How to double your IOPs
Get the system you're testing. Put artificial workload on it. Measure IOPs.
Get another system exactly the same. Put same workload on the second one. Measure IOPs. Lo and behold you've doubled your IOPs.
Like all synthetic benchmarks, unless they're showing you the exact workload you intend to use on the exact system you intend to buy, they have little value.
Playing Top Trumps with specs is not the right way to go about buying storage. How many people buy their cars based on the manufacturer's claimed top speed?