SIMT: The Same Instruction, Many Threads
The execution model that makes GPUs fast.
How the GPU Stays Busy
Thousands of cores need a smart way to be told what to do. The GPU's answer is SIMT: Single Instruction, Multiple Threads. ⚡
One Instruction, Many Threads
In SIMT, one instruction is broadcast to a whole group of threads at once. Each thread runs that same step, but on its own piece of data.
All lessons in this course
- CPU vs GPU: Latency vs Throughput
- SIMT: The Same Instruction, Many Threads
- What CUDA Actually Is
- Problems That Love the GPU