Search for: All records

While parallelism remains the main source of performance,architectural implementations and programming modelschange with each new hardware generation, often leadingto costly application re-engineering. Most tools for perfor-mance portability require manual and costly application port-ing to yet another programming model.We propose an alternative approach that automaticallytranslates programs written in one programming model(CUDA), into another (CPU threads) based on Polygeist/MLIR.Our approach includes a representation of parallel constructsthat allows conventional compiler transformations to ap-ply transparently and without modification a nd enablesparallelism-specific optimizations. We evaluate our frame-work by transpiling and optimizing the CUDA Rodinia bench-mark suite for a multi-core CPU and achieve a 58% geomeanspeedup over handwritten OpenMP code. Further, we showhow CUDA kernels from PyTorch can efficiently run andscale on the CPU-only Supercomputer Fugaku without userintervention. Our PyTorch compatibility layer making use oftranspiled CUDA PyTorch kernels outperforms the PyTorchCPU native backend by 2.7×. 

Traits for prey acquisition form the phenotypic interface of predator–prey interactions. In venomous predators, morphological variation in venom delivery apparatus like fangs and stingers may be optimized for dispatching prey. Here, we determine how a single dimension of venom injection systems evolves in response to variation in the size, climatic conditions and dietary ecology of viperid snakes. We measured fang length in more than 1900 museum specimens representing 199 viper species (55% of recognized species). We find both phylogenetic signal and within-clade variation in relative fang length across vipers suggesting both general taxonomic trends and potential adaptive divergence in fang length. We recover positive evolutionary allometry and little static allometry in fang length. Proportionally longer fangs have evolved in larger species, which may facilitate venom injection in more voluminous prey. Finally, we leverage climatic and diet data to assess the global correlates of fang length. We find that models of fang length evolution are improved through the inclusion of both temperature and diet, particularly the extent to which diets are mammal-heavy diets. These findings demonstrate how adaptive variation can emerge among components of complex prey capture systems.