Most modern computing systems are heterogeneous, either for organic reasons because components grew independently as it is the case in desktop grids, or by design to leverage the strength of specific hardware as it is the case in accelerated systems. In any case, all computing systems have some form of hardware or software heterogeneity that must been managed, leveraged and understood. HCW is a venue to discuss and innovate in all theoretical and practical aspects of heterogeneous computing: programmability, modeling, design,applications, efficient utilization, algorithms, etc.. Authors are encouraged to submit papers on topics including but not limited to:

Heterogeneous Systems and Architecture. Accelerated systems (GPUs,Xeon Phi, FPGAs,big.LITTLE, ...). Heterogeneous distributed systems (grid, desktop grid, cloud, hybrid clusters) including software heterogeneity. Deep-memory hierarchies (HDD,DRAM, cache, NUMA) and novel explicit memory systems (SSD, NVRAM, 3D stacked memory).

Programming Models and Tools. Code reusability. Performance-abstraction tradeoff. Interoperability of heterogeneous software environments. Middleware and runtime systems. Workflows. Dataflows.

Algorithms for Heterogeneous Parallel System. Parallel algorithms for solving problems on heterogeneous systems. Algorithms for managing heterogeneous resources including allocation and scheduling.

Performance. Modeling, optimizing, improving the time to solve a problem (throughput, latency, runtime), the electric consumption (power, energy) and failure management (fault tolerance, recovery,reliability).

Applications on Heterogeneous System. Case studies. Confluence of Big Data systems and heterogeneous systems. Data-intensive computing. Deep Learning. Scientific computing.

We also recall that heterogeneous systems must consist of several different architectures working together.