sounds good. honestly, i don't know what CU means when you say 60 CUs? but it does sounds good though. hope you can shed some lights to what CU is.
The meaning of a compute unit(CU) depends on who manufactured a particular GPU, but as we know the top two GPU vendors are NVIDIA and AMD: A
compute unit is a stream multiprocessor in a NVidia GPU or a SIMD engine in an AMD GPU. Each compute unit has several processing elements (ALU/stream processor). For example, A compute unit of a HD 5000 series GPU has 80 processing elements (16 processing cores with 5 ALUs per processing core).
A compute unit (CU) on a GPU is a processing unit that performs computations on data in parallel. GPUs have many compute units, and they are the basic building blocks of the GPU's architecture.
Compute units typically consist of multiple processing elements (PEs), which can perform a variety of arithmetic and logical operations on data. These PEs are designed to work together in parallel, allowing the compute unit to perform many computations simultaneously.
Compute units are important for parallel processing tasks, such as those found in graphics rendering and scientific computing. By dividing the workload among multiple compute units, the GPU can perform many computations in parallel, which can result in significant speedups compared to serial processing on a CPU.
The number of compute units on a GPU can vary depending on the specific GPU model, and it is an important factor in determining the GPU's overall processing power. More compute units generally mean higher performance, although other factors such as clock speed, memory bandwidth, and memory capacity can also play a role in determining overall performance.
In summary, a compute unit on a GPU is a processing unit that performs computations on data in parallel, consisting of multiple processing elements working together to achieve high levels of parallelism and performance.
