not only 2-3x more hosts with Nvidia, but also 2-3x daily valid results coming from Nvidia vs AMD.
Not sure why people buy Nvidia. I've purchased thousands of cards for all sorts of uses (no not all for me, for customers), and every single time AMD gave more power for the price.
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
I think that status is a matter of attitude not aptitude.
Tom M
At least this one works. WCG has no status page, and if there was it would be full of red lights. Multi billion dollar research company can't run a Boinc server to keep up with the users!
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
not only 2-3x more hosts with Nvidia, but also 2-3x daily valid results coming from Nvidia vs AMD.
Not sure why people buy Nvidia. I've purchased thousands of cards for all sorts of uses (no not all for me, for customers), and every single time AMD gave more power for the price.
people have many reasons. Better performance. Better driver support. Less issues in general. The Nvidia cards provide better compute density as well as power efficiency here at Einstein. It’s no wonder that nearly all GPGPU projects (BOINC, Folding, rendering, etc) are dominated by Nvidia. The only project where AMD stands a chance is Milkyway with its heavy reliance on FP64, and if you look at the share of compute power coming from the various devices, even Milkyway is dominated by Nvidia.
but the specific reasons are irrelevant here. The reality is that 2/3rds of the valid results come from Nvidia devices, so they deserve the appropriate attention to make sure they have a reliable application.
people have many reasons. Better performance. Better driver support. Less issues in general. The Nvidia cards provide better compute density as well as power efficiency here at Einstein. It’s no wonder that nearly all GPGPU projects (BOINC, Folding, rendering, etc) are dominated by Nvidia. The only project where AMD stands a chance is Milkyway with its heavy reliance on FP64, and if you look at the share of compute power coming from the various devices, even Milkyway is dominated by Nvidia.
but the specific reasons are irrelevant here. The reality is that 2/3rds of the valid results come from Nvidia devices, so they deserve the appropriate attention to make sure they have a reliable application.
I'm not an AMD fanboy, I just buy whatever gives the most bang for the buck. And every time I've compared FP32 output per $, AMD has won hands down. In the CPU market, sometimes it's AMD sometimes it's Intel.
I used to build high end gaming machines for a living, and most people asked for AMD, those who asked for Nvidia, when I showed them the prices and the computing speed, they changed their mind. One wanted a specific nvidia, the brand new card literally melted when I tested it on Boinc. No overheat protection. Straight in the bin and an AMD replaced it.
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
I'm not an AMD fanboy, I just buy whatever gives the most bang for the buck. And every time I've compared FP32 output per $, AMD has won hands down. In the CPU market, sometimes it's AMD sometimes it's Intel.
To each their own...
I'm okay with what your doing and believe in. I'm even okay with your saying that AMD is superior to NVIDIA in FP32 output per $$. But I'm not okay with your trying to convince me and the rest of us on BOINC that AMD is better. Can't you let the topic rest?
I will continue to buy NVIDIA, not because of the performance per $$ when comparing FP32, but because I like the peace of mind that NVIDIA gives me in their software updates (drivers) and by purchasing from EVGA with superior product support.
Sounds like your experience is based in 10-15 years ago. Because Nvidia have been on top for quite a while. They outperform AMD in almost any metric. And Nvidia provides many other value adds to justify going with them. Like DLSS (2.0), and NVENC, of which AMD only has inferior implementations and way less supported than Nvidia. Theoretical compute per dollar means little if the drivers are so bad that cause constant crashing, or they don’t provide the features that are promised. And very few people have free or stolen electric to not care about the power efficiency. Heck, I have “free” electric for a few of my systems and I still won’t waste the power running old power hungry hardware. Makes a lot more sense to get the most compute per watt as possible.
but again, the specific reasons Nvidia is superior isn’t really relevant. They have a huge pool of Nvidia users and they need to make sure the application works and provides valid results. They’ve already hashed out the AMD cards. These tasks are still in beta, and when it’s put in production you’ll have plenty of tasks for your old AMD cards.
I'm okay with what your doing and believe in. I'm even okay with your saying that AMD is superior to NVIDIA in FP32 output per $$. But I'm not okay with your trying to convince me and the rest of us on BOINC that AMD is better. Can't you let the topic rest?
I will continue to buy NVIDIA, not because of the performance per $$ when comparing FP32, but because I like the peace of mind that NVIDIA gives me in their software updates (drivers) and by purchasing from EVGA with superior product support.
The support is the same, mostly they both work, occasionally there are glitches. The only way to choose is FP32/$.
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
Sounds like your experience is based in 10-15 years ago. Because Nvidia have been on top for quite a while.
No, between 1995 and present.
Ian&Steve C. wrote:
They outperform AMD in almost any metric.
Not here they don't. Either you're buying at a different end of the range, or they're cheaper where you live. I have no bias, I just look at the FP32 speed, and the $, then buy the best one.
Ian&Steve C. wrote:
Theoretical compute per dollar means little if the drivers are so bad that cause constant crashing,
Doesn't happen here.
Ian&Steve C. wrote:
And very few people have free or stolen electric to not care about the power efficiency. Heck, I have “free” electric for a few of my systems and I still won’t waste the power running old power hungry hardware. Makes a lot more sense to get the most compute per watt as possible.
Not if you have loads of free electricity, like a big solar array. The limitation is the cost of buying the hardware.
Ian&Steve C. wrote:
but again, the specific reasons Nvidia is superior isn’t really relevant. They have a huge pool of Nvidia users and they need to make sure the application works and provides valid results. They’ve already hashed out the AMD cards. These tasks are still in beta, and when it’s put in production you’ll have plenty of tasks for your old AMD cards.
Agreed, they do what's best for them. Or easiest (one card might be easier to get going more quickly). I was just pointing out I'm amazed anyone ever finds Nvidia a good buy. Just like I don't buy Merc cars. They're damn good, but the price, no way.
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
GWGeorge007 wrote: Uh...
)
I'm surprised nobody mentioned we're discussing raidio waves.
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
GWGeorge007 wrote: Bernd
)
I think that status is a matter of attitude not aptitude.
Tom M
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor)
Ian&Steve C. wrote: not only
)
Not sure why people buy Nvidia. I've purchased thousands of cards for all sorts of uses (no not all for me, for customers), and every single time AMD gave more power for the price.
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
Tom M wrote: I think that
)
At least this one works. WCG has no status page, and if there was it would be full of red lights. Multi billion dollar research company can't run a Boinc server to keep up with the users!
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
Peter Hucker of the Scottish
)
people have many reasons. Better performance. Better driver support. Less issues in general. The Nvidia cards provide better compute density as well as power efficiency here at Einstein. It’s no wonder that nearly all GPGPU projects (BOINC, Folding, rendering, etc) are dominated by Nvidia. The only project where AMD stands a chance is Milkyway with its heavy reliance on FP64, and if you look at the share of compute power coming from the various devices, even Milkyway is dominated by Nvidia.
but the specific reasons are irrelevant here. The reality is that 2/3rds of the valid results come from Nvidia devices, so they deserve the appropriate attention to make sure they have a reliable application.
_________________________________________________________________________
Ian&Steve C. wrote: people
)
I'm not an AMD fanboy, I just buy whatever gives the most bang for the buck. And every time I've compared FP32 output per $, AMD has won hands down. In the CPU market, sometimes it's AMD sometimes it's Intel.
I used to build high end gaming machines for a living, and most people asked for AMD, those who asked for Nvidia, when I showed them the prices and the computing speed, they changed their mind. One wanted a specific nvidia, the brand new card literally melted when I tested it on Boinc. No overheat protection. Straight in the bin and an AMD replaced it.
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
Peter Hucker of the Scottish
)
To each their own...
I'm okay with what your doing and believe in. I'm even okay with your saying that AMD is superior to NVIDIA in FP32 output per $$. But I'm not okay with your trying to convince me and the rest of us on BOINC that AMD is better. Can't you let the topic rest?
I will continue to buy NVIDIA, not because of the performance per $$ when comparing FP32, but because I like the peace of mind that NVIDIA gives me in their software updates (drivers) and by purchasing from EVGA with superior product support.
Proud member of the Old Farts Association
Sounds like your experience
)
Sounds like your experience is based in 10-15 years ago. Because Nvidia have been on top for quite a while. They outperform AMD in almost any metric. And Nvidia provides many other value adds to justify going with them. Like DLSS (2.0), and NVENC, of which AMD only has inferior implementations and way less supported than Nvidia. Theoretical compute per dollar means little if the drivers are so bad that cause constant crashing, or they don’t provide the features that are promised. And very few people have free or stolen electric to not care about the power efficiency. Heck, I have “free” electric for a few of my systems and I still won’t waste the power running old power hungry hardware. Makes a lot more sense to get the most compute per watt as possible.
but again, the specific reasons Nvidia is superior isn’t really relevant. They have a huge pool of Nvidia users and they need to make sure the application works and provides valid results. They’ve already hashed out the AMD cards. These tasks are still in beta, and when it’s put in production you’ll have plenty of tasks for your old AMD cards.
_________________________________________________________________________
GWGeorge007 wrote: I'm okay
)
The support is the same, mostly they both work, occasionally there are glitches. The only way to choose is FP32/$.
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
Ian&Steve C. wrote: Sounds
)
No, between 1995 and present.
Not here they don't. Either you're buying at a different end of the range, or they're cheaper where you live. I have no bias, I just look at the FP32 speed, and the $, then buy the best one.
Doesn't happen here.
Not if you have loads of free electricity, like a big solar array. The limitation is the cost of buying the hardware.
Agreed, they do what's best for them. Or easiest (one card might be easier to get going more quickly). I was just pointing out I'm amazed anyone ever finds Nvidia a good buy. Just like I don't buy Merc cars. They're damn good, but the price, no way.
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.