and a UsageMonitor one are solved the same way, just stick with one of them and get rid of the other. Typically, you'd use the UsageMonitor ones when handling the individual process stats, just like you did for the VRAM line in your OP screenshot.

Thinkr8 · Post by **Thinkr8** » July 11th, 2024, 6:34 pm

Typically, you'd use the UsageMonitor ones when handling the individual process stats, just like you did for the VRAM line in your OP screenshot

That's exactly what I did.

I fixed all the roundlines and text because of your help. Thank you.

What I'm left with is what I chose to put behind the front view and as of now it's in-complete and a few of what I already have there isn't producing any results. Can I ask you tomorrow as to why they aren't? My day is ending soon.

Most Excellent!

Oh, since most of the skin is using MSI AB and the SourceID number will every end user have to enter a different SID number? If so, I'll have to take action to prevent that such as providing an input field for each SID.

Post by **Yincognito** » July 11th, 2024, 7:37 pm

Thinkr8 wrote: ↑July 11th, 2024, 6:34 pm That's exactly what I did.

I fixed all the roundlines and text because of your help. Thank you.

What I'm left with is what I chose to put behind the front view and as of now it's in-complete and a few of what I already have there isn't producing any results. Can I ask you tomorrow as to why they aren't? My day is ending soon.

Most Excellent!

Oh, since most of the skin is using MSI AB and the SourceID number will every end user have to enter a different SID number? If so, I'll have to take action to prevent that such as providing an input field for each SID.

You're welcome, I'm glad it helped. Sure, you can ask about it any time you like (just not about the same things, I try to avoid repeating myself when it comes to the same person, lol). No, the SourceID (and the rest of the MSI AB measures' options) will stay the same, irrespective of the system the skin is run on, as long as we're talking about the same category of information (e.g. Memory Usage will always be SourceId=49 regardless of system):

Yincognito wrote: ↑July 10th, 2024, 6:53 pmUnlike the old HWiNFO way of getting data using the Shared Memory Viewer (which indeed required user / system specific configuration), MSI AB has a consistent syntax for the MSI AB plugin measure options, irrespective of system (I'm talking about the syntax of options like SourceId, DataSource, GPUDescriptor, etc).

That being said, the SourceID (and the rest of the MSI AB measures' options) will be different if you want to query another category of information (e.g. the SourceId of the GPU Core Clock, aka 32, will always be different than the SourceId of the Memory Usage, aka 49).

Thinkr8 · Post by **Thinkr8** » July 12th, 2024, 9:16 am

Thank you for the last reply, I agree.

Greeting's, I've checked and doubled checked the measures and meters but I'm unable to find as to why these aren't working, no results are being displayed. In MSI AB I have them ticked on, with the exception of the "ShaderCLock", I'm unable to find it in MSI AB. Is ShaderClock under a different name? All of the SourceID's are correct. In MSI AB the Temp Limit and the Power Limit sliders do slide but they don't hold (even though I don't need any limits, they seem as if they won't work properly, would that be because of my GPU doesn't allow for it?) If I'm unable to configure these measures so they produce results, what would I do, just go with what I have and hope they work on another's GPU? That wouldn't be a good thing to do to the end user, unless I could be absolutely sure they will work on another GPU I think I shouldn't release them as they are. I just need a little push in the right direction as to what I should do if I can't get any results on my end. Any advice? Thank you.

Code: Select all

; { ---GPU Shader Clock--- }

[GPU0ShaderClock]
Measure=Plugin
Plugin=MSIAfterburner.dll
GPU=0
SourceId=(no ID number, I'm unable to find ShaderClock in MSI AB)
MinValue=0
MaxValue=8000

[mShaderClock]
Meter=String
MeterStyle=Ink5
MeasureName=GPU0ShaderClock
SolidColor=0,53,91
Text=Shader Clock: %1 MHz
X=60
Y=40r
NumOfDecimals=1
DynamicVariables=1

; { ---GPU Temperature Limit--- }

[GPU0TempLimit]
Measure=Plugin
Plugin=MSIAfterburner.dll
GPU=0
SourceId=112
MinValue=0
MaxValue=200
UpdateDivider=2

[mGPU0TempLimit]
Meter=String
MeterStyle=Ink2
MeasureName=GPU0TempLimit
SolidColor=0,53,91
Text=Temp Limit: %1°
X=580r
Y=0r
NumOfDecimals=2
DynamicVariables=1

; { ---GPU Power Limit--- }

[GPU0PowerLimit]
Measure=Plugin
Plugin=MSIAfterburner.dll
GPU=0
SourceId=113
MinValue=0
MaxValue=8000
UpdateDivider=2

[mGPU0PowerLimit]
Meter=String
MeterStyle=Ink5
MeasureName=GPU0PowerLimit
SolidColor=0,53,91
Text=Power Limit: %1w
X=295r
Y=0r
NumOfDecimals=1
DynamicVariables=1

; { ---GPU Vendor--- }

[GPU0Vendor]
Measure=Plugin
Plugin=MSIAfterburner.dll
GPUDescriptor=VendorId
GPU=0
SourceID= (NULL)
UpdateDivider=3600

[mGPU0VenId]
Meter=STRING
MeterStyle=Ink5
MeasureName=GPU0Vendor
SolidColor=0,53,91
Text=VENDOR: %1
X=220r
Y=0r

And the VendorID is only producing "10DE" that's not the full ID. How do we get the full ID? I've tried a few ways to try and get it to work -but- no results.

Post by **Yincognito** » July 12th, 2024, 12:21 pm

Thinkr8 wrote: ↑July 12th, 2024, 9:16 am Thank you for the last reply, I agree.

Greeting's, I've checked and doubled checked the measures and meters but I'm unable to find as to why these aren't working, no results are being displayed. In MSI AB I have them ticked on, with the exception of the "ShaderCLock", I'm unable to find it in MSI AB. Is ShaderClock under a different name? All of the SourceID's are correct. In MSI AB the Temp Limit and the Power Limit sliders do slide but they don't hold (even though I don't need any limits, they seem as if they won't work properly, would that be because of my GPU doesn't allow for it?) If I'm unable to configure these measures so they produce results, what would I do, just go with what I have and hope they work on another's GPU? That wouldn't be a good thing to do to the end user, unless I could be absolutely sure they will work on another GPU I think I shouldn't release them as they are. I just need a little push in the right direction as to what I should do if I can't get any results on my end. Any advice? Thank you.

And the VendorID is only producing "10DE" that's not the full ID. How do we get the full ID? I've tried a few ways to try and get it to work -but- no results.

Unfortunately, I can't help much here, since for me the Temp Limit and Power Limit don't display anything in MSI AB (for my laptop, the fan speeds are also unavailable in all the tools I tried, not just MSI AB), let alone being able to slide them, and the VendorId is producing the same result as for you:

GPU Info.jpg

I didn't use them until now either, but what I can point you to is (not sure if stangowner, the MSI AB plugin developer, is still active though):
https://forums.guru3d.com/threads/rainmeter-plugin-for-msi-afterburner.319558/
There, you will find out that probably all SourceIds are available in the C:\Program Files (x86)\MSI Afterburner\SDK\Include\MAHMSharedMemory.h file, once you convert the hexadecimal numbers there into decimal (that yields 33 as the SourceId of the ShaderClock, by the way, in case it helps).

I vaguely remember working with such IDs and the registry in the past to find out what I was interested in at the time (if I'm not mistaken, it's in the thread I linked to above), but I'm not sure it was the vendor ID. As for the PNP device ID (which includes a part of the vendor ID), you can get it using PowerShell (in the case of a skin, available via a RunCommand measure), if you properly filter out or Substitute the output of Get-CimInstance -classname Win32_VideoController:

GPU Info PS.jpg

Typically, I would suggest including only the info you know for sure most users have it available (this depends on whether those systems are supported by MSI AB and the user settings regarding monitoring in MSI AB, but it's generally the most used stuff like core clock, core usage, memory clock, memory usage, temperature, etc.), and optionally have some readme.txt or meters in the skin or the Settinngs one to explain the basic steps to make the rest available in the skin if really needed.

Thinkr8 · Post by **Thinkr8** » July 12th, 2024, 5:02 pm

Ok great, thank you, I'll need a day or two. TY

Usage Monitor Plug-in and Roundlines/Dedicated GPU Process Memory Usage/Having Trouble With the Values.

Re: Usage Monitor Plug-in and Roundlines/Dedicated GPU Process Memory Usage/Having Trouble With the Values.

Re: Usage Monitor Plug-in and Roundlines/Dedicated GPU Process Memory Usage/Having Trouble With the Values.

Re: Usage Monitor Plug-in and Roundlines/Dedicated GPU Process Memory Usage/Having Trouble With the Values.

Re: Usage Monitor Plug-in and Roundlines/Dedicated GPU Process Memory Usage/Having Trouble With the Values.

Re: Usage Monitor Plug-in and Roundlines/Dedicated GPU Process Memory Usage/Having Trouble With the Values.

Re: Usage Monitor Plug-in and Roundlines/Dedicated GPU Process Memory Usage/Having Trouble With the Values.

Re: Usage Monitor Plug-in and Roundlines/Dedicated GPU Process Memory Usage/Having Trouble With the Values.

Re: Usage Monitor Plug-in and Roundlines/Dedicated GPU Process Memory Usage/Having Trouble With the Values.

Re: Usage Monitor Plug-in and Roundlines/Dedicated GPU Process Memory Usage/Having Trouble With the Values.

Re: Usage Monitor Plug-in and Roundlines/Dedicated GPU Process Memory Usage/Having Trouble With the Values.