A prototype implementation is running as a search service on an instance of the Physiome Model Repository, which is the backend of the CellML Model Repository (http://staging.physiomeproject.org/).org/
The implementation was run with the parameters summarized in Table 1, leading to the results shown in Figs 4, 5 and 6.
The algorithm was tested on several examples and test parts using the unmodified NC code obtained from the implementation were run on a Moriseiki mill turn center.
In order to make a proper comparison, an unmodified H.264/AVC JM 17.2 reference software encoder implementation was run on the same machine as the proposed algorithm, with the same configuration and with no calls to the GPU.
The one-core, quad-core, and CPU-GPGPU implementations were run and time evaluated for 500 iterations using Rawseeds dataset discussed in Section 5.1.
Correctness refers to how perfectly the software performs and how free it is from faults in terms of design, specification and implementation so that when it is run no errors occurs i.e., the product is considered to have this trait if it conforms to user's specific needs and expectations.
Using the CUDA platform from NVIDIA, we developed a Multiple Species Data Flocking implementation to be run on the NVIDIA GPU.
The tests above show that the new implementation can be run in constrained devices such as Zolertia Z1 motes which have 8-KB RAM and 92-KB ROM.
In order to measure the time spent on each task, a sequential implementation has been run on a single cluster node.
Our procedure pools the information from multiple sequence alignments, and provided software implementation can be run efficiently in parallel on a computer cluster.
