AS symbolize_keys vs Hash.keys.map.to_sym

/bin/bash -c "/home/thunder/.rvm/bin/rvm ruby-2.5.1 do /home/thunder/.rvm/rubies/ruby-2.5.1/bin/ruby /home/thunder/.RubyMine2019.3/config/scratches/scratch_1.rb"
Warming up --------------------------------------
Active Support symbolize_keys
                        37.892k i/100ms
Hash keys map to_sym    60.148k i/100ms
Calculating -------------------------------------
Active Support symbolize_keys
                        562.840k (± 1.4%) i/s -      5.608M in  10.047830s
Hash keys map to_sym    767.850k (± 1.1%) i/s -      7.699M in  10.048546s
                   with 95.0% confidence

Comparison:
Hash keys map to_sym:            767850.4 i/s
Active Support symbolize_keys:   562839.7 i/s - 1.36x  (± 0.02) slower
                   with 95.0% confidence

Code:
_________

# frozen_string_literal: true

# gem install benchmark-ips
# gem install kalibera
# gem install active_support

require 'benchmark/ips'
require 'active_support'
require 'active_support/core_ext/hash'

# Enable and start GC before each job run. Disable GC afterwards.
#
# Inspired by https://www.omniref.com/ruby/2.2.1/symbols/Benchmark/bm?#annotation=4095926&line=182
class GCSuite
  def warming(*)
    run_gc
  end

  def running(*)
    run_gc
  end

  def warmup_stats(*)
  end

  def add_report(*)
  end

  private

  def run_gc
    GC.enable
    GC.start
    GC.disable
  end
end

suite = GCSuite.new

h = { 'a' => 1, b: 2}

Benchmark.ips do |x|
  x.config(:time => 10, :warmup => 5, :stats => :bootstrap, :confidence => 95)

  x.report('Active Support symbolize_keys') do
    h.symbolize_keys
  end

  x.report('Hash keys map to_sym') do
    h.keys.map { |k| k.is_a?(Symbol) ? k : k.to_sym  }
  end

  x.compare!
end