#tiles 
go get github.com/buckhx/tiles
The godoc has good documentation and a couple examples, but here are a few as well
// Empire State Build coordinates and zoom level 18
z := 18
lat := 40.7484
lon := -73.9857
// Basic usage
t1 := tiles.FromCoordinate(lat, lon, z)
// If you want a little more granularity
x := tiles.Coordinate{Lat: lat, Lon: lon}
p := x.ToPixel(z)
t2, _ := p.ToTile()
// t1 == t2
The TileIndex allows for data to be indexed by tiles and aggregated up to their parents when requested
idx := tiles.NewTileIndex()
esb := tiles.FromCoordinate(40.7484, -73.9857, 18)
sol := tiles.FromCoordinate(40.6892, -74.0445, 18)
bbn := tiles.FromCoordinate(51.5007, -0.1246, 18)
idx.Add(esb, "EmpireStateBuilding")
idx.Add(sol, "StatueOfLiberty")
idx.Add(bbn, "BigBen")
nyc := tiles.Tile{X: 75, Y: 96, Z: 8}
den := tiles.Tile{X: 106, Y: 194, Z: 9}
fmt.Println("ESB Tile: ", idx.Values(esb))
fmt.Println("SOL Tile: ", idx.Values(sol))
fmt.Println("NYC Tile: ", idx.Values(nyc)) //contains esb & sol values!
fmt.Println("DENVER Tile: ", idx.Values(den)) //contains no values!
Here are some microbenchmarks for converting a location at zoom level 18. There's nothing really to compare them to, but should give a sense of op time on a 2.3 GHz core i7 MBP.
$ go test -bench=. -benchmem -benchtime 10s
BenchmarkTileFromCoordinate-8 100000000 165 ns/op 0 B/op 0 allocs/op
BenchmarkTileFromQuadkey-8 300000000 43.7 ns/op 0 B/op 0 allocs/op
BenchmarkQuadkeyFromTile-8 200000000 95.4 ns/op 32 B/op 1 allocs/op