Research Scientist , Revamping the Role of Modern Flash Storage in Accelerating Generative AI Pipelines

Rejoignez IBM Zurich comme Research Scientist pour innover. Profitez d'un environnement stimulant et d'avantages flexibles.

Tâches

• Explorer le rôle du stockage flash pour l'IA générationnelle.

• Analyser les pipelines RAG et leur dépendance au stockage.

• Développer des solutions pour optimiser les performances NVMe.

Compétences

• Doctorat en ingénierie ou informatique requis, avec compétences en codage.

• Compétences en Python et C++, expérience en développement logiciel.

• Excellente communication en anglais, créativité et esprit d'équipe.

Research Scientist , Revamping the Role of Modern Flash Storage in Accelerating Generative AI Pipelines

IBM Zurich Research Laboratory

Workplace Zurich, Zurich region, Switzerland Category Computer Science

Position Senior Scientist / Postdoc

Published 21 November 2024 Research Scientist

Revamping the Role of Modern Flash Storage in Accelerating Generative AI Pipelines

Ref. 2024_031

Position description

Generative AI (GenAI) pipelines, especially using the RAG pattern, have attracted a significant amount of community attention due to their widespread use, effectiveness, and potential applicability in a variety of domains. In an ideal case, all data/metadata used in these workflows will be contained in memory. However, DRAM technology faces challenges from multiple fronts as it is not scaling, has a high cost ($/GB), and is energy inefficient. As the performance of Flash-based NVMe devices keeps increasing, the key research question that we are interested in is: how can modern Flash memory and NVMe hardware/software stacks help run data-intensive GenAI pipelines (focus on RAG, i.e. information retrieval and LLM inference)?

In this context, we aim to answer the following research questions:

• What dependencies do RAG pipelines have on storage and how is storage accessed in RAG pipelines?

• How can Vector DBs benefit from using high-performance NVMe flash arrays that can support millions of small I/O operations per seconds?

• How can flash memory help accelerate the inference of LLMs?

Necessary qualifications

• The candidate has obtained a Ph.D. degree in electrical engineering or computer science with a focus on (graph) data mining, (graph) databases, data storage, or (generative) AI in the past 5 years.

• Excellent coding skills, in Python and C++, familiarity with Linux environments and software development tools (git/GitHub, IDEs, gcc, virtual machine and containers, etc.).

• High level of creativity and outstanding problem-solving ability.

• Excellent oral and written English and excellent presentation skills.

• Strong academic record and experience with publishing and presenting at international conferences.

• Strong interpersonal skills and excellent written and verbal communication.

Preferred Qualifications

• Background and experience in systems architecture and hardware design.

• Experience with systems programming and internals (kernel, memory management, hardware, CPU)

• Experience with data storage and NVMe storage internals and specification.

Diversity

IBM is committed to diversity at the workplace. With us you will find an open, multicultural environment. Excellent flexible working arrangements enable all genders to strike the desired balance between their professional development and their personal lives.

How to apply

If you are interested in this position, please submit your application here.

More information

In your application, please refer to myScience.ch and referenceJobID65889.