CCminer(SP-MOD) Modded GPU kernels.

[Cryptozillah]

I changed my swapfile from 8GB to 16Gb now on the problem rig, i hope it will run fine now.

I'm unsure if its the rig shutting down or the miner, but if its the rig I'd check for overheating problems "CPU or motherboard".
If its the miner then the swapfile may fix that.

I dont think it is the miner crashing, i am pretty sure it is the system shutting down.
GPU temps are around 65 degrees during full load so it is not a heat problem.

I now have 8GB ram in the rig and 16GB swap file, i will let it run unclocked now and see what happends.

Could be something like weak rail "IE 12v, 3.3, 5v" on  the PSU taking a dip under load tripping the PSU to shut down.

I will let it run now and see if it is working better now.
If not i will start checking other stuff. I know this PSU is working perfectly, i have used all cables i use now in another rig before putting the 980s in it.

Okay if you say so but things in a computer go bad in a blink of a eye. I've done tech support for a major Motherboard Co. for over 20 years.
I've seen just about every type of a problem. Shut down is usually caused by heat, PSU or other faulty hardware.

Yeah i know.
It is nothing going warm in this rig, it is standing next to a partly open window and i have been monitoring all the temps.
I will do some more troubleshooting if it wont run without problems now.

If the problem started immediately after instaling the 980s it's something related to that, possibly resource related,
heat, power or the cards themselves.

If it previously ran stable with the 980s it is more likely a HW failure.

The nature of the shutdown might give a clue. If the system powers off it's likely a power issue. If it drops to BIOS
it is likely some other protection tripping. If it reboots it's likely a simple HW fault or OS related.

If you have consistently reproduceable symptoms you can start juggling things around to see if things change. Since
you have other rigs available you can swap RAM, cables etc to try to move the symptoms, which would clearly identify
the faulty component.

The problem started right away and i was spending a few days to sort things out.
I have not been around when it is shutting down so i dont know exactly how it happends. But when i get back and see that it is offline it is completely turned off.
I am almost certain that it just totally crashes and shutting off like in a second, because when firing it back up i get a message that my system ran into a problem etc.
I have already swapped most parts when troubleshooting so i start to run out of options.

I hope it was memory related but time will tell

[1715649592]

1715649592

[1715649592]

1715649592

[1715649592]

1715649592

[bathrobehero]

The problem started right away and i was spending a few days to sort things out.
I have not been around when it is shutting down so i dont know exactly how it happends. But when i get back and see that it is offline it is completely turned off.
I am almost certain that it just totally crashes and shutting off like in a second, because when firing it back up i get a message that my system ran into a problem etc.
I have already swapped most parts when troubleshooting so i start to run out of options.

I hope it was memory related but time will tell

What problem? A standard disk checking or a post boot windows error message?

A sudden power loss never causes a post boot windows error message so that's useful information.

Based on your symptoms so far I would first change the riser cables, blow the dust out of the PCI-E slots and if that doesn't work I'd run the rig with only half the cards enabled and see if it crashes. Then I'd run the rig with the other half of the cards.

I had a few rigs randomly crashing periodically but mostly with cudaminer or if I used --cpu-priority 5 with ccminer. A couple of times it was with bad USB risers (no visual clues though).

[Cryptozillah]

The problem started right away and i was spending a few days to sort things out.
I have not been around when it is shutting down so i dont know exactly how it happends. But when i get back and see that it is offline it is completely turned off.
I am almost certain that it just totally crashes and shutting off like in a second, because when firing it back up i get a message that my system ran into a problem etc.
I have already swapped most parts when troubleshooting so i start to run out of options.

I hope it was memory related but time will tell

What problem? A standard disk checking or a post boot windows error message?

A sudden power loss never causes a post boot windows error message so that's useful information.

Based on your symptoms so far I would first change the riser cables, blow the dust out of the PCI-E slots and if that doesn't work I'd run the rig with only half the cards enabled and see if it crashes. Then I'd run the rig with the other half of the cards.

I had a few rigs randomly crashing periodically but mostly with cudaminer or if I used --cpu-affinity 5 with ccminer. A couple of times it was with bad USB risers (no visual clues though).

When i get back in to windows i get a black popup that say that your pc ran into a problem. If it happend again i will take a screenshot, but there is no specific info about what caused the problem.
If it wont work now i will switch the 980s with another 750Ti rig and see if that helps. That 750 rig is running fine as we speak.

But so far the 980 rig is running fine since i had some fixing done on it.

[tbearhere]

I have some problem with my 2x980 rig i started up a few days ago.
I run Win 8.1 on it like the rest of my rigs and it is a clean install.

The rig is shutting itself down like 1-2 times per day. And when i fire it back up Windows tell me that the computer ran into some problems.
This occurs with or without oc on the cards, not sure wth is going on. It happends when mining with both sp mod and the nicehash miner.

I have a 1200w Silverstone strider psu in that rig so it is not a power problem.

I have 4Gb ram in this rig just like the rest of my rigs, but could it be that i need more ram with 980 cards ?
This is the first time i mine with 980s so i am kind of out of ideas here.

Put RAM anyway to test, but... I have 1 rig with similar problem. Did you try "Benchmark" in nicehash miner? Did it run all coins or terminated some?
My rig has problems with some coins and if that is your case it might point to what the problem is

I will test to add more ram later today and see if it help.
I always benchmark all algos in the nicehash miner and i dont get any errors while doing that.

I have rig with 6 960 (4GB) that won't start with less than 8GB RAM, and another with 6 960 ( 2GB) that work with 4GB RAM. And rig with 3 980s that work perfectly with 4 GB RAM. And this 970x2 + 960 with problem with some algos ( neoscrypt, lyra2v2 and so-so with x13 (?) ) I suspect 2 unpowered risers might be a problem...
Double check all connections!

I think adding more swapfile would be enough. I have 4GB RAM in my rigs but with 16GB swapfile which lets me run 4x970 + 2x750 Ti. The swapfile is never used (as it wouldn't use 8GB memory either) but it needs it momentarily.

@bathrobehero ....... What is a swapfile?   thx

[bathrobehero]

@bathrobehero ....... What is a swapfile?   thx

Pagefile or virtual memory.

[tbearhere]

@bathrobehero ....... What is a swapfile?   thx

Pagefile or virtual memory.

Yes..thx.....I never heard it called swapfile.  

[chrysophylax]

@bathrobehero ....... What is a swapfile?   thx

Pagefile or virtual memory.

Yes..thx.....I never heard it called swapfile.  

linux / unix / posix based systems use ( and call them ) swapfiles ( and swap partitions ) ...

it seems simple enough that if you 'swap' data from memory to a hdd drive 'file' - then back again ... the file was sure to be called 'swap' 'file' ...

logical - right microsoft? ... ahem ...

...

#crysx

[chrysophylax]

@bathrobehero ....... What is a swapfile?   thx

Pagefile or virtual memory.

Yes..thx.....I never heard it called swapfile.  

linux / unix / posix based systems use ( and call them ) swapfiles ( and swap partitions ) ...

it seems simple enough that if you 'swap' data from memory to a hdd drive 'file' - then back again ... the file was sure to be called 'swap' 'file' ...

logical - right microsoft? ... ahem ...

...

#crysx

Intel manuals for x86 refer to it as paging. There's actually a good reason for this - pages of memory are mapped into physical memory (and other places.) If it's not in physical memory, it's marked "not present" and accesses to it result in a page fault, running the OSs handler for that, which loads the requested page into memory and returns the address.

and hence why posix systems simplified it into human readable terms ...

ho hum ... its all geek to me ...

hang on - i am one ...

#crysx

[Cryptozillah]

My 980 rig have now been running perfectly for 24 hours after i added 4GB more ram and increased the virtual memory to 16GB.

[frazier34567]

With my 6XEVGA 750 ti SC (No Bios mod, no extra connector, no OC, Max 1320MHz) on LyraREv2 it is 395W In from wall on digital PSU.  Windows 8.1 Pro
Quark 420W in from wall
Idle 90W in from wall

Lyra2v2 on the 750ti should do around 5MHASH each with the correct clocks. so your rig does around 30MHASH if you tune it correctly.
The 980ti only does around 18 MHASH with overclocking and costs  the same as 6 used ti's.

But the 980ti can of course do much better with the right code..

I currently do 4806 per card or 28836 for the rig.  but then I am not pushing my cards.  The only thing i have up is the fan speed at about 70% at 53C I am sure I could do more but it is not worth the time at this current pay rate.

[joblo]

My 980 rig have now been running perfectly for 24 hours after i added 4GB more ram and increased the virtual memory to 16GB.

Glad to see you're running stable.

I'd like to understand more about the pagefile size issue, particularly the notion that's it's needed but not
realy used. Clearly it is being used even if only momentarilly. Does anyone know if this issue exists on Linux?

From my understanding the memory/pagefile requirements increase with the number of GPUs. This suggests that
all GPU threads initialize in parallel, allocate the memory at the same time, use it briefly, then free it. If the threads initialized
serially only one would require large amounts of memory and would free it before the next thread initializes. The dynamic
memory requirements would be reduced and larger rigs could run with less memory and smaller swap space.

Make sense?

[bathrobehero]

My 980 rig have now been running perfectly for 24 hours after i added 4GB more ram and increased the virtual memory to 16GB.

Glad to see you're running stable.

I'd like to understand more about the pagefile size issue, particularly the notion that's it's needed but not
realy used. Clearly it is being used even if only momentarilly. Does anyone know if this issue exists on Linux?

From my understanding the memory/pagefile requirements increase with the number of GPUs. This suggests that
all GPU threads initialize in parallel, allocate the memory at the same time, use it briefly, then free it. If the threads initialized
serially only one would require large amounts of memory and would free it before the next thread initializes. The dynamic
memory requirements would be reduced and larger rigs could run with less memory and smaller swap space.

Make sense?

It's definitely parallel initialization but every software I used to monitor pagefile use, they never showed anything just a few MB tops. But if you run multiple instances of ccminer on the same device, then it will - depending on the algo - use a huge amount of memory and of course pagefile if needed. It's almost like it needs memory/pagefile just in case the GPU rans out of it.

[djm34]

My 980 rig have now been running perfectly for 24 hours after i added 4GB more ram and increased the virtual memory to 16GB.

Glad to see you're running stable.

I'd like to understand more about the pagefile size issue, particularly the notion that's it's needed but not
realy used. Clearly it is being used even if only momentarilly. Does anyone know if this issue exists on Linux?

From my understanding the memory/pagefile requirements increase with the number of GPUs.

Pagefile/RAM requirement is due to the initial memory allocation (cudamalloc) which is done for each GPU and transit through pagefile/ram.
This memory alloc being done first on ram/Pagefile then copied to gpu vram (however it is never deallocated, as it gives the possibility to copy back to cpu/ram portion of what has been allocated...)
(provided, if I am not too wrong in my representation of how it works...)

[AzzAz]

My 980 rig have now been running perfectly for 24 hours after i added 4GB more ram and increased the virtual memory to 16GB.

Glad to see you're running stable.

I'd like to understand more about the pagefile size issue, particularly the notion that's it's needed but not
realy used. Clearly it is being used even if only momentarilly. Does anyone know if this issue exists on Linux?

From my understanding the memory/pagefile requirements increase with the number of GPUs.

Pagefile/RAM requirement is due to the initial memory allocation (cudamalloc) which is done for each GPU and transit through pagefile/ram.
This memory alloc being done first on ram/Pagefile then copied to gpu vram (however it is never deallocated, as it gives the possibility to copy back to cpu/ram portion of what has been allocated...)
(provided, if I am not too wrong in my representation of how it works...)

Definitely important. My " problematic" rig ( miner crashes) had small pagefile.  I added RAM and miner now works on Lyra2v2 and x13, neoscrypt not again. Then I increased Pagefile and voila! - neoscrypt works too... So that is the most memory dependant algo?

[joblo]

My 980 rig have now been running perfectly for 24 hours after i added 4GB more ram and increased the virtual memory to 16GB.

Glad to see you're running stable.

I'd like to understand more about the pagefile size issue, particularly the notion that's it's needed but not
realy used. Clearly it is being used even if only momentarilly. Does anyone know if this issue exists on Linux?

From my understanding the memory/pagefile requirements increase with the number of GPUs.

Pagefile/RAM requirement is due to the initial memory allocation (cudamalloc) which is done for each GPU and transit through pagefile/ram.
This memory alloc being done first on ram/Pagefile then copied to gpu vram (however it is never deallocated, as it gives the possibility to copy back to cpu/ram portion of what has been allocated...)
(provided, if I am not too wrong in my representation of how it works...)

So it would seem that the CPU needs enough virtual memory to match all the GPU memory in use. I can see a preallocation
strategy being better for speed but if the CPU is swapping to disk it would be slower than dynamic memory allocation.

[djm34]

My 980 rig have now been running perfectly for 24 hours after i added 4GB more ram and increased the virtual memory to 16GB.

Glad to see you're running stable.

I'd like to understand more about the pagefile size issue, particularly the notion that's it's needed but not
realy used. Clearly it is being used even if only momentarilly. Does anyone know if this issue exists on Linux?

From my understanding the memory/pagefile requirements increase with the number of GPUs.

Pagefile/RAM requirement is due to the initial memory allocation (cudamalloc) which is done for each GPU and transit through pagefile/ram.
This memory alloc being done first on ram/Pagefile then copied to gpu vram (however it is never deallocated, as it gives the possibility to copy back to cpu/ram portion of what has been allocated...)
(provided, if I am not too wrong in my representation of how it works...)

So it would seem that the CPU needs enough virtual memory to match all the GPU memory in use. I can see a preallocation
strategy being better for speed but if the CPU is swapping to disk it would be slower than dynamic memory allocation.

this memory isn't used anymore once it has been allocated to the vram

[joblo]

My 980 rig have now been running perfectly for 24 hours after i added 4GB more ram and increased the virtual memory to 16GB.

Glad to see you're running stable.

I'd like to understand more about the pagefile size issue, particularly the notion that's it's needed but not
realy used. Clearly it is being used even if only momentarilly. Does anyone know if this issue exists on Linux?

From my understanding the memory/pagefile requirements increase with the number of GPUs.

Pagefile/RAM requirement is due to the initial memory allocation (cudamalloc) which is done for each GPU and transit through pagefile/ram.
This memory alloc being done first on ram/Pagefile then copied to gpu vram (however it is never deallocated, as it gives the possibility to copy back to cpu/ram portion of what has been allocated...)
(provided, if I am not too wrong in my representation of how it works...)

So it would seem that the CPU needs enough virtual memory to match all the GPU memory in use. I can see a preallocation
strategy being better for speed but if the CPU is swapping to disk it would be slower than dynamic memory allocation.

this memory isn't used anymore once it has been allocated to the vram

When memory is not deallocated when it is no longer needed it's usually called a leak. Is this something that cudamalloc
does transparently or does the application have any control?

[bileq]

always
Cuda error in func 'x11_simd512_cpu_init' at line 791 : out of memory.
i dont hope any profit just to test my gpu
its gt9800 with 4gb memory
cuda 7.5 installed
any working config for me?

[joblo]

My 980 rig have now been running perfectly for 24 hours after i added 4GB more ram and increased the virtual memory to 16GB.

Glad to see you're running stable.

I'd like to understand more about the pagefile size issue, particularly the notion that's it's needed but not
realy used. Clearly it is being used even if only momentarilly. Does anyone know if this issue exists on Linux?

From my understanding the memory/pagefile requirements increase with the number of GPUs.

Pagefile/RAM requirement is due to the initial memory allocation (cudamalloc) which is done for each GPU and transit through pagefile/ram.
This memory alloc being done first on ram/Pagefile then copied to gpu vram (however it is never deallocated, as it gives the possibility to copy back to cpu/ram portion of what has been allocated...)
(provided, if I am not too wrong in my representation of how it works...)

So it would seem that the CPU needs enough virtual memory to match all the GPU memory in use. I can see a preallocation
strategy being better for speed but if the CPU is swapping to disk it would be slower than dynamic memory allocation.

this memory isn't used anymore once it has been allocated to the vram

When memory is not deallocated when it is no longer needed it's usually called a leak. Is this something that cudamalloc
does transparently or does the application have any control?

Wrong - it's only LEAKED if the pointer to that memory is lost, meaning you couldn't deallocate it if you wanted to - and it happens in repeated code. To "leak," you actually have to slowly continue to eat memory until there's no more left.

You are technically correct, perhaps "hog" would be a better term. That doesn't change the point that large amounts
of CPU memory remain allocated after it is no longer needed.

[scryptr]

always
Cuda error in func 'x11_simd512_cpu_init' at line 791 : out of memory.
i dont hope any profit just to test my gpu
its gt9800 with 4gb memory
cuda 7.5 installed
any working config for me?

GT9800--

A GT9800 was top of the line once.  I have a GT9800+ and it will mine scrypt at 14kh/s with CudaMiner, early 2014 vintage software, and written to compile on CUDA 5.5.  The GT9800 simply does not have the capacity to mine at a reasonable rate.  It will not perform at all with software designed for the Maxwell chipset, the circuitry is not there.  The version of CCminer written by sp_ and discussed in this thread is specifically written to only work on Maxwell chip architecture.       --scryptr