process** (e.g. a Uvicorn process running your FastAPI application). They would all be **identical containers**, running the same thing, but each with its own process, memory, etc. That way you would take advantage of **parallelization** in **different cores** of the CPU, or even in **different machines**.

And the distributed container system with the **load balancer** would **distribute the requests** to each one of the containers with your app **in turns**. So, each request could be handled...

		for i := range cpus {
			cpuMap[cpus[i].PhysicalID] = struct{}{}
			coreMap[cpus[i].CoreID] = struct{}{}
		}
		cpu := cpus[0]
		uaAppend(" CPU ", fmt.Sprintf("(total_cpus:%d, total_cores:%d; vendor:%s; family:%s; model:%s; stepping:%d; model_name:%s)",
			len(cpuMap), len(coreMap), cpu.VendorID, cpu.Family, cpu.Model, cpu.Stepping, cpu.ModelName))
	}
	uaAppend(")", "")

### Multiple Processes - Workers

If you have more clients than what a single process can handle (for example if the virtual machine is not too big) and you have **multiple cores** in the server's CPU, then you could have **multiple processes** running with the same application at the same time, and distribute all the requests among them.

    *   Converts hardswish subgraphs into atomic ops.
*   Add support for defaulting the value of `cycle_length` argument of
    `tf.data.Dataset.interleave` to the number of schedulable CPU cores.
*   `parallel_for`: Add converter for `MatrixDiag`.
*   Add `narrow_range` attribute to `QuantizeAndDequantizeV2` and V3.
*   Added new op: `tf.strings.unsorted_segment_join`.