The frequencies and sequence numbers are comparable enough, so RR Version 7F with a bit of fudging is useful. Making all the frequencies 1048.00 gives[pre]
-----------
App Version : 6.04
-----------
Frequency : 1048
Period of task cycle = 225.3
Task sequence number = 872
runtime = 28543
phase = 0.871
principal value = 0.395
Task sequence number = 873
runtime = 28556
phase = 0.875
principal value = 0.382
Task sequence number = 874
runtime = 28671
phase = 0.88
principal value = 0.369
Task sequence number = 875
runtime = 28920
phase = 0.884
principal value = 0.356
Task sequence number = 876
runtime = 29238
phase = 0.888
principal value = 0.343
Task sequence number = 877
runtime = 29019
phase = 0.893
principal value = 0.33
Task sequence number = 878
runtime = 28719
phase = 0.897
principal value = 0.317
Task sequence number = 879
runtime = 29182
phase = 0.902
principal value = 0.304
Number of points = 8
Minimum runtime in data = 28542.67
Maximum runtime in data = 29238.28
Estimated peak runtime = 30973
Estimated average runtime = 27119
Estimated trough runtime = 24919
Estimated runtime variance = 0.195
-----------
App Version : 6.05
-----------
Frequency : 1048
Period of task cycle = 225.3
Task sequence number = 870
runtime = 23410
phase = 0.862
principal value = 0.421
Task sequence number = 871
runtime = 23243
phase = 0.866
principal value = 0.408
Task sequence number = 894
runtime = 24986
phase = 0.968
principal value = 0.099
Task sequence number = 895
runtime = 24423
phase = 0.973
principal value = 0.085
Number of points = 4
Minimum runtime in data = 23242.78
Maximum runtime in data = 24986.44
Estimated peak runtime = 25086
Estimated average runtime = 22395
Estimated trough runtime = 20860
Estimated runtime variance = 0.168
[/pre]So ( using average runtimes ) 22395/27119 ~ 83%
Yup! :-)
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
Thanks for running those numbers through your Ver 7F. That host has five or six 1048 86x in the pipe line plus 1048_925. Assume a hole needs filling. Don't know why the 1048.05 appeared, I only have those two at that freq for now.
As I have not seen any negative reports on 6.05 I am assuming that it will go mainstream soon as the default SSE2 app for windows.
If this is so then the overall cr/time will increase again. And as Project granted credit comparison has reported consistently that Einstein v Seti, for the last few weeks, is at 14 - 15% more for Einstein, I assume that a further cut is Einstein credits cannot be ruled out.
As I have not seen any negative reports on 6.05 I am assuming that it will go mainstream soon as the default SSE2 app for windows.
If this is so then the overall cr/time will increase again. And as Project granted credit comparison has reported consistently that Einstein v Seti, for the last few weeks, is at 14 - 15% more for Einstein, I assume that a further cut is Einstein credits cannot be ruled out.
I just saw a breaking news headline where the new Federal Credit Czar was informed by the credit cops (who have been monitoring the Project Granted Credit Comparison Report diligently) that Einstein@home has been giving away an over abundance of credits lately. (aprox. 15%) So the Feds have decided to not only take that 15%, but to penalize each cruncher an additional 15%, for a total of 30% of the present credit value to help offset the world wide credit shortage in hopes that it will stabilize the global credit markets.
I just saw a breaking news headline where the new Federal Credit Czar was informed by the credit cops (who have been monitoring the Project Granted Credit Comparison Report diligently) that Einstein@home has been giving away an over abundance of credits lately. (aprox. 15%) So the Feds have decided to not only take that 15%, but to penalize each cruncher an additional 15%, for a total of 30% of the present credit value to help offset the world wide credit shortage in hopes that it will stabilize the global credit markets.
Oh. Joy .... :-)
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
Has anyone else seen an increase in temps for their CPUs with 6.05? Mine seem to be running about 3-4 C higher.
It's the SSE addition. I was seeing this already on the normal SSE application. If only non-optimized apps were running (like two Primegrids), my temperature would be below 60C, but as soon as Einstein or Seti started running, the temp would go up by 4-5C.
Has anyone else seen an increase in temps for their CPUs with 6.05? Mine seem to be running about 3-4 C higher.
Normal under full load is usually 52-53 C and with the new 6.05 app its running 55-56 C
Just curious.
Actually I see temp variations even between different WUs running the same app. I did not try to correlate that to the runtime variations, but i would not be surprised if there is connection.
The frequencies and sequence
)
The frequencies and sequence numbers are comparable enough, so RR Version 7F with a bit of fudging is useful. Making all the frequencies 1048.00 gives[pre]
-----------
App Version : 6.04
-----------
Frequency : 1048
Period of task cycle = 225.3
Task sequence number = 872
runtime = 28543
phase = 0.871
principal value = 0.395
Task sequence number = 873
runtime = 28556
phase = 0.875
principal value = 0.382
Task sequence number = 874
runtime = 28671
phase = 0.88
principal value = 0.369
Task sequence number = 875
runtime = 28920
phase = 0.884
principal value = 0.356
Task sequence number = 876
runtime = 29238
phase = 0.888
principal value = 0.343
Task sequence number = 877
runtime = 29019
phase = 0.893
principal value = 0.33
Task sequence number = 878
runtime = 28719
phase = 0.897
principal value = 0.317
Task sequence number = 879
runtime = 29182
phase = 0.902
principal value = 0.304
Number of points = 8
Minimum runtime in data = 28542.67
Maximum runtime in data = 29238.28
Estimated peak runtime = 30973
Estimated average runtime = 27119
Estimated trough runtime = 24919
Estimated runtime variance = 0.195
-----------
App Version : 6.05
-----------
Frequency : 1048
Period of task cycle = 225.3
Task sequence number = 870
runtime = 23410
phase = 0.862
principal value = 0.421
Task sequence number = 871
runtime = 23243
phase = 0.866
principal value = 0.408
Task sequence number = 894
runtime = 24986
phase = 0.968
principal value = 0.099
Task sequence number = 895
runtime = 24423
phase = 0.973
principal value = 0.085
Number of points = 4
Minimum runtime in data = 23242.78
Maximum runtime in data = 24986.44
Estimated peak runtime = 25086
Estimated average runtime = 22395
Estimated trough runtime = 20860
Estimated runtime variance = 0.168
[/pre]So ( using average runtimes ) 22395/27119 ~ 83%
Yup! :-)
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
Thanks for running those
)
Thanks for running those numbers through your Ver 7F. That host has five or six 1048 86x in the pipe line plus 1048_925. Assume a hole needs filling. Don't know why the 1048.05 appeared, I only have those two at that freq for now.
Andy
As I have not seen any
)
As I have not seen any negative reports on 6.05 I am assuming that it will go mainstream soon as the default SSE2 app for windows.
If this is so then the overall cr/time will increase again. And as Project granted credit comparison has reported consistently that Einstein v Seti, for the last few weeks, is at 14 - 15% more for Einstein, I assume that a further cut is Einstein credits cannot be ruled out.
RE: As I have not seen any
)
I just saw a breaking news headline where the new Federal Credit Czar was informed by the credit cops (who have been monitoring the Project Granted Credit Comparison Report diligently) that Einstein@home has been giving away an over abundance of credits lately. (aprox. 15%) So the Feds have decided to not only take that 15%, but to penalize each cruncher an additional 15%, for a total of 30% of the present credit value to help offset the world wide credit shortage in hopes that it will stabilize the global credit markets.
RE: I just saw a breaking
)
Oh. Joy .... :-)
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
RE: Oh. Joy ....
)
I just couldn't resist it.
But seriously it won't be much longer when 1,800,000 credits and $5.00 will buy me a cup of coffee. ;-)
Back on topic..... Has
)
Back on topic.....
Has anyone else seen an increase in temps for their CPUs with 6.05? Mine seem to be running about 3-4 C higher.
Normal under full load is usually 52-53 C and with the new 6.05 app its running 55-56 C
Just curious.
RE: Has anyone else seen an
)
It's the SSE addition. I was seeing this already on the normal SSE application. If only non-optimized apps were running (like two Primegrids), my temperature would be below 60C, but as soon as Einstein or Seti started running, the temp would go up by 4-5C.
RE: Back on topic..... Has
)
Actually I see temp variations even between different WUs running the same app. I did not try to correlate that to the runtime variations, but i would not be surprised if there is connection.
CU
Bikeman
I have a power monitor on my
)
I have a power monitor on my Quad. It looked to me as though there was no significant change in switching from 6.04 to 6.05.
Of course this answer can vary by CPU architecture. And I did not take proper time-averaged data.