Kuna tofauti gani kati ya concurrency na parallelism?

Question

IT Interview Practice · Accepted Answer

**Concurrency inahusu *kushughulikia* mambo mengi kwa wakati mmoja; parallelism inahusu *kufanya* mambo mengi kwa wakati mmoja.** Concurrency ni njia ya kupanga programu ili tasks nyingi zipige hatua kwa kuingiliana (interleaving); parallelism ni kuzitekeleza kwa kweli kwa wakati mmoja kwenye cores nyingi.

## Jinsi inavyofanya kazi

CPU core moja inaweza kuwa **concurrent** lakini siyo parallel: inabadilisha kati ya tasks kwa kasi, hivyo zinaingiliana kwa muda ingawa moja tu inaendeshwa kwa kila papo hapo. **Parallelism** inahitaji cores nyingi zikiendesha instructions kwa wakati mmoja.

```text
Concurrency (1 core, interleaved):   A--B--A--B--A--B
Parallelism (2 cores, simultaneous): A--A--A--A
                                     B--B--B--B
```

## Mfano

```python
# Concurrency: one thread juggles two downloads by waiting on I/O
async def main():
    await asyncio.gather(fetch(url1), fetch(url2))  # overlap, one core

# Parallelism: two CPU cores crunch numbers at the same time
with ProcessPoolExecutor() as pool:
    pool.map(heavy_compute, data)  # truly simultaneous
```

Muhtasari wa Rob Pike: "Concurrency is about structure, parallelism is about execution." Unaweza kuwa na concurrency bila parallelism (async I/O kwenye core moja) na mara nyingi zinaunganika.

## Kwa nini ni muhimu

Wahoji hutumia hili kuangalia kama unaelewa *kwa nini* threads husaidia. Kazi ya I/O-bound hunufaika na concurrency (kuingiliana kwa kusubiri); kazi ya CPU-bound inahitaji parallelism (cores zaidi). Kuzichanganya kunapelekea designs zinazoongeza threads lakini hazioni ongezeko la kasi.

Jinsi inavyofanya kazi

Mfano

Kwa nini ni muhimu

Maswali yanayohusiana