The Future Technology Platform Project
Tuesday, June 10, 2025 3:00 PM to Thursday, June 12, 2025 4:00 PM · 2 days 1 hr. (Europe/Berlin)
Foyer D-G - 2nd floor
Project Poster
Emerging Computing TechnologiesHeterogeneous System ArchitecturesStorage Technologies and Architectures
Information
Poster is on display.
The Future Technology Platform (FTP) Project seeks to create a one-stop system to test and showcase new computing technologies and architectures, with a focus on high-performance computing, AI/ML, storage improvements, and lower power solutions.
Such a platform provides advantages at all levels of the wider computing community. Researchers gain access to a suite of up-and-coming hardware free of charge, where they can test and develop new algorithms in a risk-free environment. Industry members can use new technologies that are otherwise usually locked behind expensive acquisition and operation processes, or restricted to academic use only. Hardware vendors can profit from easier access to their systems. All levels will benefit from the work done by the project to develop documentation, training material and consulting for end users, and also to provide impartial benchmarks of computing capability and power consumption.
The project currently offers or is in the process of making available hardware such as the NVIDIA ARM DevKit which includes GPU and DPU capabilities, new AI oriented architectures such as EsperantoAI and GraphCore, and the neuromoprhic computing platform SpiNNaker. It has also been running a Scalable Storage Competition since late 2024 which is open to the public, with the goal of testing new approaches in the vital role that storage plays in high performance computing. First results of this event will be available in this poster and further discussed in a proposed BoF session at ISC 2025.
With this poster we intend to advertise the goals of the FTP project, present the various hardware solutions available for use in it, discuss with the community the advantages of the platform, and publicize the results of the current iteration of the Scalable Storage Competition as well as draw interest for the next iteration of the contest.
Contributors:
The Future Technology Platform (FTP) Project seeks to create a one-stop system to test and showcase new computing technologies and architectures, with a focus on high-performance computing, AI/ML, storage improvements, and lower power solutions.
Such a platform provides advantages at all levels of the wider computing community. Researchers gain access to a suite of up-and-coming hardware free of charge, where they can test and develop new algorithms in a risk-free environment. Industry members can use new technologies that are otherwise usually locked behind expensive acquisition and operation processes, or restricted to academic use only. Hardware vendors can profit from easier access to their systems. All levels will benefit from the work done by the project to develop documentation, training material and consulting for end users, and also to provide impartial benchmarks of computing capability and power consumption.
The project currently offers or is in the process of making available hardware such as the NVIDIA ARM DevKit which includes GPU and DPU capabilities, new AI oriented architectures such as EsperantoAI and GraphCore, and the neuromoprhic computing platform SpiNNaker. It has also been running a Scalable Storage Competition since late 2024 which is open to the public, with the goal of testing new approaches in the vital role that storage plays in high performance computing. First results of this event will be available in this poster and further discussed in a proposed BoF session at ISC 2025.
With this poster we intend to advertise the goals of the FTP project, present the various hardware solutions available for use in it, discuss with the community the advantages of the platform, and publicize the results of the current iteration of the Scalable Storage Competition as well as draw interest for the next iteration of the contest.
Contributors:
Format
On DemandOn Site
