This study focuses on the key numerical technique of stencil computations, used in many different scientific disciplines, and illustrates how auto-tuning can be used to produce very efficient implementations across a diverse set of current multicore architectures.

The recent transformation from an environment where gains in computational performance came from increasing clock frequency and other hardware engineering innovations, to an environment where gains are realized through the deployment of ever increasing numbers of modest performance cores has profoundly changed the landscape of scientific application programming. This exponential increase in core count represents both an opportunity and a challenge: access to petascale simulation capabilities and beyond will require that this concurrency be efficiently exploited. The problem for application programmers is further compounded by the diversity of multicore architectures that are now emerging [4]. From relatively complex out-of-order CPUs with complex cache structures, to relatively simple cores that support hardware multithreading, to chips that require explicit use of software controlled memory, designing optimal code for these different platforms represents a serious impediment. An emerging solution to this problem is auto-tuning: the automatic generation of many versions of a code kernel that incorporate various tuning strategies, and the benchmarking of these to select the highest performing version. Typical tuning strategies might include: maximizing incore performance with loop unrolling and restructuring; maximizing memory bandwidth by exploiting non-uniform memory access (NUMA), engaging prefetch by directives; and minimizing memory traffic by cache blocking or array padding. Often a key parameter is associated with each tuning strategy (e.g., the amount of loop unrolling or the cache blocking factor), and these parameters must be explored in addition to the layering of the basic strategies themselves.

The evolution of the traditional nuclear magic numbers away from the valley of stability is an active field of research. Experimental efforts focus on providing key spectroscopic information that will shed light into the structure of exotic nuclei and understanding the driving mechanism behind the shell evolution. In this work, we investigate the Z=6 spin-orbit shell gap towards the neutron dripline. To do so, we employed NA(p,2p)CA−1 quasi-free scattering reactions to measure the proton component of the 21+ state of 16,18,20C. The experimental findings support the notion of a moderate reduction of the proton 1p1/2−1p3/2 spin-orbit splitting, at variance to recent claims for a prevalent Z=6 magic number towards the neutron dripline.

Associated production of bottomonia and open charm hadrons in pp collisions at s=7 and 8 TeV is observed using data corresponding to an integrated luminosity of 3 fb−1 accumulated with the LHCb detector. The observation of five combinations, Y(1S)D0, Y(2S)D0, Y(1S)D+, Y(2S)D+ and Y(1S)Ds+, is reported. Production crosssections are measured for Y(1S)D0 and Y(1S)D+ pairs in the forward region. The measured cross-sections and the differential distributions indicate the dominance of double parton scattering as the main production mechanism.[Figure not available: see fulltext.]

The difference between the CP asymmetries in the decays Λc+ → pK−K+ and Λc+ → pπ−π+ is presented. Proton-proton collision data taken at centre-of-mass energies of 7 and 8 TeV collected by the LHCb detector in 2011 and 2012 are used, corresponding to an integrated luminosity of 3 fb−1. The Λc+ candidates are reconstructed as part of the Λb0 → Λc+μ−X decay chain. In order to maximize the cancellation of production and detection asymmetries in the difference, the final-state kinematic distributions of the two samples are aligned by applying phase-space-dependent weights to the Λc+ → pπ−π+ sample. This alters the definition of the integrated CP asymmetry to ACPwgt(pπ−π+). Both samples are corrected for reconstruction and selection efficiencies across the five-dimensional Λc+ decay phase space. The difference in CP asymmetries is found to be ΔACPwgt=ACP(pK−K+)−ACPwgt(pπ−π+)=(0.30±0.91±0.61)%, where the first uncertainty is statistical and the second is systematic.[Figure not available: see fulltext.].

A search for the non-resonant decays Bs0 → μ+μ−μ+μ− and B0 → μ+μ−μ+μ− is presented. The measurement is performed using the full Run 1 data set collected in proton-proton collisions by the LHCb experiment at the LHC. The data correspond to integrated luminosities of 1 and 2 fb−1 collected at centre-of-mass energies of 7 and 8 TeV, respectively. No signal is observed and upper limits on the branching fractions of the non-resonant decays at 95% confidence level are determined to beℬ(Bs0→μ+μ−μ+μ−)<2.5×10−9,[Figure not available: see fulltext.]

The first observation of the B_{s}^{0}→D[over ¯]^{0}K_{S}^{0} decay mode and evidence for the B_{s}^{0}→D[over ¯]^{*0}K_{S}^{0} decay mode are reported. The data sample corresponds to an integrated luminosity of 3.0  fb^{-1} collected in pp collisions by LHCb at center-of-mass energies of 7 and 8 TeV. The branching fractions are measured to be B(B_{s}^{0}→D[over ¯]^{0}K[over ¯]^{0})=[4.3±0.5(stat)±0.3(syst)±0.3(frag)±0.6(norm)]×10^{-4},B(B_{s}^{0}→D[over ¯]^{*0}K[over ¯]^{0})=[2.8±1.0(stat)±0.3(syst)±0.2(frag)±0.4(norm)]×10^{-4},where the uncertainties are due to contributions coming from statistical precision, systematic effects, and the precision of two external inputs, the ratio f_{s}/f_{d} and the branching fraction of B^{0}→D[over ¯]^{0}K_{S}^{0}, which is used as a calibration channel.