Sep 22 – 26, 2014
Centre des Congrès - Aix en Provence, France
Europe/Zurich timezone

NaNet: a Configurable NIC Bridging the Gap Between HPC and Real-time HEP GPU Computing

Sep 24, 2014, 5:55 PM
1m
Centre des Congrès - Aix en Provence, France

Centre des Congrès - Aix en Provence, France

14 boulevard Carnot 13100
Poster Trigger Second Poster Session

Speaker

Alessandro Lonardo (Universita e INFN, Roma I (IT))

Description

NaNet is a FPGA-BASED PCIe Network Interface Card with GPUDirect capability featuring a configurable set of channels: standard 1/10GbE and custom 34Gbps APElink and 2.5Gbps optical with deterministic latency KM3link. GPUDirect feature combined with a transport layer offload module and a data stream processing stage makes NaNet a low-latency NIC suitable for real-time GPU processing. We will describe NaNet architecture and its performances, and present two use cases for it: the GPU-based low-level trigger for the RICH detector in NA62 experiment and the on-/off-shore data link for KM3 underwater neutrino telescope.

Summary

Although GPGPU is widely accepted as an effective approach to
high performance computing, its adoption in low-latency, hard real-time
processing systems, like low level triggers in HEP experiments, still poses
several challenges.

GPUs show a rather deterministic behaviour in terms of processing
latency once input data are available in their internal memories, but
assessment of the real-time features of a whole GPGPU system takes a
careful characterization of all subsystems along data stream path.
In our analysis we identified the networking subsystem as the
most critical one because of the relevant fluctuations in its response
latency.

To overcome this issue, we designed NaNet, a FPGA-based PCIe Network
Interface Card (NIC) featuring a configurable set of network channels and
capable of receiving and sending data directly to and from Nvidia Fermi/Kepler
GPU internal memories without intermediate buffering on host memory
(GPUDirect).

The design includes a transport layer offload module with cycle-accurate deterministic latency, with support for UDP and custom KM3link and APElink protocols, added to eliminate host OS intervention on data stream and thus avoiding a possible source of jitter.

NaNet design currently supports both standard - 1GbE (1000Base-T) and
10GbE (10Base-R) - and custom - 34Gbps APElink and 2.5Gbps
deterministic latency KM3link - channels, but its modularity allows
for a straightforward inclusion of other link technologies.

An application specific module operates on input/output data streams,
performing processing on them with cycle-accurate deterministic
latency (e.g. to perform decompression and to rearrange data structures in a GPU-friendly fashion before storing them in GPU memory).

We will describe NaNet architecture and its latency/bandwidth characterization for all supported links and present NaNet usage in the NA62 and KM3 experiments.

The NA62 experiment at CERN aims at measuring the branching ratio of
the ultra-rare charged kaon decay into a pion and a neutrino/antineutrino pair.

The ~10 MHz rate of particles reaching the detectors must be reduced
by the multilevel trigger down to a ~ kHz rate, manageable by the data
storage system. First level (L0) is implemented in dedicated hardware
performing rough selections on their output reducing ~10 times the data
stream rate to match the ≤ 1MHz event target rate within 1ms time budget.
A GPU-based L0 trigger for the RICH detector using NaNet is being integrated in a parasitic mode in the experimental setup; this will allow assessing the real-time features of the system, leveraging on GPU relevant computing power to implement
more selective trigger algorithms.

The KM3 experiment aims at detecting high energy neutrinos through an
underwater Cherenkov telescope with a volume of the order of 1 km³.
In this context, NaNet is charged with two main tasks: first, global
clock and synchronization signals delivery to the off-shore electronic
system; second, reception of underwater devices data through optical
cables. Fundamental requirement for the experiment is having a known
and deterministic latency between on- and off-shore devices.

Results of NaNet performances in both experiments will be reported and
discussed.

Primary authors

Alessandro Lonardo (Universita e INFN, Roma I (IT)) Andrea Biagioni (INFN) Dr Davide Rossetti (NVIDIA Corp) Dr Elena Pastorelli (INFN sezione di Roma) Francesca Lo Cicero (INFN sezione di Roma) Dr Francesco Simula (INFN sezione di Roma) Laura Tosoratto (INFN) Dr Luca Pontisso (INFN) Dr Michele Martinelli (INFN sezione di Roma) Dr Ottorino Frezza (INFN sezione di Roma) Dr Pier Stanislao Paolucci (INFN Sezione di Roma) Piero Vicini (INFN Rome Section) Roberto Ammendola (INFN)

Co-authors

Angelo Cotta Ramusino (Universita di Ferrara (IT)) Dr Fabrizio Ameli (INFN Sezione di Roma) Francesco Simeone (INFN) Gianluca Lamanna (Sezione di Pisa (IT)) Dr Ilaria Neri (Università di Ferrara) Marco Sozzi (Sezione di Pisa (IT)) Massimiliano Fiorini (Universita di Ferrara (IT))

Presentation materials