I run 1 brp7/meerKat at a time on my 5700g. I usually am set to only use 14 out of 16 threads overall. I have not noticed the brp7 slowing down when the CPU is loaded. But I don't usually run Asteroids at home.
Hmm... that's pretty slow than, with 12 threads loaded with Milkyway (not Asteroids, that might be different with all those power limits in modern CPUs) I get one BRP7 done in about 3950 seconds, now I added one more Milkyway and that seems to add something like 50 seconds to that, which is still OK I think. I will test 14 later, still experimenting.
Tom M wrote:
I am running 4 memory modules (8m x 4 slots).
I have two sticks DDR4 @3400MT/s. Actually they are 3600, but the system wasn't stable at anything above 3400, @3400 it seems perfectly stable so far.
Tom M wrote:
It looks like the reported floating point test for Boinc is as least 10% faster than my 5700G. I have made no attempt to overclock my Cpu/iGpu's.
Not overclocked either except the RAM, just undervolted. Yes, I was pretty lucky with that chip, -30 on all cores and -20 on the iGPU. -30 on iGPU worked for Moo! but not for BRP7, that's where the one errored task actually comes from. Doesn't make much difference in power consumption, so I guess I'll leave it where it is now as long as all tasks validate.
With Milkyway using pretty little energy compared to anything else, the CPU is boosting constantly @4.6GHz even if I run 16 of their single core N-Body WUs concurrently, but than there's nothing for the GPU left and that would run @400MHz most of the time, so now I'm searching for optimal amount of threads on the CPU while watching what the hardware is doing with HWiNFO64. I guess with Asteroids or PrimeGrid with their pretty highly optimized applications, the situation might be completely different.
So I should be trying Milkyway not Asteroids? How many threads per task are you running on Milkyway? I think I have mine set to 4 cpu threads per task.
I was looking at the "dedicated memory" on my task manager. It defaults to 512MB. But the bios offers "auto" (the default) upto 2GB. So I reset that. And now the dedicated memory looks to be using about 50% out of 2GB. So there might be some iGpu speed up there?
My memory is 3400MHz.
Like I said. Interesting Questions.
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
So I should be trying Milkyway not Asteroids? How many threads per task are you running on Milkyway? I think I have mine set to 4 cpu threads per task.
Yes, Milkyway is very "lightweight". I use just one thread per task. If running on all 16 threads, compared to multithreaded tasks this seems to reduce the power consumption by about 4W while increasing the (calculated) RAC from 14k to 22k. So multithreaded N-Body is dead to me. ;-)
Tom M wrote:
I was looking at the "dedicated memory" on my task manager. It defaults to 512MB. But the bios offers "auto" (the default) upto 2GB. So I reset that. And now the dedicated memory looks to be using about 50% out of 2GB. So there might be some iGpu speed up there?
I think that's what i reserved for the GPU just as absolute minimum while it can get up to 12365MB if it needs. Not sure about it though.
Tom M wrote:
Like I said. Interesting Questions.
Yes, very interesting in particular when comming from a Core 2 Duo like me, which would simply draw all the energy it needs from the socket.
I just looked at my latest processing times for Einstein at home on the 5700G iCpu. It looks like they have dropped about 25 percent to the same range I saw for your system.
It looks like I will have to re-add the 8 thread task from PrimeGrid to see if that type of load makes a difference.
Still need to replicate your Milkyway test as well a multi-threading version of Milkyway. My default MW is 4 threads.
====edit=====
I ran the automatic MB optimizer (in the bios setup) a couple of days ago. And while I only looked at some of the settings briefly it appears my memory is now running at 2400 MHz and the CPU is running about 4.27 GHz. The memory was set on the D.O.C. aka: 3200 MHz and certain voltages. And the cpu was running on "auto" as high as 4.12 GHz but sometimes lower with less load.
If slower memory produces faster iGpu processing who am I to complain? :)
Tom M
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
It looks like I will have to re-add the 8 thread task from PrimeGrid to see if that type of load makes a difference.
I'm pretty sure it does.
As far as I can tell from the test I made yesterday, the APU is mainly limited by the 95A EDC limit resulting in some idle states mainly in the iGPU part of the chip ("effective" Speed dropping in HWiNFO). Changing PBO limits in BIOS from "Auto" to "Motherboard" raised that limit to 210A (113A were used while crunching 13x Milkyway + 1x BRP7) and both the real and the effective speed of the iGPU were pretty much constantly @2000MHz, even with 14x Milkyway.
However I'm not sure I feel comfortable with pushing the chip like that, so changed that back for now. So far I have not found any useful information about what is safe for 24/7 use like for BOINC and if it's only about cooling.
So my question is, does it share the architecture of the latest gen AMD GPU's? Or is this another you have to custom write your own code using a HPC framework?
I do understand FPGA's are programmable.
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
So how much slot space does a Radeon VII take up? I am pretty sure it is not a strick two slot card. If it is I would be tempted to run 4 like I can with Nvidia Titan V's.
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
So how much slot space does a Radeon VII take up? I am pretty sure it is not a strick two slot card. If it is I would be tempted to run 4 like I can with Nvidia Titan V's.
Tom, it is a 2 slot card with three fans. https://www.techpowerup.com/gpu-specs/radeon-vii.c3358 I'd be a little concerned about the heat build up with 3 cards in a computer unless you have something with 8 slots, or water cooled GPUs. Or use a mining rig with cables... which I know you can't get them to work.
I have seen more than one GPU listed as a two slot that were not. They usually can't be installed together like a Nvidia Titan V can.
Often it is a tiny difference but kept me from fitting them together.
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
This is interesting on the two slot issue and a blower design issue.
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
Link wrote: Tom M wrote:I
)
So I should be trying Milkyway not Asteroids? How many threads per task are you running on Milkyway? I think I have mine set to 4 cpu threads per task.
I was looking at the "dedicated memory" on my task manager. It defaults to 512MB. But the bios offers "auto" (the default) upto 2GB. So I reset that. And now the dedicated memory looks to be using about 50% out of 2GB. So there might be some iGpu speed up there?
My memory is 3400MHz.
Like I said. Interesting Questions.
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
Tom M wrote:So I should be
)
Yes, Milkyway is very "lightweight". I use just one thread per task. If running on all 16 threads, compared to multithreaded tasks this seems to reduce the power consumption by about 4W while increasing the (calculated) RAC from 14k to 22k. So multithreaded N-Body is dead to me. ;-)
I think that's what i reserved for the GPU just as absolute minimum while it can get up to 12365MB if it needs. Not sure about it though.
Yes, very interesting in particular when comming from a Core 2 Duo like me, which would simply draw all the energy it needs from the socket.
.
@Link,I just looked at my
)
@Link,
I just looked at my latest processing times for Einstein at home on the 5700G iCpu. It looks like they have dropped about 25 percent to the same range I saw for your system.
It looks like I will have to re-add the 8 thread task from PrimeGrid to see if that type of load makes a difference.
Still need to replicate your Milkyway test as well a multi-threading version of Milkyway. My default MW is 4 threads.
====edit=====
I ran the automatic MB optimizer (in the bios setup) a couple of days ago. And while I only looked at some of the settings briefly it appears my memory is now running at 2400 MHz and the CPU is running about 4.27 GHz. The memory was set on the D.O.C. aka: 3200 MHz and certain voltages. And the cpu was running on "auto" as high as 4.12 GHz but sometimes lower with less load.
If slower memory produces faster iGpu processing who am I to complain? :)
Tom M
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
Tom M wrote:It looks like I
)
I'm pretty sure it does.
As far as I can tell from the test I made yesterday, the APU is mainly limited by the 95A EDC limit resulting in some idle states mainly in the iGPU part of the chip ("effective" Speed dropping in HWiNFO). Changing PBO limits in BIOS from "Auto" to "Motherboard" raised that limit to 210A (113A were used while crunching 13x Milkyway + 1x BRP7) and both the real and the effective speed of the iGPU were pretty much constantly @2000MHz, even with 14x Milkyway.
However I'm not sure I feel comfortable with pushing the chip like that, so changed that back for now. So far I have not found any useful information about what is safe for 24/7 use like for BOINC and if it's only about cooling.
.
https://wccftech.com/amd-anno
)
https://wccftech.com/amd-announces-mass-production-of-the-alveo-v80-compute-accelerator-9495-price-tag/
So my question is, does it share the architecture of the latest gen AMD GPU's? Or is this another you have to custom write your own code using a HPC framework?
I do understand FPGA's are programmable.
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
So how much slot space does a
)
So how much slot space does a Radeon VII take up? I am pretty sure it is not a strick two slot card. If it is I would be tempted to run 4 like I can with Nvidia Titan V's.
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
Tom M wrote:So how much
)
Tom, it is a 2 slot card with three fans. https://www.techpowerup.com/gpu-specs/radeon-vii.c3358 I'd be a little concerned about the heat build up with 3 cards in a computer unless you have something with 8 slots, or water cooled GPUs. Or use a mining rig with cables... which I know you can't get them to work.
Proud member of the Old Farts Association
George, I have seen more
)
George,
I have seen more than one GPU listed as a two slot that were not. They usually can't be installed together like a Nvidia Titan V can.
Often it is a tiny difference but kept me from fitting them together.
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
https://videocardz.com/newz/a
)
https://videocardz.com/newz/asrock-first-to-launch-radeon-rx-7900-xtx-gpu-with-blower-fan-and-12v-2x6-power-connector
This is interesting on the two slot issue and a blower design issue.
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
Tom M wrote: George, I have
)
This pic is from the https://www.techpowerup.com/review/amd-radeon-vii-unboxing/ review.
It is a true 2 slot GPU.
Proud member of the Old Farts Association