Small followup to #526

[christiangnrd]

Also snuck in a readme update

[github-actions]

Metal Benchmarks

Benchmark suite	Current: 18fc6bb	Previous: 4293d1a	Ratio
private array/construct	24114.583333333336 ns	25972.166666666668 ns	0.93
private array/broadcast	462167 ns	458250 ns	1.01
private array/random/randn/Float32	790375 ns	850875 ns	0.93
private array/random/randn!/Float32	636375 ns	624291 ns	1.02
private array/random/rand!/Int64	566916 ns	571416 ns	0.99
private array/random/rand!/Float32	597333 ns	603125 ns	0.99
private array/random/rand/Int64	758083 ns	817916 ns	0.93
private array/random/rand/Float32	609375 ns	610458 ns	1.00
private array/copyto!/gpu_to_gpu	652125 ns	655833 ns	0.99
private array/copyto!/cpu_to_gpu	815125 ns	649166.5 ns	1.26
private array/copyto!/gpu_to_cpu	643250 ns	809667 ns	0.79
private array/accumulate/1d	1348396.5 ns	1338041.5 ns	1.01
private array/accumulate/2d	1389083 ns	1392583.5 ns	1.00
private array/iteration/findall/int	2111041.5 ns	2057458 ns	1.03
private array/iteration/findall/bool	1840166 ns	1836584 ns	1.00
private array/iteration/findfirst/int	1695750 ns	1702979 ns	1.00
private array/iteration/findfirst/bool	1666583 ns	1663000 ns	1.00
private array/iteration/scalar	3639250 ns	3578667 ns	1.02
private array/iteration/logical	3199292 ns	3171792 ns	1.01
private array/iteration/findmin/1d	1758458 ns	1751083.5 ns	1.00
private array/iteration/findmin/2d	1345875 ns	1354375 ns	0.99
private array/reductions/reduce/1d	1028354 ns	1041104 ns	0.99
private array/reductions/reduce/2d	662416.5 ns	658917 ns	1.01
private array/reductions/mapreduce/1d	1050958 ns	1035500 ns	1.01
private array/reductions/mapreduce/2d	658083 ns	664292 ns	0.99
private array/permutedims/4d	2535791 ns	2505042 ns	1.01
private array/permutedims/2d	1019417 ns	1027458 ns	0.99
private array/permutedims/3d	1593042 ns	1591062 ns	1.00
private array/copy	560167 ns	581625 ns	0.96
latency/precompile	8809563167 ns	8819859791 ns	1.00
latency/ttfp	3609854000 ns	3601196084 ns	1.00
latency/import	1230788375 ns	1233148709 ns	1.00
integration/metaldevrt	697708 ns	709292 ns	0.98
integration/byval/slices=1	1560792 ns	1635521 ns	0.95
integration/byval/slices=3	9990042 ns	9780625 ns	1.02
integration/byval/reference	1606584 ns	1577791 ns	1.02
integration/byval/slices=2	2607312.5 ns	2702000 ns	0.96
kernel/indexing	450459 ns	448583.5 ns	1.00
kernel/indexing_checked	447125 ns	457708 ns	0.98
kernel/launch	8916 ns	7833 ns	1.14
metal/synchronization/stream	14875 ns	14458 ns	1.03
metal/synchronization/context	15500 ns	14792 ns	1.05
shared array/construct	23913.25 ns	24288.166666666668 ns	0.98
shared array/broadcast	461917 ns	455854.5 ns	1.01
shared array/random/randn/Float32	777229 ns	808437 ns	0.96
shared array/random/randn!/Float32	639792 ns	628459 ns	1.02
shared array/random/rand!/Int64	569250 ns	570584 ns	1.00
shared array/random/rand!/Float32	604667 ns	593687 ns	1.02
shared array/random/rand/Int64	754917 ns	745583 ns	1.01
shared array/random/rand/Float32	593416.5 ns	631125 ns	0.94
shared array/copyto!/gpu_to_gpu	82792 ns	83000 ns	1.00
shared array/copyto!/cpu_to_gpu	83792 ns	84375 ns	0.99
shared array/copyto!/gpu_to_cpu	83208 ns	82166 ns	1.01
shared array/accumulate/1d	1363125 ns	1346521 ns	1.01
shared array/accumulate/2d	1394666 ns	1393625.5 ns	1.00
shared array/iteration/findall/int	1856187.5 ns	1781500 ns	1.04
shared array/iteration/findall/bool	1609375 ns	1581729 ns	1.02
shared array/iteration/findfirst/int	1398395.5 ns	1395792 ns	1.00
shared array/iteration/findfirst/bool	1362125 ns	1358000 ns	1.00
shared array/iteration/scalar	158917 ns	154000 ns	1.03
shared array/iteration/logical	2995750 ns	2965083.5 ns	1.01
shared array/iteration/findmin/1d	1464625 ns	1466813 ns	1.00
shared array/iteration/findmin/2d	1365667 ns	1361937.5 ns	1.00
shared array/reductions/reduce/1d	728042 ns	731000 ns	1.00
shared array/reductions/reduce/2d	662542 ns	661125 ns	1.00
shared array/reductions/mapreduce/1d	748167 ns	733459 ns	1.02
shared array/reductions/mapreduce/2d	665833 ns	664875 ns	1.00
shared array/permutedims/4d	2532771 ns	2519729 ns	1.01
shared array/permutedims/2d	1025062.5 ns	1019229.5 ns	1.01
shared array/permutedims/3d	1582479 ns	1610792 ns	0.98
shared array/copy	239479.5 ns	239542 ns	1.00

This comment was automatically generated by workflow using github-action-benchmark.