Well, from what little data I have, I think I can show 4.26 and 4.27 are faster than 4.15 and 4.24.
Also, the middle two charts (side by side) are the Win vs Lin comparisons for nearly the same freq, and it shows linux to be faster on my dual boot machine.
I put the two charts together for my AMD64 X2 5200 (both win and lin), and IF comparing "nearly identical" freqs is kosher, Then both 4.24 and 4.27 are faster than 4.15 or 4.26.
I put the two charts together for my AMD64 X2 5200 (both win and lin), and IF comparing "nearly identical" freqs is kosher, Then both 4.24 or 4.27 are faster than 4.15 or 4.26.
Yes, I think those frequencies are close enough for that to be a valid conclusion.
I put the two charts together for my AMD64 X2 5200 (both win and lin), and IF comparing "nearly identical" freqs is kosher, Then both 4.24 and 4.27 are faster than 4.15 or 4.26.
Well, thanks! That seems to be going towards confirming my "gut feeling" that the Linux app already had an advantage over the Windows app on the same hardware. I also felt that 4.26 would bring near-parity with the 4.20 code base (non-SSE, non-linear trig), which has also been shown, as 4.24 was not deemed much faster, if not a little bit slower, than 4.20 due to compiler issues.
Like I said elsewhere, even I will concede that I'm "wrong" most of the time, but every now and then I'll sneak a winner through...
I reckon near-parity among apps running a different OS on the same hardware in this time of turmoil with cross-project parity is probably a good idea... ;-)
Heck, Brian, just move on over to the "dark side".
Free yourself of the binds and worries of windows.
Can I get a "heck Yeahhh"??? (where's Martin ML1 when you need em??)
"Bear witness Brother".
(P.S...I kneeeeeew you'd like the post and thought of you as I was preparing it. LOL)
P.P.S You do know there are compiler differences between compilers used ...Right????. Also, differences in how they're used. That could account for it.
Heck, Brian, just move on over to the "dark side".
Free yourself of the binds and worries of windows.
I already tried that with a VM. You can see the machinations in this thread.
I entered that with the assumption that there would be somewhere around a 20% delta between the Linux application and the Windows application on the same hardware, but I had only openly stated on the boards that a 10% speedup was "major". I flat out could not prove it because the VM overhead was so severe. Your chart, while preliminary, seems to show 15-20%, which is what I originally believed.
Quote:
Can I get a "heck Yeahhh"??? (where's Martin ML1 when you need em??)
Yeah, Martin would have something to add, I'm sure... Anyway, my viewpoint on Linux is that it is too difficult for "average" users. Also as far as more advanced users, I don't understand why installation of software has to be such a bear. When I want to install something, I want to install it, not go make sure that all my libs are the right version then make sure I hold my left foot 2 inches above the floor while rubbing my belly during the install process... [/sarcasm and frustration]
Quote:
P.P.S You do know there are compiler differences between compilers used ...Right????. Also, differences in how they're used. That could account for it.
Yes, but a 20% delta on the same hardware just because of a different OS/compilation is quite extreme... If it's the best that can be done, I suppose... :shrug:
Anyway, my viewpoint on Linux is that it is too difficult for "average" users. Also as far as more advanced users, I don't understand why installation of software has to be such a bear. When I want to install something, I want to install it, not go make sure that all my libs are the right version then make sure I hold my left foot 2 inches above the floor while rubbing my belly during the install process...
This is going off topic but i just couldnt disagree more :)
Linux too difficult for average users, then what about viruses and security vulnerabilities and spyware, its MUCH harder to keep Windows fresh for a less advanced user, Debian and Debian based Linux desktop distros just work out of the box and keep on working, hows that for the average user. You have to be an expert to keep Windows in top shape, most of the procedures involved in cleaning up is just too advanced for a regular user. I feel "Linux is so difficult" is what the über-geeks WANT the rest of us to think, if normal users think Linux is too hard, then the more credit to the geeks.
Also i dont understand your argument about libraries in Linux, 99% or more of the common/handy Linux apps are available as Debian packages and you DID use Ubuntu on your VMWare project? Thats alot easier than any install method in Windows IMHO, only spyware/viruses from suspicious pornsites installs that efficiently on Windowns :D
A small story, my older bro is a very un-advanced computer user and 5-6 months ago changed to Linux (Kubuntu 7.04) on his Centrino laptop. For him that was MUCH easier than Windows ever was, and maintenance free in addition. As a bonus his wireless network had never been that stable before, and the battery he thought was already worn out became good as new, in addition to the very significant speedup for day to day tasks.
You have to be an expert to keep Windows in top shape, most of the procedures involved in cleaning up is just too advanced for a regular user.
I'm reminded of the illustrious instructor that I had for the Linux class, which I dropped due to his attitude... He insisted that Windows required frequent reinstallations to keep it cleaned up and "in top shape". This installation was performed in Nov/Dec 2004 on a computer that was assembled by myself from components that I bought from Newegg. I have never reinstalled and I don't think I need to just for the heck of it either. Also, running the built-in defragmenter is definitely not "too advanced". You analyze the disk, then you are advised whether it feels you should defragment or not, and so then you hit defragment. What is difficult about that?
The truth is, as I point out but nobody listens to (again, one of these infrequent "winners" that I feel I sneak through), the problem with slowdowns in Windows is because of 3rd party programmers, i.e. non-Microsoft programmers, adding in numerous events that launch when the system boots and stay resident in memory so as to continue to "advertise" their product. For example, and you can feel free to shudder, AOL. AOL puts in a "connectivity service" (ACSD) that is set to automatic startup along with an entry in the "run" registry key to load AOLSOFTWARE when the system starts. This gives you your wonderful System Tray icon so that AOL functionality is "right at the user's fingertips", as it were. All this though takes up memory, both main memory and stack space. You get a ton of those nice systray things going on, it is no wonder your system becomes "slow"...
The point: Those things are not directly Microsoft's fault...
Quote:
Also i dont understand your argument about libraries in Linux, 99% or more of the common/handy Linux apps are available as Debian packages and you DID use Ubuntu on your VMWare project? Thats alot easier than any install method in Windows IMHO, only spyware/viruses from suspicious pornsites installs that efficiently on Windowns :D
I had installed some package installer after the base installation that did make it easier, but my Workstation 6.0.2 license has expired and when I just tried to bring up the VM in VM Player, the mouse was moving but I could not click on anything...so dunno what that was about... I'll probably need to uninstall Workstation. I'll try to figure out what it was that I installed and get back to you... If it helps identify it, packages this way have a blue dashed circle arrow around the installation package...
Also, what I was talking about were installations of items that were not in apt or Synaptic...and I was not using (could not use?) RPM... I'll freely admit there's some n00bishness there on my part, that and the illustrious instructor was hard-core command line...
RE: So when did we change
)
AFAIK E@H was always looking for Continuous Waves (from pulsars).
CU
Bikeman
RE: AFAIK E@H was always
)
Well yes, but I'd thought we had inspiral templates too ... ah well, live & learn :-)
Cheers Mike.
I have made this letter longer than usual because I lack the time to make it shorter ...
... and my other CPU is a Ryzen 5950X :-) Blaise Pascal
Well, from what little data I
)
Well, from what little data I have, I think I can show 4.26 and 4.27 are faster than 4.15 and 4.24.
Also, the middle two charts (side by side) are the Win vs Lin comparisons for nearly the same freq, and it shows linux to be faster on my dual boot machine.
Am I reading these right?
tony
I put the two charts together
)
I put the two charts together for my AMD64 X2 5200 (both win and lin), and IF comparing "nearly identical" freqs is kosher, Then both 4.24 and 4.27 are faster than 4.15 or 4.26.
RE: I put the two charts
)
Yes, I think those frequencies are close enough for that to be a valid conclusion.
RE: I put the two charts
)
Well, thanks! That seems to be going towards confirming my "gut feeling" that the Linux app already had an advantage over the Windows app on the same hardware. I also felt that 4.26 would bring near-parity with the 4.20 code base (non-SSE, non-linear trig), which has also been shown, as 4.24 was not deemed much faster, if not a little bit slower, than 4.20 due to compiler issues.
Like I said elsewhere, even I will concede that I'm "wrong" most of the time, but every now and then I'll sneak a winner through...
I reckon near-parity among apps running a different OS on the same hardware in this time of turmoil with cross-project parity is probably a good idea... ;-)
Heck, Brian, just move on
)
Heck, Brian, just move on over to the "dark side".
Free yourself of the binds and worries of windows.
Can I get a "heck Yeahhh"??? (where's Martin ML1 when you need em??)
"Bear witness Brother".
(P.S...I kneeeeeew you'd like the post and thought of you as I was preparing it. LOL)
P.P.S You do know there are compiler differences between compilers used ...Right????. Also, differences in how they're used. That could account for it.
RE: Heck, Brian, just move
)
I already tried that with a VM. You can see the machinations in this thread.
I entered that with the assumption that there would be somewhere around a 20% delta between the Linux application and the Windows application on the same hardware, but I had only openly stated on the boards that a 10% speedup was "major". I flat out could not prove it because the VM overhead was so severe. Your chart, while preliminary, seems to show 15-20%, which is what I originally believed.
Yeah, Martin would have something to add, I'm sure... Anyway, my viewpoint on Linux is that it is too difficult for "average" users. Also as far as more advanced users, I don't understand why installation of software has to be such a bear. When I want to install something, I want to install it, not go make sure that all my libs are the right version then make sure I hold my left foot 2 inches above the floor while rubbing my belly during the install process... [/sarcasm and frustration]
Yes, but a 20% delta on the same hardware just because of a different OS/compilation is quite extreme... If it's the best that can be done, I suppose... :shrug:
RE: Anyway, my viewpoint on
)
This is going off topic but i just couldnt disagree more :)
Linux too difficult for average users, then what about viruses and security vulnerabilities and spyware, its MUCH harder to keep Windows fresh for a less advanced user, Debian and Debian based Linux desktop distros just work out of the box and keep on working, hows that for the average user. You have to be an expert to keep Windows in top shape, most of the procedures involved in cleaning up is just too advanced for a regular user. I feel "Linux is so difficult" is what the über-geeks WANT the rest of us to think, if normal users think Linux is too hard, then the more credit to the geeks.
Also i dont understand your argument about libraries in Linux, 99% or more of the common/handy Linux apps are available as Debian packages and you DID use Ubuntu on your VMWare project? Thats alot easier than any install method in Windows IMHO, only spyware/viruses from suspicious pornsites installs that efficiently on Windowns :D
A small story, my older bro is a very un-advanced computer user and 5-6 months ago changed to Linux (Kubuntu 7.04) on his Centrino laptop. For him that was MUCH easier than Windows ever was, and maintenance free in addition. As a bonus his wireless network had never been that stable before, and the battery he thought was already worn out became good as new, in addition to the very significant speedup for day to day tasks.
Team Philippines
RE: You have to be an
)
I'm reminded of the illustrious instructor that I had for the Linux class, which I dropped due to his attitude... He insisted that Windows required frequent reinstallations to keep it cleaned up and "in top shape". This installation was performed in Nov/Dec 2004 on a computer that was assembled by myself from components that I bought from Newegg. I have never reinstalled and I don't think I need to just for the heck of it either. Also, running the built-in defragmenter is definitely not "too advanced". You analyze the disk, then you are advised whether it feels you should defragment or not, and so then you hit defragment. What is difficult about that?
The truth is, as I point out but nobody listens to (again, one of these infrequent "winners" that I feel I sneak through), the problem with slowdowns in Windows is because of 3rd party programmers, i.e. non-Microsoft programmers, adding in numerous events that launch when the system boots and stay resident in memory so as to continue to "advertise" their product. For example, and you can feel free to shudder, AOL. AOL puts in a "connectivity service" (ACSD) that is set to automatic startup along with an entry in the "run" registry key to load AOLSOFTWARE when the system starts. This gives you your wonderful System Tray icon so that AOL functionality is "right at the user's fingertips", as it were. All this though takes up memory, both main memory and stack space. You get a ton of those nice systray things going on, it is no wonder your system becomes "slow"...
The point: Those things are not directly Microsoft's fault...
I had installed some package installer after the base installation that did make it easier, but my Workstation 6.0.2 license has expired and when I just tried to bring up the VM in VM Player, the mouse was moving but I could not click on anything...so dunno what that was about... I'll probably need to uninstall Workstation. I'll try to figure out what it was that I installed and get back to you... If it helps identify it, packages this way have a blue dashed circle arrow around the installation package...
Also, what I was talking about were installations of items that were not in apt or Synaptic...and I was not using (could not use?) RPM... I'll freely admit there's some n00bishness there on my part, that and the illustrious instructor was hard-core command line...