Ben Kelly [:bkelly, not reviewing] Reporter
	Description • 9 years ago

While benchmarking some "promise vs no-promise" spec questions for streams API, I wrote a small benchmark tool:

  https://github.com/wanderview/streams-promise-read

The index.html runs with native promises while bluebird.html uses the bluebird promise library.

When you increase the number of chunks in the test high enough, you can see some non-trivial differences in speed between our native promises and bluebird.

Native
------
1000 chunks:
"sync x 119 ops/sec ±27.87% (23 runs sampled)" test.js:14:3
"promise x 26.71 ops/sec ±2.97% (60 runs sampled)" test.js:14:3

10000 chunks:
"sync x 35.56 ops/sec ±20.20% (21 runs sampled)" test.js:14:3
"promise x 2.80 ops/sec ±8.60% (17 runs sampled)" test.js:14:3

Bluebird:
---------
1000 chunks:
"sync x 146 ops/sec ±16.29% (36 runs sampled)" test.js:14:3
"promise x 96.67 ops/sec ±4.96% (68 runs sampled)" test.js:14:3

10000 chunks:
"sync x 40.54 ops/sec ±21.98% (32 runs sampled)" test.js:14:3
"promise x 21.55 ops/sec ±5.12% (53 runs sampled)" test.js:14:3

Only 2.8 ops/sec for native promises vs 21.5 ops/sec for bluebird.  Thats a pretty large performance difference.

Note, you must checkout the test and tweak the values to test 10,000 chunks.

	David Bruant
	Comment 1 • 6 years ago

Running the benchmark, i get :

https://blog.wanderview.com/streams-promise-read/bluebird.html?chunks=10000&size=1
---------------
Testing 10000 chunks of 1 bytes per operation.
promise x 2.31 ops/sec ±8.30% (16 runs sampled)
Fastest is promise

https://blog.wanderview.com/streams-promise-read/native.html?chunks=10000&size=1
---------------
Testing 10000 chunks of 1 bytes per operation.
promise x 10.79 ops/sec ±3.12% (49 runs sampled)
Fastest is promise

I feel this means native promises have gotten better
RESOLVED WORKFORME

	David Bruant
	Comment 2 • 6 years ago

Sorry, i meant this to be a question

> RESOLVED WORKFORME

?

	Olli Pettay [:smaug][bugs@pettay.fi]
	Comment 3 • 6 years ago

Yeah, I get similar results. Promise scheduling uses now microtasks and IIRC we also optimize out unused tail Promises, or what to call them.