A company has an application workload that is stateless by design and can sustain occasional downtime. The application performs massively parallel computations.
Which Amazon EC2 pricing model should the company choose for its application to reduce cost?