Here are the results I obtained by running a low latency test and a high throughput test on my NV1616 production model card:
------------------------------------------------------------------------------------------------------------------------------------
Test 1 - Low Latency Test
fio -filename=/dev/nvme0n1 --direct=1 --rw=randrw --refill_buffers --norandommap --randrepeat=0 --ioengine=libaio --time_based --runtime=30 --group_reporting --name=latency --rwmixread=0 --rwmixwrite=100 --iodepth=1 --numjobs=1 --bs=512B
latency: (g=0): rw=randrw, bs=(R) 512B-512B, (W) 512B-512B, (T) 512B-512B, ioengine=libaio, iodepth=1
Starting 1 process
latency: (groupid=0, jobs=1): err= 0: pid=3689: Wed Dec 18 22:05:19 2019
write: IOPS=52.6k, BW=25.7MiB/s (26.9MB/s)(770MiB/30001msec)
lat (nsec): min=15648, max=113504, avg=17479.39, stdev=708.28
------------------------------------------------------------------------------------------------------------------------------------
Test 2 - High Throughput Test
fio --filename=/dev/nvme0n1 --direct=1 --rw=randrw --refill_buffers --norandommap --randrepeat=0 --ioengine=libaio --time_based --runtime=30 --group_reporting --name=throughput --rwmixread=0 --rwmixwrite=100 --iodepth=8 --numjobs=32 --bs=16K
throughput: (groupid=0, jobs=32): err= 0: pid=3756: Wed Dec 18 22:36:57 2019
write: IOPS=326k, BW=5100MiB/s (5347MB/s)(149GiB/30008msec)
lat (usec): min=26, max=9951, avg=777.73, stdev=264.92
--------------------------------------------------------------------------------------------------------------------------------------
Note: The results I got on my system are very close to testing results on the NVRAM Block Mode Performance App Note. Here are their results:
Low latency example:
Bandwidth: 23.0MB/s
IOPs: 46039
Latency: 15.93us
High throughput example:
Bandwidth: 5118.2MiB/s
IOPs: 327563
Latency: 773.84us