John Zawistowski, Sycomp; Leon Platts, IBM; and Hugo Saleh, Google Cloud | SC23
summary
Sycomp delivers a cloud-ready parallel file system called Sycomp Storage Fueled by IBM Storage Scale that hydrates data from Google Cloud Storage into a global namespace so on‑prem and cloud jobs access one source of truth. The platform proves out high throughput — customers hit 300 GB/s throughput for trading workloads — while removing parallel‑FS complexity and offering a workflow that can spin up VMs and spin down the parallel file system to save cost.
AI-written summary · about
chapters (15)
- 0:00Introduction to panel
- 0:44HPC plus AI transformation
- 2:11IBM's HPC market focus
- 3:33Solving customer workload challenges
- 5:23Architectural shifts in data management
- 7:28Constraints and blockers in HPC
- 9:12Access to high performance computing
- 10:58Future of HPC and AI at Sycomp
- 12:41HPC going mainstream with AI
- 14:06Grid management challenges
- 15:31Data governance in AI era
- 16:55AI tech stack and infrastructure
- 18:21IBM GPFS cloud adaptation
- 20:06Sycomp's future focus
- 21:09Closing remarks
