ERIS is the new AO instrument for VLT-UT4 led by a Consortium of Max-Planck Institut fuer Extraterrestrische Physik, UK-ATC, ETH-Zurich, NOVA-Leiden, ESO and INAF. The ERIS AO system provides NGS mode to deliver high contrast correction and LGS mode to extend high Strehl performance to large sky coverage. The AO module includes NGS and LGS wavefront sensors and, with VLT-AOF Deformable Secondary Mirror and Laser Facility, will provide AO correction to the high resolution coronagraphic imager NIX (1-5um) and the IFU spectrograph SPIFFIER (1-2.5um). In this paper, we present the final design of the ERIS AO system and the status of the of current MAIV phase.
The Enhanced Resolution Imager and Spectrograph (ERIS) is a next-generation, adaptive optics assisted, near-IR imager and integral field spectrograph (IFS) for the Cassegrain focus of the Very Large Telescope (VLT) Unit Telescope 4. It will make use of the Adaptive Optics Facility (AOF), comprising the Deformable Secondary Mirror (DSM) and the UT4 Laser Guide Star Facility (4LGSF). It is a rather complex instrument, with its state of the art AO system and two science channels. It is also meant to be a "workhorse" instrument and offers many observation modes. ERIS is being built by a Consortium of European Institutes comprising MPE Garching (D), ATC (UK), ETH Zürich (CH), Leiden University (NL) and INAF (I) in collaboration with ESO. The instrument passed Final Design Review in mid-2017 and is now in the MAIT phase. In this paper we describe the design of the ERIS Instrument Software (INS), which is in charge of controlling all instrument functions and implementing observation, calibration and maintenance procedures. The complexity of the instrument is reflected in the architecture of its control software and the number of templates required for operations. After a brief overview of the Instrument, we describe the general architecture of the ERIS control network and software. We then discuss some of the most interesting aspects of ERIS INS, like the wavefront sensors function control, AO secondary loops, IFS quick-look processing and the on-line processing for high-contrast imaging observations. Finally, we provide some information about our development process, including software quality assurance activities.
Adaptive optics is essential for the successful operation of the future Extremely Large Telescopes (ELTs). At the heart of these AO system lies the real-time control which has become computationally challenging. A majority of the previous efforts has been aimed at reducing the wavefront reconstruction latency by using many-core hardware accelerators such as Xeon Phis and GPUs. These modern hardware solutions offer a large numbers of cores combined with high memory bandwidths but have restrictive input/output (I/O). The lack of efficient I/O capability makes the data handling very inefficient and adds both to the overall latency and jitter. For example a single wavefront sensor for an ELT scale adaptive optics system can produce hundreds of millions of pixels per second that need to be processed. Passing all this data through a CPU and into GPUs or Xeon Phis, even by reducing memory copies by using systems such as GPUDirect, is highly inefficient.
The Mellanox TILE series is a novel technology offering a high number of cores and multiple 10 Gbps Ethernet ports. We present results of the TILE-Gx36 as a front-end wavefront sensor processing unit. In doing so we are able to greatly reduce the amount of data needed to be transferred to the wavefront reconstruction hardware. We show that the performance of the Mellanox TILE-GX36 is in-line with typical requirements, in terms of mean calculation time and acceptable jitter, for E-ELT first-light instruments and that the Mellanox TILE series is a serious contender for all E-ELT instruments.
We present wavefront reconstruction acceleration of high-order AO systems using an Intel Xeon Phi processor. The Xeon Phi is a coprocessor providing many integrated cores and designed for accelerating compute intensive, numerical codes. Unlike other accelerator technologies, it allows virtually unchanged C/C++ to be recompiled to run on the Xeon Phi, giving the potential of making development, upgrade and maintenance faster and less complex. We benchmark the Xeon Phi in the context of AO real-time control by running a matrix vector multiply (MVM) algorithm. We investigate variability in execution time and demonstrate a substantial speed-up in loop frequency. We examine the integration of a Xeon Phi into an existing RTC system and show that performance improvements can be achieved with limited development effort.