They should, you have twice the processing power. For example one card can do odd numbered frames and the other can do even numbers. The programmers would have to be monumentally terrible not to achieve x2.

It's amazing how much you think you know but don't actually know.

Like I said, the programmers screwed up, if you have twice the manpower, the job is done in half the time. Just like when you pause a game the GPU usage remains high. What's it calculating?

It's also cute that you post the specs of GPUs when that's only half the picture. Paper benchmarks NEVER work as desired.

Nothing does what it says on the tin, but you can use them for comparisons. If your car stated 40mpg, and mine states 30, we can agree neither will get that, but yours will be about 1/3 better.

I've owned many AMD GPUs in the past and always regret it. Last AMD card I owned was a 5700 XT and the gaming performance wasn't all that great and don't even get me started on the AMD GPU drivers. Constant GPU driver crashes until I switched that GPU out with a 3070Ti and absolutely no more crashing and smooth gameplay.

Odd I've had the exact opposite experience.

I root for AMD to come out with good GPUs just because I dislike the fact that Nvidia controls the market and the price. Until that day comes, Nvidia from here on out.

And you're funding their monopoly.

And did you say a grand earlier? A grand is the cost of a whole computer, not one part of it. I guess someone has to buy them new so I can get them 2nd hand. The ones running MW here cost me £40 each.

---

The above was double spaced between sentences, I apologise for the forum software ruining my post.

Sad to see the project gone.
I picked it, because the HD4870 GPU only got the same output as a GTX 8800GT in Folding@Home way back. So i searched for a project that could utilize the ATI cards better, and i ended up here at Milkyway@home.

I am not happy with Nvidia, they push people into their closed ecosystem.

I was looking for an MI210 Accelerator for Milkyway, but now there is no use for FP64 GPUs in Boinc it seems.

I will put my GPUs on Einstein@home, but get my CPUs running for Milkyway. Checking out the appconfig posted early here and see how well it works :)

Edit: Windows picked a good time to remove the AMD GPU driver with an update :D

I was looking for an MI210 Accelerator for Milkyway, but now there is no use for FP64 GPUs in Boinc it seems.

It likely wouldn’t have worked anyway. The new drivers needed for new AMD GPUs never seemed to work with the Separation app because of the old code.

---

Sad to see my go-to GPU project shut down; but since the scientific objective is complete, it's understandable.
My PC is not available anymore for heavy CPU crunching, and thus I have disabled n-body some time ago.

So long, and thanks for all the fish!

---

Checking out the appconfig posted early here and see how well it works :)

Those are relevant to a specific computer, the one posted (I copy below) makes no sense to me. Looks like it's for a 6 core machine, and for some reason splitting into three twos. Why not just run one six? I also make the avg lower than the nthreads, so it runs more than it should, because there's always one in startup phase using only one core, which allows the others to take over the unused cores.

<app_config>

<app_version>
<app_name>milkyway_nbody</app_name>

<plan_class>mt</plan_class>
<avg_ncpus>2</avg_ncpus>
<cmdline>--nthreads 2</cmdline>
</app_version>
<project_max_concurrent>3</project_max_concurrent>

</app_config>

---

The above was double spaced between sentences, I apologise for the forum software ruining my post.

They should, you have twice the processing power. For example one card can do odd numbered frames and the other can do even numbers. The programmers would have to be monumentally terrible not to achieve x2.

It's amazing how much you think you know but don't actually know.

Like I said, the programmers screwed up, if you have twice the manpower, the job is done in half the time. Just like when you pause a game the GPU usage remains high. What's it calculating?[/quote]

If it was possible it would have been done. There is a reason why dual GPU setups are not popular and are not being supported much anymore.

Nothing does what it says on the tin, but you can use them for comparisons. If your car stated 40mpg, and mine states 30, we can agree neither will get that, but yours will be about 1/3 better.

On your paper the AMD card is better than the 4900, which it's not. Go look at gaming benchmarks. The 4090 does better in nearly every game.

Odd I've had the exact opposite experience.[/quote]

Its a widely known problem. I'm not the only one.

And you're funding their monopoly.

I prefer to play games at the best possible experience. Nvidia offers that. I have to pay it.

And did you say a grand earlier? A grand is the cost of a whole computer, not one part of it. I guess someone has to buy them new so I can get them 2nd hand. The ones running MW here cost me £40 each.

I never said price. You are confusing me with someone else.

If it was possible it would have been done. There is a reason why dual GPU setups are not popular and are not being supported much anymore.

I assumed it was because a modern single one was fast enough. Less electricity, smaller power supply, less hassle fitting it in a big case.

Since you appear to know everything, how on earth did they manage to use twice the computing power to achieve less than twice the result?

On your paper the AMD card is better than the 4900, which it's not. Go look at gaming benchmarks. The 4090 does better in nearly every game.

Surely games use texture and pixel rates, which I quoted?

Its a widely known problem. I'm not the only one.

And when I had Nvidia problems, I saw everyone else with Nvidia problems. All drivers/software/any program is rubbish. Programmers always have been and always will be useless, they make mistakes all the time.

I prefer to play games at the best possible experience. Nvidia offers that. I have to pay it.

You could always use triple AMDs. And just what are you powering anyway? A 100 inch 8K screen at 120fps? Any card on sale now gives a decent gaming experience. I play 3D shooting games at HD with a Radeon R9 Nano!

I never said price. You are confusing me with someone else.

Solidair. He said "Your scale is off for current GPU'S 7900xtx is about a grand"

You two are so similar I thought you were the same person.

---

The above was double spaced between sentences, I apologise for the forum software ruining my post.

I assumed it was because a modern single one was fast enough. Less electricity, smaller power supply, less hassle fitting it in a big case.

Even modern GPUs struggle with high resolution games. Dual, triple or quad GPUs was the only way to get close to getting high frames on modern games with a high resolution, but it wasn't 100% increase performance per card added. Never has been.

Since you appear to know everything, how on earth did they manage to use twice the computing power to achieve less than twice the result?

Need some context here. What uses twice the power of what to achieve what?

Surely games use texture and pixel rates, which I quoted?

Well yes they do, but the 4090 provides high frame rates for nearly all games. Only a few select outliners where the AMD card was higher. Realistically, the top model AMD card is more comparable to the 4080. Similar price, similar performance.

And when I had Nvidia problems, I saw everyone else with Nvidia problems. All drivers/software/any program is rubbish. Programmers always have been and always will be useless, they make mistakes all the time.

No one is perfect. Game developers on the PC have to build a game that's compatible with thousands, millions even, of possible hardware combinations. That's not exactly easy.

You could always use triple AMDs.

Why would I buy 3x the amount of GPUs to do what a single GPU can achieve? While I do have systems with multiple GPUs and I've even had systems that I gamed on with SLI'd GPUs in the past. It's not worth it. Outside of those previous points, not all games even benefit at all from multiple GPUs.

And just what are you powering anyway? A 100 inch 8K screen at 120fps?

The size of the display is irrelevant. A 10 inch 8k display or a 100 inch 8k display will require the same processing power to achieve the same results.

Any card on sale now gives a decent gaming experience. I play 3D shooting games at HD with a Radeon R9 Nano!

I have dual 4k displays with a third I plan on purchasing soon. These displays have a 144hz refresh rate where I try to get the FPS as close to 144 FPS as possible for the best possible experience.

Solidair. He said "Your scale is off for current GPU'S 7900xtx is about a grand"

You two are so similar I thought you were the same person.

I'm sure he's just as offended as I am about that assumption.

Since you appear to know everything, how on earth did they manage to use twice the computing power to achieve less than twice the result?

your mistake is assuming that it's in fact using twice the "computing power". having two cards installed does not mean they are both being used 100%

it came down to a combination of optimization (or lack of), game adoption (or lack of), and other system bottlenecks. this is where your two-workers analogy also fails, since it makes assumptions that they are working in a wide open area and allowed to work in parallel without any interference. when SLI/Xfire was rarely optimized to allow them both operate at full capacity. it was not uncommon for the secondary card to be operating at 0-30% GPU load. that can't be equated to "twice the computing power"

it was extra work for game devs to make the game work properly with either SLI or Xfire so many elected to not even bother. there were only a handful of games or benchmarks properly coded and optimized for SLI to get it to scale close to 1:1 when adding additional cards. games and benchmarks that took the time to optimize SLI/Xfire did indeed see close to 2x performance with a second card and the second card did in those cases show full utilization.

---

Thank you, but my donations here on Milkyway@home have been exclusively GPUs. Reason for that is that CPUs are doing other, more important, down to Earth projects. While my GPUs, from the old days of SETI@home, have been doing the work for the mysteries of the Universe.

I am not so glad that Milkyway@home is ending their GPU units, as I was really fond of your app. Mostly, I could work without noticing that Milkyway@home is done in the background.

So now I will be switching to other options, that I donate to. Some of whom do not have badges on their progress & donation(s).
Wishing you best of luck & if there is a GPU return, my computers will be ready. 👍😎

---

non-profit org. Play4Life in Zagreb, Croatia, EU

Well if you like Space related projects as previously mentioned Einstein has currently a number of applications that support various GPU's ... AMD IBM NVIDIA Even MAC OS running on Intel has GPU options. Several of these applications are waiting with tasks available. Einstein has the most flexible User friendly settings for With our Without CPU GPU etc so you can have what you want. You can look at Server Status for which applications have work now.

Enjoy
Bill F

---

In October of 1969 I took an oath to support and defend the Constitution of the United States against all enemies, foreign and domestic;
There was no expiration date.

My "goal" is contribute the most I can per unit of heat I am outputting into my room, in the middle of summer. Way to discount our intentions,Tom...

Checking out the appconfig posted early here and see how well it works :)

Those are relevant to a specific computer, the one posted (I copy below) makes no sense to me. Looks like it's for a 6 core machine, and for some reason splitting into three twos. Why not just run one six? I also make the avg lower than the nthreads, so it runs more than it should, because there's always one in startup phase using only one core, which allows the others to take over the unused cores.

No it's for any of my multi-core pc's that are also running tasks from other projects and I'm limiting it to 6 cpu cores for Milky Way to use. I prefer the 3 double cpu tasks because they take longer for me to crunch and they also don't get hung up like the 16 cpu core tasks do. I never run both cpu and gpu tasks from the same project on my pc's, and no I'm not running multiple Boinc setups on my pc's, so this is an easy way for me to limit the tasks in this case for Milky Way and also run tasks from other projects at the same time. Just for info I was also running the Yoyo ecmP2 tasks at the same time and had the last line set to 1 to 4 tasks depending on how much ram was in each pc, I deleted all the other lines except the final line when I put it in the Yoyo folder.

<app_config>

<app_version>
<app_name>milkyway_nbody</app_name>

<plan_class>mt</plan_class>
<avg_ncpus>2</avg_ncpus>
<cmdline>--nthreads 2</cmdline>
</app_version>
<project_max_concurrent>3</project_max_concurrent>

</app_config>

OK lots of people are looking for Astrophysics related projects for their GPU's ... at least for NVIDIA I have a question.

Asteroids shows two windows apps that show "(cuda55)" in the Name / version. Is this an indication of a NVIDIA Application ?

Bill F

Yes.
It also has cuda102 and cuda118

Checking out the appconfig posted early here and see how well it works :)

Those are relevant to a specific computer, the one posted (I copy below) makes no sense to me. Looks like it's for a 6 core machine, and for some reason splitting into three twos. Why not just run one six? I also make the avg lower than the nthreads, so it runs more than it should, because there's always one in startup phase using only one core, which allows the others to take over the unused cores.

No it's for any of my multi-core pc's that are also running tasks from other projects and I'm limiting it to 6 cpu cores for Milky Way to use. I prefer the 3 double cpu tasks because they take longer for me to crunch and they also don't get hung up like the 16 cpu core tasks do. I never run both cpu and gpu tasks from the same project on my pc's, and no I'm not running multiple Boinc setups on my pc's, so this is an easy way for me to limit the tasks in this case for Milky Way and also run tasks from other projects at the same time. Just for info I was also running the Yoyo ecmP2 tasks at the same time and had the last line set to 1 to 4 tasks depending on how much ram was in each pc, I deleted all the other lines except the final line when I put it in the Yoyo folder.

<app_config>

<app_version>
<app_name>milkyway_nbody</app_name>

<plan_class>mt</plan_class>
<avg_ncpus>2</avg_ncpus>
<cmdline>--nthreads 2</cmdline>
</app_version>
<project_max_concurrent>3</project_max_concurrent>

</app_config>

I do wonder which would be more efficient: a bunch of small (2T) Nbody tasks, or a smaller number of larger (4T) tasks, or a single very large (8T) task. Right now I've got mine set up as 2 avg cpus, 2 threads per task for the command line, and 4 max concurrent tasks for 4 tasks, 8 threads taken up by MW@H. Guess I'll have to experiment and see how my RAC changes.

Even modern GPUs struggle with high resolution games.

And yet I play fine in HD with an old 8000GFlops Nano.

Since you appear to know everything, how on earth did they manage to use twice the computing power to achieve less than twice the result?

Need some context here. What uses twice the power of what to achieve what?

I thought that was obvious. Two cards, twice the computing power, should achieve twice the calculations and render twice as many frames in the game, just like a twice as fast card.

Surely games use texture and pixel rates, which I quoted?

Well yes they do

So why can't I use those rates to estimate how fast they are at games?

but the 4090 provides high frame rates for nearly all games. Only a few select outliners where the AMD card was higher. Realistically, the top model AMD card is more comparable to the 4080. Similar price, similar performance.

No one is perfect. Game developers on the PC have to build a game that's compatible with thousands, millions even, of possible hardware combinations. That's not exactly easy.

I think it's more to do with things getting so much more complex, programmers divide things into chunks. Just like building a massive building, everyone works on a different part, then finds something the plumber wants to do ruins the structural integrity.

You could always use triple AMDs.

Why would I buy 3x the amount of GPUs to do what a single GPU can achieve?

Because you said you hated Nvidia. And since AMDs are cheaper the three of them wouldn't set you back much more.

And just what are you powering anyway? A 100 inch 8K screen at 120fps?

The size of the display is irrelevant. A 10 inch 8k display or a 100 inch 8k display will require the same processing power to achieve the same results.

A larger screen requires a higher resolution to be sharp.

I have dual 4k displays with a third I plan on purchasing soon. These displays have a 144hz refresh rate where I try to get the FPS as close to 144 FPS as possible for the best possible experience.

Ah, you live in another world, one where people have an extra zero on their salary.

---

The above was double spaced between sentences, I apologise for the forum software ruining my post.

your mistake is assuming that it's in fact using twice the "computing power". having two cards installed does not mean they are both being used 100%

it came down to a combination of optimization (or lack of), game adoption (or lack of), and other system bottlenecks. this is where your two-workers analogy also fails, since it makes assumptions that they are working in a wide open area and allowed to work in parallel without any interference. when SLI/Xfire was rarely optimized to allow them both operate at full capacity. it was not uncommon for the secondary card to be operating at 0-30% GPU load. that can't be equated to "twice the computing power"

it was extra work for game devs to make the game work properly with either SLI or Xfire so many elected to not even bother. there were only a handful of games or benchmarks properly coded and optimized for SLI to get it to scale close to 1:1 when adding additional cards. games and benchmarks that took the time to optimize SLI/Xfire did indeed see close to 2x performance with a second card and the second card did in those cases show full utilization.

Surely a game is pretty similar to Boinc, a mass of very similar calculations to do (in the case of games, raytracing). A very easy thing for the program on the CPU to send half to one card and half to the other. Alternate lines, alternate frames, whatever.

---

The above was double spaced between sentences, I apologise for the forum software ruining my post.

No it's for any of my multi-core pc's that are also running tasks from other projects and I'm limiting it to 6 cpu cores for Milky Way to use. I prefer the 3 double cpu tasks because they take longer for me to crunch and they also don't get hung up like the 16 cpu core tasks do.

So it's specific to your requirements. I have no problem with any Nbody right up to 16, they never get stuck. I also don't see the point in making them take longer to crunch.

I never run both cpu and gpu tasks from the same project on my pc's

Why not? I do if they're different things. Pointfull here, pointless at Einstein (where they can make any of their work run on a GPU).

---

The above was double spaced between sentences, I apologise for the forum software ruining my post.

OK lots of people are looking for Astrophysics related projects for their GPU's ... at least for NVIDIA I have a question.

Asteroids shows two windows apps that show "(cuda55)" in the Name / version. Is this an indication of a NVIDIA Application ?

Bill F

Update... For people looking for Astrophysics related projects for their GPU's ...

Asteroids@home has NVIDIA app's for the following system groups

Microsoft Windows (98 or later) running on an Intel x86-compatible CPU
Microsoft Windows running on an AMD x86_64 or Intel EM64T CPU
Linux running on an AMD x86_64 or Intel EM64T CPU

Enjoy
Bill F