Hi Ron, I would like to see a research component in ESC as well. I don't have a percentage in mind but getting to Exascale won't be evolutionary (or if it is, we don't need an initiative). Regards, Sudip On 3/22/11 10:48 AM, "ronald minnich" <[email protected]> wrote:
My biggest single concern about what I've seen on the ESC roadmap, and I really very reluctant to say this because of the hard work done to date, is that it seems very evolutionary, and I just can't tell how that's going to play out on the new big machines. Is there room for revolution in the ESC? If so, what fraction of the spend plan? I think it's a problem if it's under 60%
But am I being unfair? If so, just tell me!
ron _______________________________________________ Esc-team mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/esc-team
Sudip: We are not saying research is not needed, we are saying that ESC is not the place for it. Bronis On Tue, 22 Mar 2011, Dosanjh, Sudip S wrote:
Hi Ron,
I would like to see a research component in ESC as well. I don't have a percentage in mind but getting to Exascale won't be evolutionary (or if it is, we don't need an initiative).
Regards,
Sudip
On 3/22/11 10:48 AM, "ronald minnich" <[email protected]> wrote:
My biggest single concern about what I've seen on the ESC roadmap, and I really very reluctant to say this because of the hard work done to date, is that it seems very evolutionary, and I just can't tell how that's going to play out on the new big machines. Is there room for revolution in the ESC? If so, what fraction of the spend plan? I think it's a problem if it's under 60%
But am I being unfair? If so, just tell me!
ron _______________________________________________ Esc-team mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/esc-team
_______________________________________________ Esc-team mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/esc-team
Bronis: Not sure why that should be the case. ESC actually started more as a research effort. At a minimum, I don't think you can plan a deployment activity for Exascale without having a connection to the research community. If ESC has no opportunity to influence Exascale software research, I am not sure how it will succeed. Sudip On 3/22/11 11:41 AM, "Bronis R. de Supinski" <[email protected]> wrote:
Sudip:
We are not saying research is not needed, we are saying that ESC is not the place for it.
Bronis
On Tue, 22 Mar 2011, Dosanjh, Sudip S wrote:
Hi Ron,
I would like to see a research component in ESC as well. I don't have a percentage in mind but getting to Exascale won't be evolutionary (or if it is, we don't need an initiative).
Regards,
Sudip
On 3/22/11 10:48 AM, "ronald minnich" <[email protected]> wrote:
My biggest single concern about what I've seen on the ESC roadmap, and I really very reluctant to say this because of the hard work done to date, is that it seems very evolutionary, and I just can't tell how that's going to play out on the new big machines. Is there room for revolution in the ESC? If so, what fraction of the spend plan? I think it's a problem if it's under 60%
But am I being unfair? If so, just tell me!
ron _______________________________________________ Esc-team mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/esc-team
_______________________________________________ Esc-team mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/esc-team
Sudip: Re:
Not sure why that should be the case. ESC actually started more as a research effort. At a minimum, I don't think you can plan a deployment activity for Exascale without having a connection to the research community. If ESC has no opportunity to influence Exascale software research, I am not sure how it will succeed.
Perhaps we are merely using different words for the same thing. ESC needs to influence exascale software research. ESC needs to monitor exascale software research. ESC needs to adopt and to productize successful exascale software research. ESC does not need to perform exascale software research. Doing so would expand the scope of ESC significantly and require it to have much more funding than the already large amount being discussed. Bronis
Sudip
On 3/22/11 11:41 AM, "Bronis R. de Supinski" <[email protected]> wrote:
Sudip:
We are not saying research is not needed, we are saying that ESC is not the place for it.
Bronis
On Tue, 22 Mar 2011, Dosanjh, Sudip S wrote:
Hi Ron,
I would like to see a research component in ESC as well. I don't have a percentage in mind but getting to Exascale won't be evolutionary (or if it is, we don't need an initiative).
Regards,
Sudip
On 3/22/11 10:48 AM, "ronald minnich" <[email protected]> wrote:
My biggest single concern about what I've seen on the ESC roadmap, and I really very reluctant to say this because of the hard work done to date, is that it seems very evolutionary, and I just can't tell how that's going to play out on the new big machines. Is there room for revolution in the ESC? If so, what fraction of the spend plan? I think it's a problem if it's under 60%
But am I being unfair? If so, just tell me!
ron _______________________________________________ Esc-team mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/esc-team
_______________________________________________ Esc-team mailing list [email protected] https://lists.mcs.anl.gov/mailman/listinfo/esc-team
participants (2)
-
Bronis R. de Supinski -
Dosanjh, Sudip S