AWP ODC GPU - Revision history

Maechlin at 18:40, 13 February 2014

2014-02-13T18:40:16Z

Maechlin at 18:01, 20 November 2012

2012-11-20T18:01:12Z

Maechlin at 18:00, 20 November 2012

2012-11-20T18:00:34Z

Maechlin at 01:58, 3 November 2012

2012-11-03T01:58:12Z

Maechlin at 15:46, 25 October 2012

2012-10-25T15:46:00Z

Maechlin: Created page with 'Efecan is a research intern working with me on SCEC projects. Thanks for the support to him as well as to our first major GPU simulation in preparation. What Efecan is trying to …'

2012-10-25T15:44:49Z

Created page with 'Efecan is a research intern working with me on SCEC projects. Thanks for the support to him as well as to our first major GPU simulation in preparation. What Efecan is trying to …'

New page

Efecan is a research intern working with me on SCEC projects. Thanks for the support to him as well as to our first major GPU simulation in preparation. What Efecan is trying to run is a full scale Chino Hills validation ground motion simulation, at size of 3500x2500x1500 problem size, 16m resolution, we used 250 GPUs on Keeneland, the initial test went well. We are seeing 55x speedup comparing to the CPU code for the same setting, we know Keeneland may have some scaling issue when using high number of nodes, plus this simulation has serious I/O - around 0.5TB, requiring frequent data transfer from GPU to CPU through PCIe. However, we are pretty happy with the performance getting so far. One issue with current GPU code is that we don't have validation prints about velocity max/min value, for that the simulation results couldn't be validated at run time. We are working with our visualization expert to analyze the initial results, and also work with seismologist Kim Olsen. Once everything is set, we'll make reservation next week or so for a full size run which possibly take 5 hours to complete.

← Older revision		Revision as of 18:40, 13 February 2014
Line 1:		Line 1:
−	~~Efecan Poyraz, a research intern working with~~ Yifeng Cui, Kim Olsen, and others on SCEC projects on SCEC's first major GPU ~~simulation in preparation~~. ~~Efecan~~ is ~~trying to run~~ a full scale Chino Hills validation ground motion simulation, at size of 3500x2500x1500 problem size, 16m resolution, we used 250 GPUs on Keeneland, the initial test went well. We are seeing 55x speedup comparing to the CPU code for the same setting, we know Keeneland may have some scaling issue when using high number of nodes, plus this simulation has serious I/O - around 0.5TB, requiring frequent data transfer from GPU to CPU through PCIe. However, we are pretty happy with the performance getting so far. One issue with current GPU code is that we don't have validation prints about velocity max/min value, for that the simulation results couldn't be validated at run time. We are working with our visualization expert to analyze the initial results, and also work with seismologist Kim Olsen. Once everything is set, we'll make reservation next week or so for a full size run which possibly take 5 hours to complete.	+	Yifeng Cui, Kim Olsen, Efecan Poyraz, and others on SCEC projects have developed SCEC's first major GPU software. This group is running a full scale Chino Hills validation ground motion simulation, at size of 3500x2500x1500 problem size, 16m resolution, we used 250 GPUs on Keeneland, the initial test went well. We are seeing 55x speedup comparing to the CPU code for the same setting, we know Keeneland may have some scaling issue when using high number of nodes, plus this simulation has serious I/O - around 0.5TB, requiring frequent data transfer from GPU to CPU through PCIe. However, we are pretty happy with the performance getting so far. One issue with current GPU code is that we don't have validation prints about velocity max/min value, for that the simulation results couldn't be validated at run time. We are working with our visualization expert to analyze the initial results, and also work with seismologist Kim Olsen. Once everything is set, we'll make reservation next week or so for a full size run which possibly take 5 hours to complete.
	[[Image:Screen Shot 2012-11-20 at 9.57.51 AM.png\|right\|150px]]		[[Image:Screen Shot 2012-11-20 at 9.57.51 AM.png\|right\|150px]]

← Older revision		Revision as of 18:01, 20 November 2012
Line 1:		Line 1:
	Efecan Poyraz, a research intern working with Yifeng Cui, Kim Olsen, and others on SCEC projects on SCEC's first major GPU simulation in preparation. Efecan is trying to run a full scale Chino Hills validation ground motion simulation, at size of 3500x2500x1500 problem size, 16m resolution, we used 250 GPUs on Keeneland, the initial test went well. We are seeing 55x speedup comparing to the CPU code for the same setting, we know Keeneland may have some scaling issue when using high number of nodes, plus this simulation has serious I/O - around 0.5TB, requiring frequent data transfer from GPU to CPU through PCIe. However, we are pretty happy with the performance getting so far. One issue with current GPU code is that we don't have validation prints about velocity max/min value, for that the simulation results couldn't be validated at run time. We are working with our visualization expert to analyze the initial results, and also work with seismologist Kim Olsen. Once everything is set, we'll make reservation next week or so for a full size run which possibly take 5 hours to complete.		Efecan Poyraz, a research intern working with Yifeng Cui, Kim Olsen, and others on SCEC projects on SCEC's first major GPU simulation in preparation. Efecan is trying to run a full scale Chino Hills validation ground motion simulation, at size of 3500x2500x1500 problem size, 16m resolution, we used 250 GPUs on Keeneland, the initial test went well. We are seeing 55x speedup comparing to the CPU code for the same setting, we know Keeneland may have some scaling issue when using high number of nodes, plus this simulation has serious I/O - around 0.5TB, requiring frequent data transfer from GPU to CPU through PCIe. However, we are pretty happy with the performance getting so far. One issue with current GPU code is that we don't have validation prints about velocity max/min value, for that the simulation results couldn't be validated at run time. We are working with our visualization expert to analyze the initial results, and also work with seismologist Kim Olsen. Once everything is set, we'll make reservation next week or so for a full size run which possibly take 5 hours to complete.
−	[[Image:~~usgs~~-~~logo~~-~~color~~.~~jpg~~\|right\|150px]]	+	[[Image:Screen Shot 2012-11-20 at 9.57.51 AM.png\|right\|150px]]

← Older revision		Revision as of 18:00, 20 November 2012
Line 1:		Line 1:
	Efecan Poyraz, a research intern working with Yifeng Cui, Kim Olsen, and others on SCEC projects on SCEC's first major GPU simulation in preparation. Efecan is trying to run a full scale Chino Hills validation ground motion simulation, at size of 3500x2500x1500 problem size, 16m resolution, we used 250 GPUs on Keeneland, the initial test went well. We are seeing 55x speedup comparing to the CPU code for the same setting, we know Keeneland may have some scaling issue when using high number of nodes, plus this simulation has serious I/O - around 0.5TB, requiring frequent data transfer from GPU to CPU through PCIe. However, we are pretty happy with the performance getting so far. One issue with current GPU code is that we don't have validation prints about velocity max/min value, for that the simulation results couldn't be validated at run time. We are working with our visualization expert to analyze the initial results, and also work with seismologist Kim Olsen. Once everything is set, we'll make reservation next week or so for a full size run which possibly take 5 hours to complete.		Efecan Poyraz, a research intern working with Yifeng Cui, Kim Olsen, and others on SCEC projects on SCEC's first major GPU simulation in preparation. Efecan is trying to run a full scale Chino Hills validation ground motion simulation, at size of 3500x2500x1500 problem size, 16m resolution, we used 250 GPUs on Keeneland, the initial test went well. We are seeing 55x speedup comparing to the CPU code for the same setting, we know Keeneland may have some scaling issue when using high number of nodes, plus this simulation has serious I/O - around 0.5TB, requiring frequent data transfer from GPU to CPU through PCIe. However, we are pretty happy with the performance getting so far. One issue with current GPU code is that we don't have validation prints about velocity max/min value, for that the simulation results couldn't be validated at run time. We are working with our visualization expert to analyze the initial results, and also work with seismologist Kim Olsen. Once everything is set, we'll make reservation next week or so for a full size run which possibly take 5 hours to complete.
		+	[[Image:usgs-logo-color.jpg\|right\|150px]]

← Older revision		Revision as of 01:58, 3 November 2012
Line 1:		Line 1:
−	Efecan Poyraz is a research intern working with Yifeng Cui, Kim Olsen, and others on SCEC projects on SCEC's first major GPU simulation in preparation. Efecan is trying to run a full scale Chino Hills validation ground motion simulation, at size of 3500x2500x1500 problem size, 16m resolution, we used 250 GPUs on Keeneland, the initial test went well. We are seeing 55x speedup comparing to the CPU code for the same setting, we know Keeneland may have some scaling issue when using high number of nodes, plus this simulation has serious I/O - around 0.5TB, requiring frequent data transfer from GPU to CPU through PCIe. However, we are pretty happy with the performance getting so far. One issue with current GPU code is that we don't have validation prints about velocity max/min value, for that the simulation results couldn't be validated at run time. We are working with our visualization expert to analyze the initial results, and also work with seismologist Kim Olsen. Once everything is set, we'll make reservation next week or so for a full size run which possibly take 5 hours to complete.	+	Efecan Poyraz, a research intern working with Yifeng Cui, Kim Olsen, and others on SCEC projects on SCEC's first major GPU simulation in preparation. Efecan is trying to run a full scale Chino Hills validation ground motion simulation, at size of 3500x2500x1500 problem size, 16m resolution, we used 250 GPUs on Keeneland, the initial test went well. We are seeing 55x speedup comparing to the CPU code for the same setting, we know Keeneland may have some scaling issue when using high number of nodes, plus this simulation has serious I/O - around 0.5TB, requiring frequent data transfer from GPU to CPU through PCIe. However, we are pretty happy with the performance getting so far. One issue with current GPU code is that we don't have validation prints about velocity max/min value, for that the simulation results couldn't be validated at run time. We are working with our visualization expert to analyze the initial results, and also work with seismologist Kim Olsen. Once everything is set, we'll make reservation next week or so for a full size run which possibly take 5 hours to complete.

← Older revision		Revision as of 15:46, 25 October 2012
Line 1:		Line 1:
−	Efecan is a research intern working with me on SCEC projects~~. Thanks for the support to him as well as to our~~ first major GPU simulation in preparation. ~~What~~ Efecan is trying to run is a full scale Chino Hills validation ground motion simulation, at size of 3500x2500x1500 problem size, 16m resolution, we used 250 GPUs on Keeneland, the initial test went well. We are seeing 55x speedup comparing to the CPU code for the same setting, we know Keeneland may have some scaling issue when using high number of nodes, plus this simulation has serious I/O - around 0.5TB, requiring frequent data transfer from GPU to CPU through PCIe. However, we are pretty happy with the performance getting so far. One issue with current GPU code is that we don't have validation prints about velocity max/min value, for that the simulation results couldn't be validated at run time. We are working with our visualization expert to analyze the initial results, and also work with seismologist Kim Olsen. Once everything is set, we'll make reservation next week or so for a full size run which possibly take 5 hours to complete.	+	Efecan Poyraz is a research intern working with Yifeng Cui, Kim Olsen, and others on SCEC projects on SCEC's first major GPU simulation in preparation. Efecan is trying to run a full scale Chino Hills validation ground motion simulation, at size of 3500x2500x1500 problem size, 16m resolution, we used 250 GPUs on Keeneland, the initial test went well. We are seeing 55x speedup comparing to the CPU code for the same setting, we know Keeneland may have some scaling issue when using high number of nodes, plus this simulation has serious I/O - around 0.5TB, requiring frequent data transfer from GPU to CPU through PCIe. However, we are pretty happy with the performance getting so far. One issue with current GPU code is that we don't have validation prints about velocity max/min value, for that the simulation results couldn't be validated at run time. We are working with our visualization expert to analyze the initial results, and also work with seismologist Kim Olsen. Once everything is set, we'll make reservation next week or so for a full size run which possibly take 5 hours to complete.