|
Sep 15th, 2023 |
|
|
|
Query 1: all publications |
|
Expected: Listing all the publications |
|
|
|
Actual: only some selected publications, missing some publications |
|
document_content: # Publication |
|
title=Jupiter: a networked computing architecture” |
|
venue=UCC Companion 2021: 28:1-28:8 |
|
authors=['P Ghosh', 'Q Nguyen', 'P Sakulkar', 'J Tran', 'A Knezevic', 'J Wang', 'Z Lin', 'B Krishnamachari', 'M Annavaram', 'S Avestimehr'] |
|
abstract=Modern latency-sensitive applications such as real-time multi-camera video analytics require networked computing to meet the time constraints. We present Jupiter, an open-source networked computing system that inputs a Directed Acyclic Graph (DAG)-based computational task graph to efficiently distribute the tasks among a set of networked compute nodes and orchestrates the execution thereafter. This Kubernetes container-orchestration-based system includes a range of profilers: network profilers, resource profilers, and execution time profilers; to support both centralized and decentralized scheduling algorithms. While centralized scheduling algorithms with global knowledge have been popular among the grid/cloud computing community, we argue that a distributed scheduling approach is better suited for networked computing due to lower communication and computation overhead in the face of network dynamics. We propose a new class of distributed scheduling algorithms called WAVE and show that despite using more localized knowledge, the WAVE algorithm can match the performance of a well-known centralized scheduling algorithm called Heterogeneous Earliest Finish Time (HEFT). To this, we present a set of real-world experiments on two separate testbeds: (1) a worldwide network of 90 cloud computers across eight cities and (2) a cluster of 30 Raspberry pi nodes. |
|
|
|
# Information |
|
links.pdf=/static/public/papers/JUPITER.pdf |
|
links.semantic_scholar=https://www.semanticscholar.org/paper/ef3485930a0eca1cd98b8dd7dfc7a919501ec06c |
|
type=Conference Papers |
|
year=2021 |
|
paper_id=bb6a0958 |
|
ss_title=Jupiter: a networked computing architecture |
|
ss_authors=[{'authorId': '49934897', 'name': 'Pradipta Ghosh'}, {'authorId': '145628959', 'name': 'Quynh Nguyen'}, {'authorId': '2254069', 'name': 'Pranav Sakulkar'}, {'authorId': '2073070356', 'name': 'Aleksandra Knezevic'}, {'authorId': '40553305', 'name': 'Jason A. Tran'}, {'authorId': '2109016686', 'name': 'Jiatong Wang'}, {'authorId': '46268272', 'name': 'Zhifeng Lin'}, {'authorId': '1701475', 'name': 'B. Krishnamachari'}, {'authorId': '145599558', 'name': 'M. Annavaram'}, {'authorId': '5877233', 'name': 'A. Avestimehr'}] |
|
ss_venue=UCC Companion |
|
ss_year=2019 |
|
ss_abstract=Modern latency-sensitive applications such as real-time multi-camera video analytics require networked computing to meet the time constraints. We present Jupiter, an open-source networked computing system that inputs a Directed Acyclic Graph (DAG)-based computational task graph to efficiently distribute the tasks among a set of networked compute nodes and orchestrates the execution thereafter. This Kubernetes container-orchestration-based system includes a range of profilers: network profilers, resource profilers, and execution time profilers; to support both centralized and decentralized scheduling algorithms. While centralized scheduling algorithms with global knowledge have been popular among the grid/cloud computing community, we argue that a distributed scheduling approach is better suited for networked computing due to lower communication and computation overhead in the face of network dynamics. We propose a new class of distributed scheduling algorithms called WAVE and show that despite using more localized knowledge, the WAVE algorithm can match the performance of a well-known centralized scheduling algorithm called Heterogeneous Earliest Finish Time (HEFT). To this, we present a set of real-world experiments on two separate testbeds: (1) a worldwide network of 90 cloud computers across eight cities and (2) a cluster of 30 Raspberry pi nodes. |
|
ss_paper_id=ef3485930a0eca1cd98b8dd7dfc7a919501ec06c |
|
Sure! Here are the details of all the publications mentioned in the input text: |
|
|
|
1. Publication: |
|
- Title: Jupiter: a networked computing architecture |
|
- Venue: UCC Companion 2021: 28:1-28:8 |
|
- Authors: P Ghosh, Q Nguyen, P Sakulkar, J Tran, A Knezevic, J Wang, Z Lin, B Krishnamachari, M Annavaram, S Avestimehr |
|
- Abstract: Modern latency-sensitive applications such as real-time multi-camera video analytics require networked computing to meet the time constraints. We present Jupiter, an open-source networked computing system that inputs a Directed Acyclic Graph (DAG)-based computational task graph to efficiently distribute the tasks among a set of networked compute nodes and orchestrates the execution thereafter. This Kubernetes container-orchestration-based system includes a range of profilers: network profilers, resource profilers, and execution time prof |
|
Hi! What question do you have for ANGR? press 0 to exit |
|
|
|
|
|
Query 2: math-related publications |
|
Expected: Solving Math Word Problems Concerning Systems of Equations with GPT-3 |
|
|
|
Actual: fits the expected (only search by titles but not sure if any other publication contents are also related) |
|
document_content: # Publication |
|
title=Solving Math Word Problems Concerning Systems of Equations with GPT-3 |
|
venue=Thirteenth AAAI Symposium on Educational Advances in Artificial Intelligence (EAAI-23) to be held February 11-12, 2023. |
|
authors=['M Zong', 'B Krishnamachari'] |
|
abstract=None |
|
|
|
# Information |
|
links.pdf=/static/public/papers/Solving_Math_Word_Problems_with_GPT3-2022.pdf |
|
type=Conference Papers |
|
year=2023 |
|
paper_id=42a8fad3 |
|
|
|
|
|
Query 3: all authors |
|
Expected: only author names |
|
|
|
Actual: publication names also included, and some authors are not included, such as E. Ondula who publicated “Using Reinforcement Learning for Operating Educational Campuses Safely during a Pandemic (Student Abstract)“, AAAI 2022. |
|
document_content: # Publication |
|
title=Jupiter: a networked computing architecture” |
|
venue=UCC Companion 2021: 28:1-28:8 |
|
authors=['P Ghosh', 'Q Nguyen', 'P Sakulkar', 'J Tran', 'A Knezevic', 'J Wang', 'Z Lin', 'B Krishnamachari', 'M Annavaram', 'S Avestimehr'] |
|
abstract=Modern latency-sensitive applications such as real-time multi-camera video analytics require networked computing to meet the time constraints. We present Jupiter, an open-source networked computing system that inputs a Directed Acyclic Graph (DAG)-based computational task graph to efficiently distribute the tasks among a set of networked compute nodes and orchestrates the execution thereafter. This Kubernetes container-orchestration-based system includes a range of profilers: network profilers, resource profilers, and execution time profilers; to support both centralized and decentralized scheduling algorithms. While centralized scheduling algorithms with global knowledge have been popular among the grid/cloud computing community, we argue that a distributed scheduling approach is better suited for networked computing due to lower communication and computation overhead in the face of network dynamics. We propose a new class of distributed scheduling algorithms called WAVE and show that despite using more localized knowledge, the WAVE algorithm can match the performance of a well-known centralized scheduling algorithm called Heterogeneous Earliest Finish Time (HEFT). To this, we present a set of real-world experiments on two separate testbeds: (1) a worldwide network of 90 cloud computers across eight cities and (2) a cluster of 30 Raspberry pi nodes. |
|
|
|
# Information |
|
links.pdf=/static/public/papers/JUPITER.pdf |
|
links.semantic_scholar=https://www.semanticscholar.org/paper/ef3485930a0eca1cd98b8dd7dfc7a919501ec06c |
|
type=Conference Papers |
|
year=2021 |
|
paper_id=bb6a0958 |
|
ss_title=Jupiter: a networked computing architecture |
|
ss_authors=[{'authorId': '49934897', 'name': 'Pradipta Ghosh'}, {'authorId': '145628959', 'name': 'Quynh Nguyen'}, {'authorId': '2254069', 'name': 'Pranav Sakulkar'}, {'authorId': '2073070356', 'name': 'Aleksandra Knezevic'}, {'authorId': '40553305', 'name': 'Jason A. Tran'}, {'authorId': '2109016686', 'name': 'Jiatong Wang'}, {'authorId': '46268272', 'name': 'Zhifeng Lin'}, {'authorId': '1701475', 'name': 'B. Krishnamachari'}, {'authorId': '145599558', 'name': 'M. Annavaram'}, {'authorId': '5877233', 'name': 'A. Avestimehr'}] |
|
ss_venue=UCC Companion |
|
ss_year=2019 |
|
ss_abstract=Modern latency-sensitive applications such as real-time multi-camera video analytics require networked computing to meet the time constraints. We present Jupiter, an open-source networked computing system that inputs a Directed Acyclic Graph (DAG)-based computational task graph to efficiently distribute the tasks among a set of networked compute nodes and orchestrates the execution thereafter. This Kubernetes container-orchestration-based system includes a range of profilers: network profilers, resource profilers, and execution time profilers; to support both centralized and decentralized scheduling algorithms. While centralized scheduling algorithms with global knowledge have been popular among the grid/cloud computing community, we argue that a distributed scheduling approach is better suited for networked computing due to lower communication and computation overhead in the face of network dynamics. We propose a new class of distributed scheduling algorithms called WAVE and show that despite using more localized knowledge, the WAVE algorithm can match the performance of a well-known centralized scheduling algorithm called Heterogeneous Earliest Finish Time (HEFT). To this, we present a set of real-world experiments on two separate testbeds: (1) a worldwide network of 90 cloud computers across eight cities and (2) a cluster of 30 Raspberry pi nodes. |
|
ss_paper_id=ef3485930a0eca1cd98b8dd7dfc7a919501ec06c |
|
The authors of the publication are: |
|
1. P Ghosh |
|
2. Q Nguyen |
|
3. P Sakulkar |
|
4. J Tran |
|
5. A Knezevic |
|
6. J Wang |
|
7. Z Lin |
|
8. B Krishnamachari |
|
9. M Annavaram |
|
10. S Avestimehr |
|
|
|
Query 4: What publications does B Krishnamachari contribute to |
|
Expected: |
|
"Search and Rescue on the Line", |
|
"Solving Math Word Problems Concerning Systems of Equations with GPT-3", |
|
"Graph Convolutional Network-based Scheduler for Distributing Computation in the Internet of Robotic Things", |
|
"Characterizing ML training performance at the tactical edge", |
|
"Optimal Trading on a Dynamic Curve Automated Market Maker", |
|
"Using Reinforcement Learning for Operating Educational Campuses Safely during a Pandemic (Student Abstract)", |
|
"\u201cMulti-objective network synthesis for dispersed computing in tactical environments\u201d", |
|
"\u201cIntelligent Communication over Realistic Wireless Networks in Multi-Agent Cooperative Games\u201d", |
|
"DEFER: Distributed Edge Inference for Deep Neural Networks", |
|
"Network Synthesis for Tactical Environments: Scenario, Challenges, and Opportunities", |
|
"Neural Networks for DDoS Attack Detection using an Enhanced Urban IoT Dataset", |
|
"Trip Planning for Autonomous Vehicles with Wireless Data Transfer Needs Using Reinforcement Learning", |
|
"\u201cLearning Practical Communication Strategies in Cooperative Multi-Agent Reinforcement Learning\u201d", |
|
"GCNScheduler: Scheduling distributed computing applications using graph convolutional networks", |
|
"\u201cRevealing a Hidden, Stable Spectral Structure of Urban Vehicular Traffic\u201d", |
|
"\u201cDataset: Large-scale Urban IoT Activity Data for DDoS Attack Emulation\u201d", |
|
"Jupiter: a networked computing architecture\u201d", |
|
"Course Scheduling to Minimize Student Wait Times For University Buildings During Epidemics", |
|
"\u201cSimulation-Based Analysis of COVID-19 Spread Through Classroom Transmission on a University Campus,\u201d", |
|
"A Decentralized Review System for Data Marketplaces", |
|
"DAISIM: A Computational Simulator for the MakerDAO Stablecoin", |
|
"Dynamic Curves for Decentralized Autonomous Cryptocurrency Exchanges", |
|
"Simulating the MakerDAO Stablecoin", |
|
"Dynamic Automated Market Makers for Decentralized Cryptocurrency Exchange", |
|
"CONTAIN: Privacy-oriented Contact Tracing Protocols for Epidemics", |
|
"Large-scale Urban IoT Activity Data for DDoS Attack Emulation", |
|
"TEAM: Trilateration for Exploration and Mapping with Robotic Networks" |
|
|
|
Actual: missing a lot publications |
|
document_content: # Publication |
|
title=https://anrg.usc.edu/www/papers/scheduling.pdf |
|
venue=IEEE BigData 2021: 4365-4370 |
|
authors=['A Hekmati', 'B Krishnamachari', 'M Mataric', 'Course Scheduling to Minimize Student Wait Times For University Buildings During Epidemics'] |
|
abstract=ABSTRACT Of the 17 years (1946–64) Jawaharlal Nehru was India's Prime Minister, his Congress Party government had a senior politician holding the Finance Ministry for only 5: Morarji Desai (1958–63). Otherwise, this crucial portfolio was held by a succession of experts; an illustration of consociational power sharing. This article is about two of those, C.D. Deshmukh and T.T. Krishnamachari, and some of their chequered governmental experiences. Through these, it seeks to cast a critical light on the changing contours of relations between and within the party and the government in the economic sphere during its decade of political dominance. |
|
|
|
# Information |
|
links.pdf=/static/public/papers/scheduling.pdf |
|
links.semantic_scholar=https://www.semanticscholar.org/paper/de91d2dbca72d5115198139618222094e0fa7cf0 |
|
type=Conference Papers |
|
year=2021 |
|
paper_id=0a9ed062 |
|
ss_title=Bearing ‘financial responsibility’ for the Government and the Party: C.D. Deshmukh (1950–56) & T.T. Krishnamachari (1956–58) |
|
ss_authors=[{'authorId': '90288069', 'name': 'R. Ankit'}] |
|
ss_venue= |
|
ss_year=2020 |
|
ss_abstract=ABSTRACT Of the 17 years (1946–64) Jawaharlal Nehru was India's Prime Minister, his Congress Party government had a senior politician holding the Finance Ministry for only 5: Morarji Desai (1958–63). Otherwise, this crucial portfolio was held by a succession of experts; an illustration of consociational power sharing. This article is about two of those, C.D. Deshmukh and T.T. Krishnamachari, and some of their chequered governmental experiences. Through these, it seeks to cast a critical light on the changing contours of relations between and within the party and the government in the economic sphere during its decade of political dominance. |
|
ss_paper_id=de91d2dbca72d5115198139618222094e0fa7cf0 |
|
Based on the given information, B Krishnamachari has contributed to the following publication: |
|
|
|
- Title: "Course Scheduling to Minimize Student Wait Times For University Buildings During Epidemics" |
|
- Authors: A Hekmati, B Krishnamachari, M Mataric |
|
- Venue: IEEE BigData 2021 |
|
- Pages: 4365-4370 |
|
- Year: 2021 |
|
- Publication Link: [here](https://anrg.usc.edu/www/papers/scheduling.pdf) |
|
|
|
Please note that this is the only publication listed for B Krishnamachari based on the provided information. |
|
|