parallelism
a clear similarity between two things
a noticeable similarity or correspondence between two or more things, especially in their structure, development, or circumstances
using the same grammar pattern in several parts of a sentence or text
the use of the same or a very similar grammatical pattern in related words, phrases, clauses, or lines, especially to create balance, rhythm, or emphasis
when a computer does several pieces of work at the same time
the use or presence of simultaneous execution of multiple operations or tasks in a computer system, usually to improve speed or performance
the condition of two lines or surfaces being parallel
the property or state of two or more lines, surfaces, or planes being parallel: staying the same distance apart and not meeting
- Base: parallelism
- Plural: parallelism, parallelisms
data parallelism
1same operation on data parts
doing the same computer operation on many pieces of data at once
- The image-processing code uses data parallelism, applying the same filter to many pixels at once.
- GPUs are well suited to data parallelism because they can run one operation across large arrays.
- Data parallelism helps the program process many records at the same time.
instruction-level parallelism
1CPU instructions overlap
when a CPU handles several small instructions at the same time
- Modern processors use instruction-level parallelism to overlap simple operations inside the CPU.
- The compiler can sometimes expose more instruction-level parallelism by reordering instructions safely.
- A CPU can use instruction-level parallelism to do small steps at the same time.
massive parallelism
1large-scale simultaneous processing
using a very large number of parallel operations at once
- Modern neural-network training depends on massive parallelism across thousands of GPU cores.
- DNA sequencing platforms achieve massive parallelism by reading millions of fragments at once.
- Massive parallelism can make very large calculations faster.
task parallelism
1different tasks run together
when a computer runs different jobs at the same time
- The server uses task parallelism: one thread reads requests while another writes results to the database.
- Task parallelism is useful when a program has several independent jobs to do.
- Task parallelism lets two different jobs run at once.