a am now running Einstein@home, physics,SiDock@home, biology,Numberfield, mathematics. I use to cooperate with CERN, which has jectme a letter of thanks because I helped them, in atarting a project using VirtualBox, Testing4Theory@home which has now a different name, The letter was signed by Ben Segal, a member of the internet Council. I have I also a letter by Roger Penrose in 1995, before he was made Nobel Prize.
Hello Tullio,
I know you're from Italy and nearly 80 years old, but your message doesn't seem to have a question in it related to the topic. Are you just making a statement? Or did you need help with something related to the topic?
I need the iGPU to run Einstein.I could not run GPUgrid since it needs CUDA.
I am no longer in the hospital but in a Residence home for old people. I am almost 88.
Tullio
Incredible. My utmost respect. Handling details of Boinc projects and CPU specifics. I've never heard of anyone of your generation being a geek. My father is 20 years younger and only knows web browsers and MS Word. I'm totally blown away. I wish you a long and healthy life and that your curiosity for new things never dries up.
Thanks to all for comments. Keith, I believe you made some helpful comments for me over a year ago. THANKS! Following your advice I started to try and stop usage of igpu to speed up other processing, but I think I screwed that up. Because my numbers did not look good, I removed all restrictions on Milky to try and get a new baseline. I will try and leave the new, do all, settings in place for another week or two. To give an example, the first month (Sep) I ran the G5-5 (Ryzen 5800 H, 8 core, 16 threads, at 3.2 GHz, maybe can be boosted to 4.4 GHz, but don't know, and Nvidia GeForce RTX 3050.) I generated 150,300 mean daily credits on that host and approx only 119,000 daily credits in January.) I will continue to let it ride for a while. The performance of my two Dell 8930s is even worse. I will post again after I get more data. By the way, because of processing time for Milky N Body Separation 146, running about 1 hr, 10 min proc time, do you people even do them? Anyone know how credits are generated for those long processing files?
Keith, love your rank of 6 on SETI! When they shut down my rank was 783 U.S. and 1,642 world, so you can see that I have to closely monitor performance of my small number of hosts.
JIm, it depends a lot on the project and the applications whether it is either worthwhile for processing on the cpu for the credit or you just value the science results from the cpu campaigns and deem it worthwhile.
At Milkyway the credit allotted for the N-body tasks is typically about a fifth of the credit assigned to the Separation gpu tasks for example. The gpu tasks compute in 1/70th of the time that the cpu tasks take and use much less energy for the computations.
I do believe for N-body that the tasks are using the CreditNew credit algorithm while the Separations tasks are using a fixed calculated credit assigned by the admins. The credit on N-body tasks is very variable as typical with CreditNew but the Separation tasks are assigned around ~226 +/- 1 more or less for every task.
I generally believe that if a project offers both cpu and gpu tasks, it is better to only run the gpu tasks as most efficient and generate the highest RAC.
Gpu only projects only have gpu devices running the tasks and cpu only projects only have cpu devices running the tasks. So best to mix and match several projects for best utilization of your resources.
I generated 150,300 mean daily credits on that host and approx only 119,000 daily credits in January.) I will continue to let it ride for a while. The performance of my two Dell 8930s is even worse.
Thanks again, folks. Will keep you informed.
Jim
One thing that can affect your RAC here at Einstein is the type of task you are running, ie the Meerkat tasks give out 3,333 credits per valid task, the GRP#1 tasks give out 3,465 credits per valid task, the O3 cpu tasks give out 1,000 credits per valid tasks and the O3 gpu tasks give out 1,000 credits per valid task. All that leads to if you got alot more O3 tasks in January then yes your RAC will drop accordingly.
I need the iGPU to run Einstein.I could not run GPUgrid since it needs CUDA.
I am no longer in the hospital but in a Residence home for old people. I am almost 88.
Tullio
Incredible. My utmost respect. Handling details of Boinc projects and CPU specifics. I've never heard of anyone of your generation being a geek. My father is 20 years younger and only knows web browsers and MS Word. I'm totally blown away. I wish you a long and healthy life and that your curiosity for new things never dries up.
What constitutes "your generation"?
I agree with Scrooge, at 88 Tullio is from another generation.
I need the iGPU to run Einstein.I could not run GPUgrid since it needs CUDA.
I am no longer in the hospital but in a Residence home for old people. I am almost 88.
Tullio
Incredible. My utmost respect. Handling details of Boinc projects and CPU specifics. I've never heard of anyone of your generation being a geek. My father is 20 years younger and only knows web browsers and MS Word. I'm totally blown away. I wish you a long and healthy life and that your curiosity for new things never dries up.
What constitutes "your generation"?
I agree with Scrooge, at 88 Tullio is from another generation.
That's what I'm asking, what number makes it "another generation" 85+ 80+ 75+ 70+ 65+ 60+ 55+ etc, I don't disagree that he is the ONLY 88 year old I know that can do with computers what he does!!
That's what I'm asking, what number makes it "another generation" 85+ 80+ 75+ 70+ 65+ 60+ 55+ etc, I don't disagree that he is the ONLY 88 year old I know that can do with computers what he does!!
Okay, I have to explain my astonishment a little bit... My comment was based on my experience of growing up behind the iron curtain. When I was 11 or 12, I first saw a computer (Robotron KC 87), an east german 8-bit micro computer, running an illegal replica of the Zilog Z80 CPU—because of embargo on western IT technology. A special school in my town had a few of these where I could learn programming BASIC one hour a week. After german reunification in 1990, money was tight for some years (economic collapse). At the age of 16 I got my first own PC (486DX2). Outside of universities, the few research centers, and large state-owned companies, my generation was the first to use computers widely. After 1990, fathers and mothers learned basic computer skills from their kids and first western home computers (8 bit Commodore's C64, C128, or Motorola 68K based AMIGAs (expensive)). They struggled when the first IBM compatible PCs replaced typewriters in their jobs in the 1990s. But the generation of today's 80+ grandmothers and grandfathers often only got tablets from their grandchildren 10 years ago. It's a fairly recent statement that residences for seniors are now also required to offer viable WiFi. But operating a powerful computer in a residence home for seniors to run BOINC projects—you've never heard of that.
That's what I'm asking, what number makes it "another generation" 85+ 80+ 75+ 70+ 65+ 60+ 55+ etc, I don't disagree that he is the ONLY 88 year old I know that can do with computers what he does!!
Okay, I have to explain my astonishment a little bit... My comment was based on my experience of growing up behind the iron curtain. When I was 11 or 12, I first saw a computer (Robotron KC 87), an east german 8-bit micro computer, running an illegal replica of the Zilog Z80 CPU—because of embargo on western IT technology. A special school in my town had a few of these where I could learn programming BASIC one hour a week. After german reunification in 1990, money was tight for some years (economic collapse). At the age of 16 I got my first own PC (486DX2). Outside of universities, the few research centers, and large state-owned companies, my generation was the first to use computers widely. After 1990, fathers and mothers learned basic computer skills from their kids and first western home computers (8 bit Commodore's C64, C128, or Motorola 68K based AMIGAs (expensive)). They struggled when the first IBM compatible PCs replaced typewriters in their jobs in the 1990s. But the generation of today's 80+ grandmothers and grandfathers often only got tablets from their grandchildren 10 years ago. It's a fairly recent statement that residences for seniors are now also required to offer viable WiFi. But operating a powerful computer in a residence home for seniors to run BOINC projects—you've never heard of that.
WOW that's a sad story that most people never hear, I'm sorry you had to go thru that when just across the wall things were vastly different!!
I grew up in a US Military without anything digital until I got to high school and got to play with a large computer that used paper tapes to keep your programs on and you had to both save the program on them and then reload the tape into the machine to play your game, that's what we were making, everytime you wanted to play it, that was in the late 60's. I made a game on playing pool and golf, both were very basic choose your angle and how hard to hit it type games. I got out of high school in '71 and joined the Navy and didn't touch another pc the whole 6 years I was in but once I got out I bought myself a Commodore 64 and then found myself having to write a program to help me decide which retirement program I should go with. They were offering to let us stay on the 20 year/50% of our pay or 30 years/75% of our pay or switch to a 'defined contribution' plan where they put a defined amount into an account every year, I chose the later based on being able to give any money that was left over when I died to my future family. I got a real pc when my dad bought me a brand new IBM PC with Windows 3 on it and I loved it and after that one thing led to another and here I am today retired with a 'computer room' in my garage full of desktops and other stuff crunching for various projects.
I was working in 1980 in SGS, a microelectronix company in Agrate near Milano. I was hired to provide UNIX user manuals on a Olivetti dot matrix printer using an Onyx UNIX computer designed by Scott McNealy. From then on my career followed the UNIX development, ending with a RISC machine in Trieste Area Research Park where I was manager of a Bull UNIX Laboratory. When retired, I bought a PC, installed Linux on it and joined BOIC.
tullio wrote: a am now
)
Hello Tullio,
I know you're from Italy and nearly 80 years old, but your message doesn't seem to have a question in it related to the topic. Are you just making a statement? Or did you need help with something related to the topic?
Proud member of the Old Farts Association
Scrooge McDuck wrote: tullio
)
What constitutes "your generation"?
Thanks to all for comments.
)
Thanks to all for comments. Keith, I believe you made some helpful comments for me over a year ago. THANKS! Following your advice I started to try and stop usage of igpu to speed up other processing, but I think I screwed that up. Because my numbers did not look good, I removed all restrictions on Milky to try and get a new baseline. I will try and leave the new, do all, settings in place for another week or two. To give an example, the first month (Sep) I ran the G5-5 (Ryzen 5800 H, 8 core, 16 threads, at 3.2 GHz, maybe can be boosted to 4.4 GHz, but don't know, and Nvidia GeForce RTX 3050.) I generated 150,300 mean daily credits on that host and approx only 119,000 daily credits in January.) I will continue to let it ride for a while. The performance of my two Dell 8930s is even worse. I will post again after I get more data. By the way, because of processing time for Milky N Body Separation 146, running about 1 hr, 10 min proc time, do you people even do them? Anyone know how credits are generated for those long processing files?
Keith, love your rank of 6 on SETI! When they shut down my rank was 783 U.S. and 1,642 world, so you can see that I have to closely monitor performance of my small number of hosts.
Thanks again, folks. Will keep you informed.
Jim
JIm, it depends a lot on the
)
JIm, it depends a lot on the project and the applications whether it is either worthwhile for processing on the cpu for the credit or you just value the science results from the cpu campaigns and deem it worthwhile.
At Milkyway the credit allotted for the N-body tasks is typically about a fifth of the credit assigned to the Separation gpu tasks for example. The gpu tasks compute in 1/70th of the time that the cpu tasks take and use much less energy for the computations.
I do believe for N-body that the tasks are using the CreditNew credit algorithm while the Separations tasks are using a fixed calculated credit assigned by the admins. The credit on N-body tasks is very variable as typical with CreditNew but the Separation tasks are assigned around ~226 +/- 1 more or less for every task.
I generally believe that if a project offers both cpu and gpu tasks, it is better to only run the gpu tasks as most efficient and generate the highest RAC.
Gpu only projects only have gpu devices running the tasks and cpu only projects only have cpu devices running the tasks. So best to mix and match several projects for best utilization of your resources.
James Bradshaw wrote: I
)
One thing that can affect your RAC here at Einstein is the type of task you are running, ie the Meerkat tasks give out 3,333 credits per valid task, the GRP#1 tasks give out 3,465 credits per valid task, the O3 cpu tasks give out 1,000 credits per valid tasks and the O3 gpu tasks give out 1,000 credits per valid task. All that leads to if you got alot more O3 tasks in January then yes your RAC will drop accordingly.
mikey wrote: Scrooge McDuck
)
I agree with Scrooge, at 88 Tullio is from another generation.
Proud member of the Old Farts Association
GWGeorge007 wrote: mikey
)
That's what I'm asking, what number makes it "another generation" 85+ 80+ 75+ 70+ 65+ 60+ 55+ etc, I don't disagree that he is the ONLY 88 year old I know that can do with computers what he does!!
mikey schrieb: That's what
)
Scrooge McDuck wrote: mikey
)
WOW that's a sad story that most people never hear, I'm sorry you had to go thru that when just across the wall things were vastly different!!
I grew up in a US Military without anything digital until I got to high school and got to play with a large computer that used paper tapes to keep your programs on and you had to both save the program on them and then reload the tape into the machine to play your game, that's what we were making, everytime you wanted to play it, that was in the late 60's. I made a game on playing pool and golf, both were very basic choose your angle and how hard to hit it type games. I got out of high school in '71 and joined the Navy and didn't touch another pc the whole 6 years I was in but once I got out I bought myself a Commodore 64 and then found myself having to write a program to help me decide which retirement program I should go with. They were offering to let us stay on the 20 year/50% of our pay or 30 years/75% of our pay or switch to a 'defined contribution' plan where they put a defined amount into an account every year, I chose the later based on being able to give any money that was left over when I died to my future family. I got a real pc when my dad bought me a brand new IBM PC with Windows 3 on it and I loved it and after that one thing led to another and here I am today retired with a 'computer room' in my garage full of desktops and other stuff crunching for various projects.
I was working in 1980 in SGS,
)
I was working in 1980 in SGS, a microelectronix company in Agrate near Milano. I was hired to provide UNIX user manuals on a Olivetti dot matrix printer using an Onyx UNIX computer designed by Scott McNealy. From then on my career followed the UNIX development, ending with a RISC machine in Trieste Area Research Park where I was manager of a Bull UNIX Laboratory. When retired, I bought a PC, installed Linux on it and joined BOIC.