Sotabase

Career

· Research Scientist, Meta AI2019–
· Research Staff Member, IBM Thomas J. Watson Research Center2017–2019
· Research Intern, IBM T.J. Watson Research Center2016–2016
· Research Intern, Lawrence Livermore National Laboratory2014–2014
· Teaching Assistant, University of Illinois at Urbana-Champaign2013–2013
· PhD Student, University of Illinois at Urbana-Champaign2012–2017
· Research Assistant, University of Illinois at Urbana-Champaign2012–2017
· Intern, Microsoft2011–
· Intern, TUBITAK UEKAE2011–
· Erasmus Exchange Program, Uppsala University2011–
· Intern, Middle East Technical University2010–
· Bachelor's degree, Bilkent University2008–2012

Publications (29)

Sustainable AI: Environmental Implications, Challenges and Opportunities
Conference on Machine Learning and Systems · 2021
553
cited
LayerSkip: Enabling Early Exit Inference and Self-Speculative Decoding
Annual Meeting of the Association for Computational Linguistics · 2024
192
cited
Parallel Programming with Migratable Objects: Charm++ in Practice
International Conference for High Performance Computing, Networking, Storage and Analysis · 2014
159
cited
Carbon Explorer: A Holistic Framework for Designing Carbon Aware Datacenters
International Conference on Architectural Support for Programming Languages and Operating Systems · 2022
150
cited
DataPerf: Benchmarks for Data-Centric AI Development
Neural Information Processing Systems · 2022
129
cited
Understanding Training Efficiency of Deep Learning Recommendation Models at Scale
International Symposium on High-Performance Computer Architecture · 2020
123
cited
TT-Rec: Tensor Train Compression for Deep Learning Recommendation Models
Conference on Machine Learning and Systems · 2021
115
cited
RecShard: statistical feature-based memory optimization for industry-scale neural recommendation
International Conference on Architectural Support for Programming Languages and Operating Systems · 2022
78
cited
Towards realizing the potential of malleable jobs
International Conference on High Performance Computing · 2014
52
cited
A Holistic Approach for Designing Carbon Aware Datacenters
arXiv.org · 2022
30
cited
Beyond Efficiency: Scaling AI Sustainably
IEEE Micro · 2024
27
cited
MP-Rec: Hardware-Software Co-design to Enable Multi-path Recommendation
International Conference on Architectural Support for Programming Languages and Operating Systems · 2023
20
cited
Carbon Dependencies in Datacenter Design and Management
ACM SIGEnergy Energy Informatics Review · 2023
14
cited
Datacenter-Scale Analysis and Optimization of GPU Machine Learning Workloads
IEEE Micro · 2021
14
cited
MAD-Max Beyond Single-Node: Enabling Large Machine Learning Model Acceleration on Distributed Systems
International Symposium on Computer Architecture · 2023
14
cited
Carbon Responder: Coordinating Demand Response for the Datacenter Fleet
arXiv.org · 2023
13
cited
Data Acquisition: A New Frontier in Data-centric AI
arXiv.org · 2023
13
cited
Generative AI Beyond LLMs: System Implications of Multi-Modal Generation
IEEE International Symposium on Performance Analysis of Systems and Software · 2023
13
cited
SecNDP: Secure Near-Data Processing with Untrusted Memory
International Symposium on High-Performance Computer Architecture · 2022
13
cited
CHAI: Clustered Head Attention for Efficient LLM Inference
International Conference on Machine Learning · 2024
12
cited
Sotabase
Bilge Acun | Researcher Profile | Sotabase | Sotabase