Analysis assistant for determining execution inefficiencies in dataflow programs
In a method, in response to an interface a computer-implemented analysis assistant initiates a presentation of inefficiency results, determined an efficiency analyzer based on a mapping of a dataflow program to execute on hardware of a computing system. The assistant receives an inefficiency included among the inefficiency results and composes formatted inefficiency results comprising a presentation format of the inefficiency to assist a developer of the dataflow program to interpret the inefficiency. The analysis assistant outputs the formatted inefficiency results to an interface, which can comprise an interface to output the formatted inefficiency results for use by the developer to improve the dataflow program in association with the inefficiency. In implementations the presentation can comprise an interactive presentation with a developer of the dataflow program. A computer program product and a computing system can implement the method. The computing system can execute the assistant, and can include the interfaces.
1. A method, the method comprising:
initiating, by a computer-implemented analysis assistant, responsive to an interface of a first computing system, a presentation session comprising a presentation of inefficiency results associated with a mapping of a set of operators of a dataflow program to execute the set of operators on hardware of a second computing system, the analysis assistant included in the first computing system;
receiving, by the analysis assistant, an inefficiency included among the inefficiency results;
composing by the analysis assistant, formatted inefficiency results comprising the inefficiency having a presentation format to assist a developer of the dataflow program to interpret the inefficiency; and,
outputting, by the analysis assistant, to an interface of a third computing system, the formatted inefficiency results, the interface of the third computing system comprising at least a portion to output the formatted inefficiency results for use by the developer to improve the dataflow program in association with the inefficiency.
2. The method of claim 1 , wherein the method further comprises:
determining, by the analysis assistant, a suggestion to modify the dataflow program to improve the inefficiency; and,
outputting, by the analysis assistant, using a second output interface of the third computing system, the suggestion.
3. The method of claim 2 , wherein the suggestion to modify the dataflow program to improve the inefficiency comprises one of a suggestion to modify the dataflow program to reduce the inefficiency and a suggestion to modify the dataflow program to remove the inefficiency.
4. The method of claim 1 , wherein the presentation session comprises an interactive presentation of the formatted inefficiency results.
5. The method of claim 4 , wherein the interactive presentation is included in an interaction of a compiler with a developer to program the dataflow program.
6. The method of claim 1 , wherein the second computing system comprises a coarse-grain reconfigurable system.
7. The method of claim 1 , wherein the hardware of the second computing system comprises at least one reconfigurable processor;
wherein the mapping comprises executing an operator, among the set of operators, on a processor among the at least one reconfigurable processor; and
wherein the inefficiency results are further associated with the executing the operator on the processor among the at least one reconfigurable processor.
8. The method of claim 1 , wherein at least one of the interface of the first computing system and the interface of the third computing system comprises an interface of a compiler for compiling the dataflow program.
9. The method of claim 1 , wherein the presentation format is selected from the group consisting of a tabular format; a graphical format; a pie chart; a bar chart; a latency waterfall chart; and, a combination of a bar chart and a latency waterfall chart.
10. The method of claim 1 , wherein the inefficiency is included among an inefficiency category selected from the group consisting of an inefficiency in an operator program to execute an operator of the dataflow program; a pipeline imbalance; a memory stall; a transient effect; an unused hardware component of the second computing system; and, an underutilized hardware component of the second computing system.
11. A computer program product, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, wherein the program instructions are executable by at least one processor to cause the at least one processor to:
initiate, responsive to an interface of a first computing system, a presentation session comprising a presentation of inefficiency results associated with a mapping of a set of operators of a dataflow program to execute the set of operators on hardware of a second computing system;
receive an inefficiency included among the inefficiency results;
compose formatted inefficiency results comprising the inefficiency having a presentation format to assist a developer of the dataflow program to interpret the inefficiency; and,
output, to an interface of a third computing system, the formatted inefficiency results, the interface of the third computing system comprising at least a portion an interface to output the formatted inefficiency results for use by the developer to improve the dataflow program in association with the inefficiency.
12. The computer program product of claim 11 , wherein the program instructions are executable by the at least one processor to further cause the at least one processor to:
determine, based on the inefficiency, a suggestion to modify the dataflow program to improve the inefficiency; and,
output, using a second output interface of the third computing system, the suggestion.
13. A first computing system, the first computing system comprising:
a first processor;
a second computing system comprising a second processor;
an efficiency analyzer configured to execute on the second processor to determine inefficiency results associated with executing a dataflow program on a third computing system; and,
an analysis assistant configured to execute on the first processor to:
initiate, responsive to an interface of the first computing system, a presentation session comprising a presentation of the inefficiency results;
receive the inefficiency results;
select, from among the inefficiency results, an inefficiency;
compose formatted inefficiency results comprising the inefficiency having a presentation format to assist a developer of the dataflow program to interpret the inefficiency; and,
output, to an interface of a fourth computing system, the formatted inefficiency results, the interface of the fourth computing system comprising at least a portion to output the formatted inefficiency results for use by the developer to improve the dataflow program in association with the inefficiency.
14. The first computing system of claim 13 , wherein the analysis assistant is configured to execute on the first processor to further:
determine a suggestion to modify the dataflow program to improve the inefficiency; and,
output, using a second output interface of the fourth computing system, the suggestion.
15. The first computing system of claim 13 , wherein the analysis assistant configured to execute on the first processor to initiate, responsive to an interface of the first computing system,
the presentation session comprises the analysis assistant configured to execute on the first processor to further initiate the presentation session to comprise an interactive presentation of the formatted inefficiency results.
16. The first computing system of claim 15 , wherein the interactive presentation is included in an interaction of a compiler with a developer to program the dataflow program.
17. The first computing system of claim 13 , wherein the presentation format is selected from the group consisting of a tabular format; a graphical format; a pie chart; a bar chart; a latency waterfall chart; and, a combination of a bar chart and a latency waterfall chart.
18. The first computing system of claim 13 , wherein the third computing system comprises at least one reconfigurable processor; and,
wherein the inefficiency results associated with the executing the dataflow program on the third computing system comprise inefficiency results associated with the executing an operator of the dataflow program on a processor among the at least one reconfigurable processor.
19. The first computing system of claim 13 , wherein at least one of the interface of the first computing system and the interface of the third computing system comprises an interface of a compiler for compiling the dataflow program.
20. The first computing system of claim 13 , wherein the inefficiency is included among an inefficiency category selected from the group consisting of an inefficiency in an operator program to execute an operator of the dataflow program; a pipeline imbalance; a memory stall; a transient effect; an unused hardware component of the second computing system; and, an underutilized hardware component of the second computing system.