Patents by Inventor Emre Kultursay
Emre Kultursay has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).
-
Publication number: 20230153087Abstract: The invention comprises (i) a compilation method for automatically converting a single-threaded software program into an application-specific supercomputer, and (ii) the supercomputer system structure generated as a result of applying this method. The compilation method comprises: (a) Converting an arbitrary code fragment from the application into customized hardware whose execution is functionally equivalent to the software execution of the code fragment; and (b) Generating interfaces on the hardware and software parts of the application, which (i) Perform a software-to-hardware program state transfer at the entries of the code fragment; (ii) Perform a hardware-to-software program state transfer at the exits of the code fragment; and (iii) Maintain memory coherence between the software and hardware memories. If the resulting hardware design is large, it is divided into partitions such that each partition can fit into a single chip. Then, a single union chip is created which can realize any of the partitions.Type: ApplicationFiled: January 16, 2023Publication date: May 18, 2023Applicant: Global Supercomputing CorporationInventors: Kemal Ebcioglu, Emre Kultursay
-
Patent number: 11579854Abstract: The invention comprises (i) a compilation method for automatically converting a single-threaded software program into an application-specific supercomputer, and (ii) the supercomputer system structure generated as a result of applying this method. The compilation method comprises: (a) Converting an arbitrary code fragment from the application into customized hardware whose execution is functionally equivalent to the software execution of the code fragment; and (b) Generating interfaces on the hardware and software parts of the application, which (i) Perform a software-to-hardware program state transfer at the entries of the code fragment; (ii) Perform a hardware-to-software program state transfer at the exits of the code fragment; and (iii) Maintain memory coherence between the software and hardware memories. If the resulting hardware design is large, it is divided into partitions such that each partition can fit into a single chip. Then, a single union chip is created which can realize any of the partitions.Type: GrantFiled: August 25, 2021Date of Patent: February 14, 2023Assignee: Global Supercomputing CorporationInventors: Kemal Ebcioglu, Emre Kultursay
-
Publication number: 20210389936Abstract: The invention comprises (i) a compilation method for automatically converting a single-threaded software program into an application-specific supercomputer, and (ii) the supercomputer system structure generated as a result of applying this method. The compilation method comprises: (a) Converting an arbitrary code fragment from the application into customized hardware whose execution is functionally equivalent to the software execution of the code fragment; and (b) Generating interfaces on the hardware and software parts of the application, which (i) Perform a software-to-hardware program state transfer at the entries of the code fragment; (ii) Perform a hardware-to-software program state transfer at the exits of the code fragment; and (iii) Maintain memory coherence between the software and hardware memories. If the resulting hardware design is large, it is divided into partitions such that each partition can fit into a single chip. Then, a single union chip is created which can realize any of the partitions.Type: ApplicationFiled: August 25, 2021Publication date: December 16, 2021Applicant: Global Supercomputing CorporationInventors: Kemal Ebcioglu, Emre Kultursay
-
Patent number: 11132186Abstract: The invention comprises (i) a compilation method for automatically converting a single-threaded software program into an application-specific supercomputer, and (ii) the supercomputer system structure generated as a result of applying this method. The compilation method comprises: (a) Converting an arbitrary code fragment from the application into customized hardware whose execution is functionally equivalent to the software execution of the code fragment; and (b) Generating interfaces on the hardware and software parts of the application, which (i) Perform a software-to-hardware program state transfer at the entries of the code fragment; (ii) Perform a hardware-to-software program state transfer at the exits of the code fragment; and (iii) Maintain memory coherence between the software and hardware memories. If the resulting hardware design is large, it is divided into partitions such that each partition can fit into a single chip. Then, a single union chip is created which can realize any of the partitions.Type: GrantFiled: March 16, 2020Date of Patent: September 28, 2021Assignee: Global Supercomputing CorporationInventors: Kemal Ebcioglu, Emre Kultursay
-
Publication number: 20200278848Abstract: The invention comprises (i) a compilation method for automatically converting a single-threaded software program into an application-specific supercomputer, and (ii) the supercomputer system structure generated as a result of applying this method. The compilation method comprises: (a) Converting an arbitrary code fragment from the application into customized hardware whose execution is functionally equivalent to the software execution of the code fragment; and (b) Generating interfaces on the hardware and software parts of the application, which (i) Perform a software-to-hardware program state transfer at the entries of the code fragment; (ii) Perform a hardware-to-software program state transfer at the exits of the code fragment; and (iii) Maintain memory coherence between the software and hardware memories. If the resulting hardware design is large, it is divided into partitions such that each partition can fit into a single chip. Then, a single union chip is created which can realize any of the partitions.Type: ApplicationFiled: March 16, 2020Publication date: September 3, 2020Applicant: Global Supercomputing CorporationInventors: Kemal Ebcioglu, Emre Kultursay
-
Patent number: 10642588Abstract: The invention comprises (i) a compilation method for automatically converting a single-threaded software program into an application-specific supercomputer, and (ii) the supercomputer system structure generated as a result of applying this method. The compilation method comprises: (a) Converting an arbitrary code fragment from the application into customized hardware whose execution is functionally equivalent to the software execution of the code fragment; and (b) Generating interfaces on the hardware and software parts of the application, which (i) Perform a software-to-hardware program state transfer at the entries of the code fragment; (ii) Perform a hardware-to-software program state transfer at the exits of the code fragment; and (iii) Maintain memory coherence between the software and hardware memories. If the resulting hardware design is large, it is divided into partitions such that each partition can fit into a single chip. Then, a single union chip is created which can realize any of the partitions.Type: GrantFiled: October 22, 2018Date of Patent: May 5, 2020Assignee: Global Supercomputing CorporationInventors: Kemal Ebcioglu, Emre Kultursay
-
Publication number: 20190114158Abstract: The invention comprises (i) a compilation method for automatically converting a single-threaded software program into an application-specific supercomputer, and (ii) the supercomputer system structure generated as a result of applying this method. The compilation method comprises: (a) Converting an arbitrary code fragment from the application into customized hardware whose execution is functionally equivalent to the software execution of the code fragment; and (b) Generating interfaces on the hardware and software parts of the application, which (i) Perform a software-to-hardware program state transfer at the entries of the code fragment; (ii) Perform a hardware-to-software program state transfer at the exits of the code fragment; and (iii) Maintain memory coherence between the software and hardware memories. If the resulting hardware design is large, it is divided into partitions such that each partition can fit into a single chip. Then, a single union chip is created which can realize any of the partitions.Type: ApplicationFiled: October 22, 2018Publication date: April 18, 2019Applicant: Global Supercomputing CorporationInventors: Kemal Ebcioglu, Emre Kultursay
-
Patent number: 10146516Abstract: The invention comprises (i) a compilation method for automatically converting a single-threaded software program into an application-specific supercomputer, and (ii) the supercomputer system structure generated as a result of applying this method. The compilation method comprises: (a) Converting an arbitrary code fragment from the application into customized hardware whose execution is functionally equivalent to the software execution of the code fragment; and (b) Generating interfaces on the hardware and software parts of the application, which (i) Perform a software-to-hardware program state transfer at the entries of the code fragment; (ii) Perform a hardware-to-software program state transfer at the exits of the code fragment; and (iii) Maintain memory coherence between the software and hardware memories. If the resulting hardware design is large, it is divided into partitions such that each partition can fit into a single chip. Then, a single union chip is created which can realize any of the partitions.Type: GrantFiled: September 6, 2016Date of Patent: December 4, 2018Assignee: Global Supercomputing CorporationInventors: Kemal Ebcioglu, Emre Kultursay
-
Publication number: 20170017476Abstract: The invention comprises (i) a compilation method for automatically converting a single-threaded software program into an application-specific supercomputer, and (ii) the supercomputer system structure generated as a result of applying this method. The compilation method comprises: (a) Converting an arbitrary code fragment from the application into customized hardware whose execution is functionally equivalent to the software execution of the code fragment; and (b) Generating interfaces on the hardware and software parts of the application, which (i) Perform a software-to-hardware program state transfer at the entries of the code fragment; (ii) Perform a hardware-to-software program state transfer at the exits of the code fragment; and (iii) Maintain memory coherence between the software and hardware memories. If the resulting hardware design is large, it is divided into partitions such that each partition can fit into a single chip. Then, a single union chip is created which can realize any of the partitions.Type: ApplicationFiled: September 6, 2016Publication date: January 19, 2017Applicant: Global Supercomputing CorporationInventors: Kemal Ebcioglu, Emre Kultursay
-
Patent number: 9495223Abstract: The invention comprises (i) a compilation method for automatically converting a single-threaded software program into an application-specific supercomputer, and (ii) the supercomputer system structure generated as a result of applying this method. The compilation method comprises: (a) Converting an arbitrary code fragment from the application into customized hardware whose execution is functionally equivalent to the software execution of the code fragment; and (b) Generating interfaces on the hardware and software parts of the application, which (i) Perform a software-to-hardware program state transfer at the entries of the code fragment; (ii) Perform a hardware-to-software program state transfer at the exits of the code fragment; and (iii) Maintain memory coherence between the software and hardware memories. If the resulting hardware design is large, it is divided into partitions such that each partition can fit into a single chip. Then, a single union chip is created which can realize any of the partitions.Type: GrantFiled: December 23, 2014Date of Patent: November 15, 2016Assignee: Global Supercomputing CorporationInventors: Kemal Ebcioglu, Emre Kultursay
-
Patent number: 9262300Abstract: A method for capturing breakpoint information from a debuggee software process includes generating a breakpoint condition based upon a breakpoint request received from a user computing device corresponding to a user and transmitting the generated breakpoint condition to debuglets, each corresponding to a software process executed by a debuggee service. The debuggee service executes on a distributed system, and each debuglet translates the generated breakpoint condition to a physical breakpoint condition set to the respective software process. The method also includes receiving a request from one of the debuglets to update active breakpoint information captured by the debuglet upon the physical breakpoint condition being hit by one of the software processes and transmitting a notification from the processing device indicating the physical breakpoint condition being hit to the user computing device.Type: GrantFiled: March 19, 2015Date of Patent: February 16, 2016Assignee: Google Inc.Inventors: Erez Haba, Emre Kultursay, Vladimir Lifliand, Amnon Omri Horowitz
-
Publication number: 20150317190Abstract: The invention comprises (i) a compilation method for automatically converting a single-threaded software program into an application-specific supercomputer, and (ii) the supercomputer system structure generated as a result of applying this method. The compilation method comprises: (a) Converting an arbitrary code fragment from the application into customized hardware whose execution is functionally equivalent to the software execution of the code fragment; and (b) Generating interfaces on the hardware and software parts of the application, which (i) Perform a software-to-hardware program state transfer at the entries of the code fragment; (ii) Perform a hardware-to-software program state transfer at the exits of the code fragment; and (iii) Maintain memory coherence between the software and hardware memories. If the resulting hardware design is large, it is divided into partitions such that each partition can fit into a single chip. Then, a single union chip is created which can realize any of the partitions.Type: ApplicationFiled: December 23, 2014Publication date: November 5, 2015Applicant: GLOBAL SUPERCOMPUTING CORPORATIONInventors: Kemal Ebcioglu, Emre Kultursay
-
Patent number: 8966457Abstract: The invention comprises (i) a compilation method for automatically converting a single-threaded software program into an application-specific supercomputer, and (ii) the supercomputer system structure generated as a result of applying this method. The compilation method comprises: (a) Converting an arbitrary code fragment from the application into customized hardware whose execution is functionally equivalent to the software execution of the code fragment; and (b) Generating interfaces on the hardware and software parts of the application, which (i) Perform a software-to-hardware program state transfer at the entries of the code fragment; (ii) Perform a hardware-to-software program state transfer at the exits of the code fragment; and (iii) Maintain memory coherence between the software and hardware memories. If the resulting hardware design is large, it is divided into partitions such that each partition can fit into a single chip. Then, a single union chip is created which can realize any of the partitions.Type: GrantFiled: November 15, 2011Date of Patent: February 24, 2015Assignee: Global Supercomputing CorporationInventors: Kemal Ebcioglu, Emre Kultursay, Mahmut Taylan Kandemir
-
Patent number: 8825982Abstract: A method is described to partition the memory of application-specific hardware compiled from a software program. Applying the invention generates multiple small memories that need not be kept coherent and are defined over a specific region of the program. The invention creates application specific hardware which preserves the memory image and addressing model of the original software program. The memories are dynamically initialized and flushed at the entries and exits of the program region they are defined in.Type: GrantFiled: June 9, 2011Date of Patent: September 2, 2014Assignee: Global Supercomputing CorporationInventors: Emre Kultursay, Kemal Ebcioglu, Mahmut Taylan Kandemir
-
Publication number: 20130125097Abstract: The invention comprises (i) a compilation method for automatically converting a single-threaded software program into an application-specific supercomputer, and (ii) the supercomputer system structure generated as a result of applying this method. The compilation method comprises: (a) Converting an arbitrary code fragment from the application into customized hardware whose execution is functionally equivalent to the software execution of the code fragment; and (b) Generating interfaces on the hardware and software parts of the application, which (i) Perform a software-to-hardware program state transfer at the entries of the code fragment; (ii) Perform a hardware-to-software program state transfer at the exits of the code fragment; and (iii) Maintain memory coherence between the software and hardware memories. If the resulting hardware design is large, it is divided into partitions such that each partition can fit into a single chip. Then, a single union chip is created which can realize any of the partitions.Type: ApplicationFiled: November 15, 2011Publication date: May 16, 2013Applicant: Global Supercomputing CorporationInventors: Kemal Ebcioglu, Emre Kultursay, Mahmut Taylan Kandemir
-
Publication number: 20110307663Abstract: A method is described to partition the memory of application-specific hardware compiled from a software program. Applying the invention generates multiple small memories that need not be kept coherent and are defined over a specific region of the program. The invention creates application specific hardware which preserves the memory image and addressing model of the original software program. The memories are dynamically initialized and flushed at the entries and exits of the program region they are defined in.Type: ApplicationFiled: June 9, 2011Publication date: December 15, 2011Applicant: GLOBAL SUPERCOMPUTING CORPORATIONInventors: Emre Kultursay, Kemal Ebcioglu, Mahmut Taylan Kandemir