Speaker
Mr
Maiken Pedersen
Description
The Nordic Tier-1 site for LHC data and computing is a distributed one,
where computing and storage is spread over several countries. In order
to do efficient computing on the data, a local low-latency cache is needed
close to each computing cluster.
This talk will give an overview on how this works in production today, with
data throughputs in the 100Gbit/s range, and then talk about the challenges
ahead where scaling of a factor of 10-20 on top of that will be needed to
keep up with the increased data rates from the High-Luminosity LHC upgrade
that comes online in 2030.
The challenges of staging data at 1 Tbit/s will be discussed from networking,
hardware, software, and operational points of view.
| Suggested Contribution Type | Regular Talk (15-30 min) |
|---|
Authors
Mr
Maiken Pedersen
Mattias Wadenstein
(University of Umeå (SE))