# 벤치마크

독립적인 TechEmpower 벤치마크에 따르면 **FastAPI** 애플리케이션이 Uvicorn을 사용하여 <a href="https://www.techempower.com/benchmarks/#section=test&runid=7464e520-0dc2-473d-bd34-dbdfd7e85911&hw=ph&test=query&l=zijzen-7" class="external-link" target="_blank">가장 빠른 Python 프레임워크 중 하나</a>로 실행되며, Starlette와 Uvicorn 자체(내부적으로 FastAPI가 사용하는 도구)보다 조금 아래에 위치합니다.

그러나 벤치마크와 비교를 확인할 때 다음 사항을 염두에 두어야 합니다.

## 벤치마크와 속도

벤치마크를 확인할 때, 일반적으로 여러 가지 유형의 도구가 동등한 것으로 비교되는 것을 볼 수 있습니다.

    * extremely easy tests based on HTTPX and `pytest`
    * **CORS**
    * **Cookie Sessions**
    * ...and more.

## Performance

import java.util.TreeSet;
import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.ConcurrentMap;
import java.util.concurrent.ConcurrentSkipListMap;

/**
 * Helper classes for various benchmarks.
 *
 * @author Christopher Swenson
 */
final class BenchmarkHelpers {
  /** So far, this is the best way to test various implementations of {@link Set} subclasses. */
  public interface CollectionsImplEnum {

import java.util.TreeSet;
import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.ConcurrentMap;
import java.util.concurrent.ConcurrentSkipListMap;

/**
 * Helper classes for various benchmarks.
 *
 * @author Christopher Swenson
 */
final class BenchmarkHelpers {
  /** So far, this is the best way to test various implementations of {@link Set} subclasses. */
  public interface CollectionsImplEnum {

 * #offer} and a call to {@link #topK}, with O(k) memory. In comparison, quickselect has the same
 * asymptotics but requires O(n) memory, and a {@code PriorityQueue} implementation takes O(n log
 * k). In benchmarks, this implementation performs at least as well as either implementation, and
 * degrades more gracefully for worst-case input.
 *
 * <p>The implementation does not necessarily use a <i>stable</i> sorting algorithm; when multiple

   * than this method, and this method in turn generally performs better than copying the list and
   * calling {@link Collections#sort(List)}.
   */
  // TODO(kevinb): rerun benchmarks including new options
  public <E extends T> List<E> sortedCopy(Iterable<E> elements) {
    @SuppressWarnings("unchecked") // does not escape, and contains only E's
    E[] array = (E[]) Iterables.toArray(elements);

   * than this method, and this method in turn generally performs better than copying the list and
   * calling {@link Collections#sort(List)}.
   */
  // TODO(kevinb): rerun benchmarks including new options
  public <E extends T> List<E> sortedCopy(Iterable<E> elements) {
    @SuppressWarnings("unchecked") // does not escape, and contains only E's
    E[] array = (E[]) Iterables.toArray(elements);

* Add translation to Portuguese for [Benchmarks - Comparações](https://fastapi.tiangolo.com/pt/benchmarks/). PR [#1274](https://github.com/tiangolo/fastapi/pull/1274) by [@Serrones](https://github.com/Serrones).

# CPU WHEEL
test:linux_cpu_wheel_test_filters --test_tag_filters=-no_oss,-oss_excluded,-oss_serial,-gpu,-tpu,-benchmark-test,-v1only,-no_oss_py38,-no_oss_py39,-no_oss_py310
test:linux_cpu_wheel_test_filters --build_tag_filters=-no_oss,-oss_excluded,-oss_serial,-gpu,-tpu,-benchmark-test,-v1only,-no_oss_py38,-no_oss_py39,-no_oss_py310
test:linux_cpu_wheel_test_filters --test_lang_filters=py --test_size_filters=small,medium

package com.google.common.collect;

import static java.util.Arrays.sort;

import com.google.caliper.BeforeExperiment;
import com.google.caliper.Benchmark;
import com.google.caliper.Param;
import java.util.Arrays;
import java.util.Comparator;
import java.util.Random;

/**
 * A benchmark to determine the overhead of sorting with {@link Ordering#from(Comparator)}, or with