The maximum entropy method (MEM) is a well known deconvolution technique in radio-interferometry. This method solves a non-linear optimization problem with an entropy regularization term. Other heuristics such as CLEAN are faster but highly user dependent. Nevertheless, MEM has the following advantages: it is unsupervised, it has a statistical basis, it has a better resolution and better image quality under certain conditions. This work presents a high performance GPU version of non-gridding MEM, which is tested using real and simulated data. We propose a single-GPU and a multi-GPU implementation for single and multi-spectral data, respectively. We also make use of the Peer-to-Peer and Unified Virtual Addressing features of newer GPUs which allows to exploit transparently and efficiently multiple GPUs. Several ALMA data sets are used to demonstrate the effectiveness in imaging and to evaluate GPU performance. The results show that a speedup from 1000 to 5000 times faster than a sequential version can be achieved, depending on data and image size. This allows to reconstruct the HD142527 CO(6-5) short baseline data set in 2.1 minutes, instead of 2.5 days that takes a sequential version on CPU.

We present ALMA (Cycle 0) band-6 and band-3 observations of the transition disk Sz\,91. The disk inclination and position angle are determined to be $i=49.5\degr\pm3.5\degr$ and $\mathrm{PA}=18.2\degr\pm3.5\degr$ and the dusty and gaseous disk are detected up to $\sim220$ au and $\sim400$ au from the star, respectively. Most importantly, our continuum observations indicate that the cavity size in the mm-sized dust distribution must be $\sim97$ au in radius, the largest cavity observed around a T Tauri star. Our data clearly confirms the presence of \co(2-1) well inside the dust cavity. Based on these observational constrains we developed a disk model that simultaneously accounts for the \co and continuum observations (i.e., gaseous and dusty disk). According to our model, most of the millimeter emission comes from a ring located between 97 and 140 au. We also find that the dust cavity is divided into an innermost region largely depleted of dust particles ranging from the dust sublimation radius up to 85 au, and a second, moderately dust-depleted region, extending from 85 to 97 au. The extremely large size of the dust cavity, the presence of gas and small dust particles within the cavity and the accretion rate of Sz\,91 are consistent with the formation of multiple (giant) planets.

Gaseous giant planet formation is thought to occur in the first few million years following stellar birth. Models predict that giant planet formation carves a deep gap in the dust component (shallower in the gas). Infrared observations of the disk around the young star HD142527, at ~140pc, found an inner disk ~10AU in radius, surrounded by a particularly large gap, with a disrupted outer disk beyond 140AU, indicative of a perturbing planetary-mass body at ~90 AU. From radio observations, the bulk mass is molecular and lies in the outer disk, whose continuum emission has a horseshoe morphology. The vigorous stellar accretion rate would deplete the inner disk in less than a year, so in order to sustain the observed accretion, matter must flow from the outer-disk into the cavity and cross the gap. In dynamical models, the putative protoplanets channel outer-disk material into gap-crossing bridges that feed stellar accretion through the inner disk. Here we report observations with the Atacama Large Millimetre Array (ALMA) that reveal diffuse CO gas inside the gap, with denser HCO+ gas along gap-crossing filaments, and that confirm the horseshoe morphology of the outer disk. The estimated flow rate of the gas is in the range 7E-9 to 2E-7 Msun/yr, which is sufficient to maintain accretion onto the star at the present rate.