I have no idea of how to get this working
 

So ive tried to follow the available guides and tried searching through various forum posts for hours but to no avail. I have a 5950x that im happily running prime95 on but i want to set up my 3090 to do trial-factoring. I believe i have setup MISFIT/GPU72 to get and turn in my work but i cant get either mfaktc or mfaktx controller to get going. Pls help.

Welcome to the forum.
Here is a pointer to a reference guide. [url]https://www.mersenneforum.org/showthread.php?p=488518[/url]
Someone with more experience can help answer questions that you have.

[QUOTE=bluecat0531;596676]I believe i have setup MISFIT/GPU72 to get and turn in my work but i cant get either mfaktc or mfaktx controller to get going. Pls help.[/QUOTE]Log everything. Read the logs for clues. Divide and conquer; first get mfaktc working alone, without Misfit/gpu72. If you still can't figure it out, show us some logs. As things stand now, we don't know what issues you're having, because you've provided almost zero information. (Did mfaktc run at all? Did it pass the selftest? Is there anything in the worktodo file? Does the worktodo file even exist? Where did you install it? What error messages or console output do you get when you try to run mfaktc? Misfit? Etc, etc, etc.)

[QUOTE=kriesel;596740]Log everything. Read the logs for clues. Divide and conquer; first get mfaktc working alone, without Misfit/gpu72. If you still can't figure it out, show us some logs. As things stand now, we don't know what issues you're having, because you've provided almost zero information. (Did mfaktc run at all? Did it pass the selftest? Is there anything in the worktodo file? Does the worktodo file even exist? Where did you install it? What error messages or console output do you get when you try to run mfaktc? Misfit? Etc, etc, etc.)[/QUOTE]

So i cant figure out how to run mfaktc at all. Ive tried using a cmdline but i dont know any commands at all and trying a batch file was no help either, my only other option was just double clicking the .exe file and all that happens is a cmdline appears for one second then disappears.

What operating system are you using?

Where have you installed mfaktc?

Open a terminal window, cd into wherever you have installed mfaktc and run if from there. (Under Linux or other UNIX variant type [c]./mfaktc[/c] and press Enter.) Then post *all* the output you get from it here.

Do you have a worktodo.txt established? And IIRC there is a need for a config file to be established before it will stay running. That part was confusing to me too. (I tried running it on a machine with an integrated GPU. But it used shared memory and lowered my desired Prime95 performance.)

Until the OP posts some useful detail of what he's tried and what it produced, makes some effort to provide data about what's occurring, I suggest we all stop guessing in the dark and trying to help. Our effort should not exceed his.

My answer was commensurate with the OP's most recent post.

[QUOTE=Uncwilly;596760]the OP's most recent post.[/QUOTE]I don't have an issue with your posts or Chris's. Look again at post 4. He quoted my entire post, that included several questions, answered none, and provided little information or specifics. We don't know if he has installed mfaktc in Program Files or system32 or somewhere sensible, has a worktodo file, has proper permissions on the install folder and mfaktc files, has a suitable CUDA computing compatible driver installed, what batch file he has attempted, what cmd line he has attempted, etc etc etc. The details which we would need to see to provide effective help are all being kept secret. I think we may have another "help me in spite of my refusing to provide any useful diagnostic data or answer any questions" troll.

[QUOTE=bluecat0531;596745]So i cant figure out how to run mfaktc at all. Ive tried using a cmdline but i dont know any commands at all and trying a batch file was no help either, my only other option was just double clicking the .exe file and all that happens is a cmdline appears for one second then disappears.[/QUOTE]

You need to make a text file called worktodo.txt, where you will have the work for the mfaktc program. It is in the following format:

[CODE]
Factor=[exponent],[bitlevel-start],[bitlevel-end]

Example:
Factor=500000071,82,83
[/CODE]
This example will search through all potential factors between 2^82 and 2^83. You can run it as a test, and it should take about 11 hours - in case you finish it, it will output a line to a [U][I]text file "results.txt", which you should beforehand create as an empty file in case it doesn't do it automatically[/I][/U]. There are of course other exponents to do TF (trial factoring) on, and those needed the most (above 120M exponents) take at most 10 minutes with your GPU, as they currently need TF from 2^74 to 2^75.

This should be all to make it work as to having the worktodo file.

It's however almost certain you don't have the necessary version of mfaktc, as I too didn't find a version supporting 3090 (I have 3080), but I didn't ask about it yet. You will also need a cudart64_(version number).dll - the number is most probably 104 or 105, as it mentions it needs mfaktc compiled for CUDA 10.4 or 10.5. You might also need to install the CUDA files necessary for running programs on GPU.


That's about it. If you have any questions, do ask.

[QUOTE=Viliam Furik;596771]You need to make a text file called worktodo.txt, where you will [STRIKE]have[/STRIKE][B]insert[/B] the work for the mfaktc program. It is in the following format:

[CODE]
Factor=[exponent],[bitlevel-start],[bitlevel-end]

Example:
Factor=500000071,82,83
[/CODE]This example will search through all potential factors between 2^82 and 2^83. You can run it as a test, and it should take about 11 hours - in case you finish it, it will output a line to a [U][I]text file "results.txt"[STRIKE], which you should beforehand create as an empty file in case it doesn't do it automatically[/STRIKE][/I][/U]. There are of course other exponents to do TF (trial factoring) on, and those needed the most (above 120M exponents) take at most 10 minutes with your GPU, as they currently need TF from 2^74 to 2^75.
...

It's however almost certain you don't have the necessary version of mfaktc, as I too didn't find a version supporting 3090 (I have 3080), but I didn't ask about it yet. You will also need a cudart64_(version number).dll - the number is most probably 104 or 105, as it mentions it needs mfaktc compiled for CUDA 10.4 or 10.5. You might also need to install the CUDA files necessary for running programs on GPU.
[/QUOTE]In all my years of running numerous CUDA versions of mfaktc on multiple GPUs, I don't recall ever needing to create results.txt. Checking in the [URL="http://www.mersenneforum.org/mfaktc/mfaktc-0.21/mfaktc-0.21.tar.gz"]source code[/URL] shows it uses an append option call of fopen to write results lines.
[URL]https://www.cplusplus.com/reference/cstdio/fopen/[/URL] states the file will be created if it does not exist. In module output.c:
[CODE]void print_result_line(mystuff_t *mystuff, int factorsfound)
/* printf the final result line (STDOUT and resultfile) */
{
char UID[110]; /* 50 (V5UserID) + 50 (ComputerID) + 8 + spare */
char string[200];

FILE *resultfile=NULL;


if(mystuff->V5UserID[0] && mystuff->ComputerID[0])
sprintf(UID, "UID: %s/%s, ", mystuff->V5UserID, mystuff->ComputerID);
else
UID[0]=0;

if(mystuff->mode == MODE_NORMAL)
{
resultfile = [B]fopen(mystuff->resultfile, "a");[/B]
if(mystuff->print_timestamp == 1)print_timestamp(resultfile);
}
if(factorsfound)
{
#ifndef MORE_CLASSES
if((mystuff->mode == MODE_NORMAL) && (mystuff->stats.class_counter < 96))
#else
if((mystuff->mode == MODE_NORMAL) && (mystuff->stats.class_counter < 960))
#endif
{
sprintf(string, "found %d factor%s for %s%u from 2^%2d to 2^%2d (partially tested) [mfaktc %s %s]", factorsfound, (factorsfound > 1) ? "s" : "", NAME_NUMBERS, mystuff->exponent, mystuff->bit_min, mystuff->bit_max_stage, MFAKTC_VERSION, mystuff->stats.kernelname);
}
else
{
sprintf(string, "found %d factor%s for %s%u from 2^%2d to 2^%2d [mfaktc %s %s]", factorsfound, (factorsfound > 1) ? "s" : "", NAME_NUMBERS, mystuff->exponent, mystuff->bit_min, mystuff->bit_max_stage, MFAKTC_VERSION, mystuff->stats.kernelname);
}
}
else
{
sprintf(string, "no factor for %s%u from 2^%d to 2^%d [mfaktc %s %s]", NAME_NUMBERS, mystuff->exponent, mystuff->bit_min, mystuff->bit_max_stage, MFAKTC_VERSION, mystuff->stats.kernelname);
}

if(mystuff->mode != MODE_SELFTEST_SHORT)
{
printf("%s\n", string);
}
if(mystuff->mode == MODE_NORMAL)
{
fprintf(resultfile, "%s%s\n", UID, string);
fclose(resultfile);
}
}
[/CODE]RTX30xx are compute capability 8.6. That will require a compile with & dlls for CUDA 11.x. [URL]https://en.wikipedia.org/wiki/CUDA[/URL] Download mirror at [URL]https://download.mersenne.ca/mfaktc/mfaktc-0.21[/URL] Go for the 2047M GpuSieveSize capable version to get full performance from these fast GPUs.
CUDA 10.x is required for RTX20xx & GTX16xx.
MacOS support has been dropped after CUDA 10.2.

[QUOTE=kriesel;596764]I don't have an issue with your posts or Chris's. Look again at post 4. He quoted my entire post, that included several questions, answered none, and provided little information or specifics. We don't know if he has installed mfaktc in Program Files or system32 or somewhere sensible, has a worktodo file, has proper permissions on the install folder and mfaktc files, has a suitable CUDA computing compatible driver installed, what batch file he has attempted, what cmd line he has attempted, etc etc etc. The details which we would need to see to provide effective help are all being kept secret. I think we may have another "help me in spite of my refusing to provide any useful diagnostic data or answer any questions" troll.[/QUOTE]

I do apologize for my brief answers. I was not fully thinking them out as each time I was in a rush to the get to work for my ~10 hr shifts. This has been so for the past few days since i first made the thread. I did try to answer some of your questions but I believe that some of them i could not answer as i previously stated could not get the program to run at all. And then perhaps some of your questions were categorized under "etc." which i may have brushed of as I could not ascertain what you wanted. To answer what questions I can see in this quote, I have the program installed on my desktop inside of a folder, I believe i have installed the latest cuda driver which i know is seperate from a geforce driver(but on that note i did see in another post i need some .dll file ill be sure to go and grab that), the batch file was just basically a copy of the cmds i was running something like "./mfaktc.exe -h and some other extensions like the -st or -st2, commandline im using is just the one that pops up when you search cmd in windows 10 (which gives you my OS which i have the latest version of), and lastly no i dont believe i have any diagnostic data to give you due to my current predicament just info on my system/setup.

[QUOTE=Uncwilly;596755]Do you have a worktodo.txt established? And IIRC there is a need for a config file to be established before it will stay running. That part was confusing to me too. (I tried running it on a machine with an integrated GPU. But it used shared memory and lowered my desired Prime95 performance.)[/QUOTE]

I do have a worktodo.txt file and it does have the necessary data as seen from an earlier reply but as for the config file.I would have no idea the format or info it would need. Did try getting some configs to work from mfaktx but was having no luck there either. But for now I am just working to get mfaktc running as suggested i can mess with other helper programs later.

Still MUCH too vague. Until you provide lots of accurate specifics, effective help is not possible.

GIMPS GPU applications are all command line applications. If you don't know how to use the command line in your OS, learn that first. The disappearing command window is a symptom of someone who doesn't know even the basics of using the command line in their OS. Learn about cmd /k in Windows, or opening a terminal window in Linux.

Document in a post in the thread the [B]exact[/B] cmd lines you use, working directory, OS & OS version, hardware, mfaktc version, driver version, cudart version, and the exact resulting content of mfaktc output redirected to a file. Also exact mfaktc worktodo content except mask AID. Copy/paste to avoid (more) typos.
For a rough idea of level of detail, see the [URL="https://www.mersenneforum.org/showpost.php?p=527911&postcount=4"]Colab example[/URL]. The level of detail required, is that which would be required if you were dictating keystrokes to someone else to reproduce your run attempt EXACTLY, plus include its output too. Every character of input. Exact reproduction of any [URL="https://mersenneforum.org/showpost.php?p=585491&postcount=3503"]info section[/URL] and any error messages at the very least.
[QUOTE=bluecat0531;596867][B]something like[/B] "./mfaktc.exe -h and some other extensions like the -st or -st2, commandline im using is just the one that pops up when you search cmd in windows 10 (which gives you my OS which i have the latest version of), and lastly no i dont believe i have any diagnostic data to give you due to my current predicament just info on my system/setup.[/QUOTE]Linux syntax there vs Win 10 reference. "Something like" is pretty useless. We need to see EXACTLY what you attempt and EXACTLY what it generates as output. One wrong character, assumption, or omission can kill a run or send the helpful in the wrong direction.
Four posts in by the person requesting help, we know little more than there's a 5950 and a 3090 involved and a user who apparently needs a basic OS cmd line tutorial who claims he's set it all up correctly, but only sees a cmd box flash on and off the screen and can't or won't provide any useful detail such as program output or input.
I'm about done trying to help the OP due to the persistent lack of useful information from him. I wonder if he's ever debugged anything or is a frequent help desk user or how little computer experience or knowledge he has.

So I have navigated to my installation folder and ran the following commands which didnt seem to give me any output or keep the window open even using /k "start mfaktc-win-64.exe /k -h > OutputTest.txt", "start mfaktc-win-64.exe /k -st > OutputTest2.txt" , and "start mfaktc-win-64.exe /k -st2 > OutputTest3.txt". The directory im using is "C:\Users\Bluecat\Desktop\PRIME STUFF".

OS is Windows 10 version 21H2 build 19044.1415 and experience? Windows Feature Experience Pack 120.2212.3920.0.

Hardware 5950x, rtx 3090 (Asus strix OC 24GB), Gigabyte x570 Aorus Master (Bios version F35E), 2 2x 16GB G. SKill Trident Z RGB for AMD F4-3200C16-16GTZRX, Corsair 1000W power supply that im not going to open my pc to find out the exact model for.

Mfaktc version is 0.21, Cuda driver is 11.5(also Geforce driver 497.29), ive downloaded cudart64_110.dll but tbh no idea where to put it other than in the above directory.

Each of those commands did create a file with the stated name and each were blank.

And lastly as far as worktodo.txt content "Factor=2021-12-30,21245513,73,74
Factor=2021-12-30,21245479,73,74"

[QUOTE=bluecat0531;596928]So I have navigated to my installation folder and ran the following commands which didnt seem to give me any output or keep the window open even using /k "start mfaktc-win-64.exe /k -h > OutputTest.txt", "start mfaktc-win-64.exe /k -st > OutputTest2.txt" , and "start mfaktc-win-64.exe /k -st2 > OutputTest3.txt". The directory im using is "C:\Users\Bluecat\Desktop\PRIME STUFF".

OS is Windows 10 version 21H2 build 19044.1415 and experience? Windows Feature Experience Pack 120.2212.3920.0.

Hardware 5950x, rtx 3090 (Asus strix OC 24GB), Gigabyte x570 Aorus Master (Bios version F35E), 2 2x 16GB G. SKill Trident Z RGB for AMD F4-3200C16-16GTZRX, Corsair 1000W power supply that im not going to open my pc to find out the exact model for.

Mfaktc version is 0.21, Cuda driver is 11.5(also Geforce driver 497.29), ive downloaded cudart64_110.dll but tbh no idea where to put it other than in the above directory.

Each of those commands did create a file with the stated name and each were blank.

And lastly as far as worktodo.txt content "Factor=2021-12-30,21245513,73,74
Factor=2021-12-30,21245479,73,74"[/QUOTE][i]Now[/i], you're talking! :tu:

[QUOTE=bluecat0531;596928]So I have navigated to my installation folder and ran the following commands which didnt seem to give me any output or keep the window open even using /k "start mfaktc-win-64.exe /k -h > OutputTest.txt", "start mfaktc-win-64.exe /k -st > OutputTest2.txt" , and "start mfaktc-win-64.exe /k -st2 > OutputTest3.txt". The directory im using is "C:\Users\Bluecat\Desktop\PRIME STUFF".

OS is Windows 10 version 21H2 build 19044.1415 and experience? Windows Feature Experience Pack 120.2212.3920.0.

Hardware 5950x, rtx 3090 (Asus strix OC 24GB), Gigabyte x570 Aorus Master (Bios version F35E), 2 2x 16GB G. SKill Trident Z RGB for AMD F4-3200C16-16GTZRX, Corsair 1000W power supply that im not going to open my pc to find out the exact model for.

Mfaktc version is 0.21, Cuda driver is 11.5(also Geforce driver 497.29), ive downloaded cudart64_110.dll but tbh no idea where to put it other than in the above directory.

Each of those commands did create a file with the stated name and each were blank.

And lastly as far as worktodo.txt content "Factor=2021-12-30,21245513,73,74
Factor=2021-12-30,21245479,73,74"[/QUOTE]
There are errors, and info gaps:
1) Worktodo.txt should be without dates. To my knowledge no legit [URL="https://www.mersenneforum.org/showpost.php?p=522098&postcount=22"]worktodo entry format[/URL] in any GIMPS application includes dates of assignment or any dates.
[CODE]Factor=21245513,73,74
Factor=21245479,73,74[/CODE]or[CODE]Factor=WhateverTheFirstAIDis,21245513,73,74
Factor=WhateverTheSecondAIDis,21245479,73,74[/CODE]2) "start mfaktc-win-64.exe /k -h > OutputTest.txt" is not a legal Win 10 command. There is [URL="https://docs.microsoft.com/en-us/windows-server/administration/windows-commands/start"]no /k option for the Windows start command[/URL] or for mfaktc*.exe. (I think the /k is just getting passed to mfaktc and ignored though.)

3) What's being redirected to OutputTest.txt is the output of the start command, which is null, so an empty zero-bytes file is created. Mfaktc's console output would go to a command window that blinks quickly in and out of existence. There are ways around that, using " " correctly, or otherwise correcting the command line used.

4) Mfaktc has not only multiple version numbers, but there are [URL="https://download.mersenne.ca/mfaktc/mfaktc-0.21"]a lot of different compiled flavors in multiple dimensions[/URL]:
[LIST][*]more-classes vs less-classes,[*]standard 128M max GpuSieveSize vs. 2047M max size,[*]many with different CUDA levels supported;[*]Windows or Linux;[*]moot in this case, but in older CUDA levels there were both 32-bit and 64-bit compiles.[/LIST]The CUDA level compatibility could be a looming problem after the other issues are resolved. Can't tell from the info provided, which of the many flavors of mfaktc v0.21 is being used, or whether there is CUDA level compatibility throughout.

5) I think you're trying an inferior method to launch the mfaktc program. I don't see a need for the start command. It has to be already at a command line to use start to spawn another cmd.exe process, so just use the current cmd session instead.

Windows shortcuts can be much easier, faster, and less prone to pilot error in the long run than typing command lines. Creating a shortcut to launch and run mfaktc, a flavor of cmd /k mfaktc etc is the command line for the shortcut, your working mfaktc directory is the directory to start the process in.

I have shortcuts defined, configured with the equivalent of the command line
C:\Windows\System32\[B]cmd.exe /k[/B] C:\Users\username\Documents\mfaktc-rtx2080\mf.bat
and start in C:\Users\username\Documents\mfaktc-rtx2080
All the device specification, executable flavor selection, setting window title, output redirection, restart or chaining to another task upon crash, etc, is housed in mf.bat for easy reliable quick production running.

6) We don't know from the preceding post whether the installation folder is ok, with correct file and folder permissions & location, or somewhere the OS restricts. But he earlier posted somewhere under desktop. Which should be ok unless properties got oddly set.

7) We still don't know whether there's an mfaktc.ini present, whether it's been customized, what it contains, or whether the OP's introduced issues there. The earlier issues are enough to prevent productive troubleshooting or use, so this can wait a bit.


See also [URL]https://www.mersenneforum.org/showpost.php?p=596919&postcount=8[/URL]


Putting the CUDArt .dll file in the working directory is fine, and simpler than messing with search paths. (Though possibly less space efficient.)

[QUOTE=kriesel;596963]1) Worktodo.txt should be without dates. To my knowledge no legit [URL="https://www.mersenneforum.org/showpost.php?p=522098&postcount=22"]worktodo entry format[/URL] in any GIMPS application includes dates of assignment or any dates.
[/QUOTE]
Small correction: mfaktX doesn't give a dime about the AID field, but it MUST be present. If you have no AID, you can put a comma, or write "N/A" there, or.... use a date. Using a date is a common practice when you have lots of assignments, for many days, to know when you got them (and eventually, where did you get them from, like gimps, gpu72, mersenne.ca, etc), and for example, MISFIT has options/checkboxes to automatically substitute AID with dates or any custom text. So, the OP's assignment lines are correct. What (mess) he does with the "start" command is another story. You explained right for the points 2 to 5.

I usually open a command prompt and type "mfaktc" there :razz:
The rest is organized by the ini file, etc.

(well, I am telling lies, in fact, I only type "misfit" :razz:, but a batch file launching the adequate copies of mfaktc in their specific folders was the working mode for years).

With the help so far I have gotten the program working hooray! I tried two things at once so I dont know which fixed it. One was fixxing the command to only "mfaktc-win-64.exe and the appendages ie -h -st and got it output to a file. The second was just wiping all previous mfaktc files and redownloading it (which btw was mfaktc-0.21.win.cuda11.2-2047). For now I am just using the command but now that its working hopefully I can eventually setup a shortcut. Ideally now id like to get misfit and/or mfaktx controller working so that its a bit more automated.

Oh I guess I should mention how I knew it worked. I just went to the main website and got some manual assignments to copy/paste into my worktodo file and when the program finished running i copy/pasted the contents of my results file into the form and I got credited for my GHZ/Days.

[QUOTE=LaurV;597007]Small correction: mfaktX doesn't give a dime about the AID field, but [B]it MUST be present[/B]. [/QUOTE](bold emphasis in preceding quote mine) No, the AID field is clearly optional for mfaktc. Check the source code. Or try it. I'm running currently, on a GPU, the worktodo line "Factor=1250000029,86,87" which is accepted just fine. Mersenne.ca assignments such as for OBD don't include AIDs and can be run unchanged.[CODE]
got assignment: exp=1250000029 bit_min=86 bit_max=87 (12537.18 GHz-days)
[/CODE]From mfaktc's source code module parse.c, routine parse_worktodo_line, the AID field and its associated trailing comma ARE OPTIONAL:
[CODE] if ((2!=number_of_commas) && (3!=number_of_commas)) // [SIZE=3][B]must have 2 or 3 commas[/B][/SIZE]...
return INVALID_FORMAT;

if(2==number_of_commas)
assignment->assignment_key[0] = '\0';
else
{
strncpy(assignment->assignment_key,ptr,1+(strstr(ptr,",")-ptr) ); // copy the comma..
*strstr(assignment->assignment_key,",") = '\0'; // null-terminate key
ptr=1 + strstr(ptr,",");
}[/CODE]It does appear to accept almost-random text excluding commas in the optional AID field, subject to the max line length 100 and some other restrictions. Thanks for that tip. Reference info is modified to reflect that. That mfaktc accepts almost any arbitrary text as AID input, and discards valid AIDs, not including them in result output, is ... just wrong, in multiple ways. But it is confirmed to be so. Probably the same applies to mfakto, since it was created from mfaktc source code. It would be good to get that corrected in both. The status quo makes it too easy to poach any TF assignment, or forge a fake result to terminate someone's assignment.
[quote]I usually open a command prompt and type "mfaktc" there. The rest is organized by the ini file, etc.[/quote]There is no facility documented in mfaktc.ini for specifying device number. There is apparently no provision for input of device number via mfaktc.ini present in the source code; strings "Device" or "device" do not appear in source module read_config.c. So any system with multiple CUDA GPUs will need the -d option on the command line for running mfaktc on GPU device number > 0.

Well done, next step is to install misfit (if you run windows) to automatize what you did by hand (getting assignments and reporting results). If you run Linux, there are some python scrips doing the same as misfit, but without the graphic interface.
edit: crosspost, I was replying to OP. For the "optional" part, I may be confused... is the gpuowl or the cudaLucas the one requiring the N/A or AID to be present? I know for sure that assignments without it didnt work in the past. All my mfaktc lines contains dates (from gpu72/misfit), generally the date when the exponent was assigned, but not always.

[url]https://www.mersenneforum.org/showpost.php?p=522098&postcount=22[/url]

Gpuowl and Mlucas had requirements for either an AID or a dummy value (0 for gpuowl, 32 0's for Mlucas). Gpuowl was extended to also support "N/A", and didn't require anything for its brief support of TF. I think Mlucas was relaxed to no longer require the fixed length 0's. All the rest treated AID as optional. None support date input, per their documentation, or the fraction of source code I've scanned.

[url]http://v5.mersenne.org/v5design/v5webAPI_0.97.html[/url]

Parameter data types defined in the API (section 2.0) do not include dates, except as a subset of general strings.
Get Assignment (section 5.3) makes no provisions for dates in the request or response.
Assignment progress reports
5.6.2.1 AP "delta date estimated next update " is really specified as delta time in seconds (integer)
similarly "delta date estimated completion" (integer)
5.6.2.2 shows however a mm/dd/yyyy hh:mm:ss example request
I've done AP manually by using curl. It takes seconds input for those deltas. I've never tried date strings.

There appears to be no provision for sending or receiving dates in assignments.

If GPU72/misfit are including dates, it appears to me to be some nonstandard usage.
That probably depends on mfaktx basically discarding the AID string if present.