NANOBENCH version 2
/usr/bin/node bench.js

# local fs: 1000 * 4K (batch: 1)
ok ~48 ms (0 s + 48091632 ns)

# hyperdrive: 1000 * 4K (batch: 1)
ok ~675 ms (0 s + 675026782 ns)

# client: 1000 * 4K (batch: 1)
ok ~1.31 s (1 s + 306483537 ns)


---

# local fs: 1000 * 4K (batch: 100)
ok ~67 ms (0 s + 66589082 ns)

# hyperdrive: 1000 * 4K (batch: 100)
ok ~488 ms (0 s + 487873347 ns)

# client: 1000 * 4K (batch: 100)
ok ~727 ms (0 s + 727270347 ns)


---

# local fs: 1000 * 4K (batch: 1000)
ok ~79 ms (0 s + 78934663 ns)

# hyperdrive: 1000 * 4K (batch: 1000)
ok ~436 ms (0 s + 435552247 ns)

# client: 1000 * 4K (batch: 1000)
ok ~698 ms (0 s + 697718412 ns)


---

# local fs: 200 * 2M (batch: 1)
ok ~147 ms (0 s + 146825916 ns)

# hyperdrive: 200 * 2M (batch: 1)
ok ~775 ms (0 s + 775041586 ns)

# client: 200 * 2M (batch: 1)
ok ~2.3 s (2 s + 302370689 ns)