�ݺ�ߣshows by User: chettyharish

�ݺ�ߣshows by User: chettyharish / http://www.slideshare.net/images/logo.gif �ݺ�ߣshows by User: chettyharish / Sun, 24 Jan 2016 22:33:27 GMT �ݺ�ߣShare feed for �ݺ�ߣshows by User: chettyharish Scourge /slideshow/scourge-57438736/57438736 scourge-160124223327
A survival based horror game based on Blender Game Engine v1.72b . Uses a combination of vector texture shaders , UV texture mappers , NURBS paths and 3D sound detectors. I carried out the job of modeling ,scripting and level development of the game.]]>
A survival based horror game based on Blender Game Engine v1.72b . Uses a combination of vector texture shaders , UV texture mappers , NURBS paths and 3D sound detectors. I carried out the job of modeling ,scripting and level development of the game.]]> Sun, 24 Jan 2016 22:33:27 GMT /slideshow/scourge-57438736/57438736 chettyharish@slideshare.net(chettyharish) Scourge chettyharish A survival based horror game based on Blender Game Engine v1.72b . Uses a combination of vector texture shaders , UV texture mappers , NURBS paths and 3D sound detectors. I carried out the job of modeling ,scripting and level development of the game. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/scourge-160124223327-thumbnail.jpg?width=120&height=120&fit=bounds" /> A survival based horror game based on Blender Game Engine v1.72b . Uses a combination of vector texture shaders , UV texture mappers , NURBS paths and 3D sound detectors. I carried out the job of modeling ,scripting and level development of the game.

Scourge from Harish Chetty

]]> 716 5 https://cdn.slidesharecdn.com/ss_thumbnails/scourge-160124223327-thumbnail.jpg?width=120&height=120&fit=bounds presentation 000000 http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Silent stores /slideshow/silent-stores-57438543/57438543 silentstores-160124222457
The research project is geared towards analyzing the nature and ratio of silent stores in X86 architecture. We use the gem5 simulator to generate the required simulation of various benchmarks and cloud servers to process the massively time consuming processes. Clustering between stores is determined using Run Length Encoding of Top 200 sequences (one set of inter-related silent and non-silent stores and other set of only silent-stores)]]>
The research project is geared towards analyzing the nature and ratio of silent stores in X86 architecture. We use the gem5 simulator to generate the required simulation of various benchmarks and cloud servers to process the massively time consuming processes. Clustering between stores is determined using Run Length Encoding of Top 200 sequences (one set of inter-related silent and non-silent stores and other set of only silent-stores)]]> Sun, 24 Jan 2016 22:24:57 GMT /slideshow/silent-stores-57438543/57438543 chettyharish@slideshare.net(chettyharish) Silent stores chettyharish The research project is geared towards analyzing the nature and ratio of silent stores in X86 architecture. We use the gem5 simulator to generate the required simulation of various benchmarks and cloud servers to process the massively time consuming processes. Clustering between stores is determined using Run Length Encoding of Top 200 sequences (one set of inter-related silent and non-silent stores and other set of only silent-stores) <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/silentstores-160124222457-thumbnail.jpg?width=120&height=120&fit=bounds" /> The research project is geared towards analyzing the nature and ratio of silent stores in X86 architecture. We use the gem5 simulator to generate the required simulation of various benchmarks and cloud servers to process the massively time consuming processes. Clustering between stores is determined using Run Length Encoding of Top 200 sequences (one set of inter-related silent and non-silent stores and other set of only silent-stores)

Silent stores from Harish Chetty

]]> 169 5 https://cdn.slidesharecdn.com/ss_thumbnails/silentstores-160124222457-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Polling server /slideshow/polling-server-57438542/57438542 pollingserver-160124222456
The project is a polling server for aperiodic tasks (SCHED_POLL) based on Earliest Deadline First (SCHED_DEADLINE) on Linux Kernel v3.15.1. The scheduler uses various threshold limiters for each tasks based on RB trees and also stores address of the leftmost element/node (to determine earliest task ). Currently it does extremely quick task scheduling (due to storing of leftmost element) but has an error ratio of 5% w.r.t. to ideal conditions.]]>
The project is a polling server for aperiodic tasks (SCHED_POLL) based on Earliest Deadline First (SCHED_DEADLINE) on Linux Kernel v3.15.1. The scheduler uses various threshold limiters for each tasks based on RB trees and also stores address of the leftmost element/node (to determine earliest task ). Currently it does extremely quick task scheduling (due to storing of leftmost element) but has an error ratio of 5% w.r.t. to ideal conditions.]]> Sun, 24 Jan 2016 22:24:56 GMT /slideshow/polling-server-57438542/57438542 chettyharish@slideshare.net(chettyharish) Polling server chettyharish The project is a polling server for aperiodic tasks (SCHED_POLL) based on Earliest Deadline First (SCHED_DEADLINE) on Linux Kernel v3.15.1. The scheduler uses various threshold limiters for each tasks based on RB trees and also stores address of the leftmost element/node (to determine earliest task ). Currently it does extremely quick task scheduling (due to storing of leftmost element) but has an error ratio of 5% w.r.t. to ideal conditions. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/pollingserver-160124222456-thumbnail.jpg?width=120&height=120&fit=bounds" /> The project is a polling server for aperiodic tasks (SCHED_POLL) based on Earliest Deadline First (SCHED_DEADLINE) on Linux Kernel v3.15.1. The scheduler uses various threshold limiters for each tasks based on RB trees and also stores address of the leftmost element/node (to determine earliest task ). Currently it does extremely quick task scheduling (due to storing of leftmost element) but has an error ratio of 5% w.r.t. to ideal conditions.

Polling server from Harish Chetty

]]> 204 4 https://cdn.slidesharecdn.com/ss_thumbnails/pollingserver-160124222456-thumbnail.jpg?width=120&height=120&fit=bounds presentation 000000 http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 N way merge sort /slideshow/n-way-merge-sort/57438478 n-waymergesort-160124222234
A large scale distributed external sorter which can handle N machines simultaneously. The system supports a master-slave architecture where the slaves handle sorting and the master handles transmission, socket control , network flow and final N-way merge to achieve the final result.]]>
A large scale distributed external sorter which can handle N machines simultaneously. The system supports a master-slave architecture where the slaves handle sorting and the master handles transmission, socket control , network flow and final N-way merge to achieve the final result.]]> Sun, 24 Jan 2016 22:22:33 GMT /slideshow/n-way-merge-sort/57438478 chettyharish@slideshare.net(chettyharish) N way merge sort chettyharish A large scale distributed external sorter which can handle N machines simultaneously. The system supports a master-slave architecture where the slaves handle sorting and the master handles transmission, socket control , network flow and final N-way merge to achieve the final result. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/n-waymergesort-160124222234-thumbnail.jpg?width=120&height=120&fit=bounds" /> A large scale distributed external sorter which can handle N machines simultaneously. The system supports a master-slave architecture where the slaves handle sorting and the master handles transmission, socket control , network flow and final N-way merge to achieve the final result.

N way merge sort from Harish Chetty

]]> 621 4 https://cdn.slidesharecdn.com/ss_thumbnails/n-waymergesort-160124222234-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 Access pattern of tags /slideshow/access-pattern-of-tags-57438348/57438348 accesspatternoftags-160124221744
The project is based on detecting frequent patterns on a web-server by using the server log. It generates multi-level statistics based on Apriori algorithm with each level determined by Pattern length. Currently this is a dynamic programming form of Apriori where we have implemented the rule that to be a frequent pattern at a level k (i.e patter of size k), one of the elements from its powerset of size (k-1) must be frequent. This helps in improving the efficiency to less than quadratic time on average. ]]>
The project is based on detecting frequent patterns on a web-server by using the server log. It generates multi-level statistics based on Apriori algorithm with each level determined by Pattern length. Currently this is a dynamic programming form of Apriori where we have implemented the rule that to be a frequent pattern at a level k (i.e patter of size k), one of the elements from its powerset of size (k-1) must be frequent. This helps in improving the efficiency to less than quadratic time on average. ]]> Sun, 24 Jan 2016 22:17:44 GMT /slideshow/access-pattern-of-tags-57438348/57438348 chettyharish@slideshare.net(chettyharish) Access pattern of tags chettyharish The project is based on detecting frequent patterns on a web-server by using the server log. It generates multi-level statistics based on Apriori algorithm with each level determined by Pattern length. Currently this is a dynamic programming form of Apriori where we have implemented the rule that to be a frequent pattern at a level k (i.e patter of size k), one of the elements from its powerset of size (k-1) must be frequent. This helps in improving the efficiency to less than quadratic time on average. <img style="border:1px solid #C3E6D8;float:right;" alt="" src="https://cdn.slidesharecdn.com/ss_thumbnails/accesspatternoftags-160124221744-thumbnail.jpg?width=120&height=120&fit=bounds" /> The project is based on detecting frequent patterns on a web-server by using the server log. It generates multi-level statistics based on Apriori algorithm with each level determined by Pattern length. Currently this is a dynamic programming form of Apriori where we have implemented the rule that to be a frequent pattern at a level k (i.e patter of size k), one of the elements from its powerset of size (k-1) must be frequent. This helps in improving the efficiency to less than quadratic time on average.

Access pattern of tags from Harish Chetty

]]> 206 4 https://cdn.slidesharecdn.com/ss_thumbnails/accesspatternoftags-160124221744-thumbnail.jpg?width=120&height=120&fit=bounds presentation Black http://activitystrea.ms/schema/1.0/post

http://activitystrea.ms/schema/1.0/posted

0 https://cdn.slidesharecdn.com/profile-photo-chettyharish-48x48.jpg?cb=1571551419 https://cdn.slidesharecdn.com/ss_thumbnails/scourge-160124223327-thumbnail.jpg?width=320&height=320&fit=bounds slideshow/scourge-57438736/57438736 Scourge https://cdn.slidesharecdn.com/ss_thumbnails/silentstores-160124222457-thumbnail.jpg?width=320&height=320&fit=bounds slideshow/silent-stores-57438543/57438543 Silent stores https://cdn.slidesharecdn.com/ss_thumbnails/pollingserver-160124222456-thumbnail.jpg?width=320&height=320&fit=bounds slideshow/polling-server-57438542/57438542 Polling server