1. Designing parallel programs is harder than designing sequential programs.
2. We start our design by using the best sequential program we know and then we must solve the following additional problems

• we must coordinate parallel processes to achieve
• noninterference or
• synchronized access
• we must prevent
• deadlock
• livelock
• we must construct correct concurrent programs with
• a well-defined design and construction procedure
• and synchronization derived from global invariants
• we must verify programs with orders of magnitude larger state spaces
• we must distribute processes across multiple processors
• in shared and distributed memory systems
• and balance the workload among processors

a. Application level modeling

• Partial Differential Equations: heat transfer, atmosperic models, soil models, radiation transport
• systems of linear equations: circuit simulation,
• nonlinear of equations: natural gas pipelines
• particle interaction: gravitational pull and molecular dynamics
• recurrence relations: optimization, biological sequence analysis
• etc.

b. Discretization of the Problem (finite difference, finite element, grid simulations, linear algebra, cellular automata, etc.)

c. Parallel Program Design

• Partitioning to find maximal parallelism in model
• Finding paths of Communication between objects in parallel model
• involves defining synchronization constraints: global invariants
• Agglomerating objects to increase granularity for best performance
• Distributing (Mapping) computation among multiple processors for best performance

d. Construction of Correct Concurrent Programs

• using patterns and parallel programming paradigms (if they exist)
• global invariant constraints
• compilation procedures to implement global constraints

e. Verification of Concurrent Software

• building a good test harness
• eliminating deadlock and livelock
• checking all global states for correctness properties