A CPU differs from a GPU mainly in which way?
Why C is the answer
Fewer cores, tuned for serial work
GPUs have many parallel cores.
In detail
- CPUs have fewer cores, suited to serial work.
- Most systems use both.
Use case
An illustrative scenario, not a real incident.
A CPU runs a database well, with a few strong cores, while a GPU runs thousands of parallel threads.
See it in different ways
flowchart LR A["CPU: few fast cores"] --> B["Serial control work"] C["GPU: many parallel cores"] --> D["Matrix math"]
mindmap
root((AI))
Machine learning
Deep learning
Generative AI
Drivers
GPUs
Data
Algorithms
Stages
Training
Inferencemindmap
root(("Processors"))
n1["CPU"]
n2["Fewer cores"]
n3["Serial work"]
n4["GPU"]
n5["Many cores"]
n6["Parallel work"]Key point: CPUs have fewer cores, tuned for serial work.
Why the other options are wrong
- A. No use of memory while it runs
- They use memory.
- B. No ability to run an operating system
- They run operating systems.
- D. More cores, tuned for parallel work
- That describes GPUs.