sjvanrossum commented on PR #29633:
URL: https://github.com/apache/beam/pull/29633#issuecomment-1843130690

   Update with results from a n2d-standard-48 instance on GCP.
   Much to my amusement, there's a slight drop in performance on 2-byte writes 
on this platform.
   I'll run this again with exact length array construction, that benched a 
higher on the low end, but lower on the high end.
   
   ```
   Benchmark                                      Mode  Cnt        Score       
Error  Units
   VarIntBenchmark.benchmarkNewVarIntAll         thrpt    5  1016584.280 ± 
30523.261  ops/s
   VarIntBenchmark.benchmarkNewVarIntExtraLarge  thrpt    5   476104.305 ±  
8071.387  ops/s
   VarIntBenchmark.benchmarkNewVarIntLarge       thrpt    5   689945.440 ± 
19918.354  ops/s
   VarIntBenchmark.benchmarkNewVarIntMedium      thrpt    5  1000459.069 ± 
17035.831  ops/s
   VarIntBenchmark.benchmarkNewVarIntSmall       thrpt    5  2810764.124 ± 
70178.225  ops/s
   VarIntBenchmark.benchmarkOldVarIntAll         thrpt    5  1017242.096 ± 
16247.798  ops/s
   VarIntBenchmark.benchmarkOldVarIntExtraLarge  thrpt    5   271332.667 ±  
4024.960  ops/s
   VarIntBenchmark.benchmarkOldVarIntLarge       thrpt    5   629245.307 ± 
12227.888  ops/s
   VarIntBenchmark.benchmarkOldVarIntMedium      thrpt    5  1251259.777 ± 
24432.231  ops/s
   VarIntBenchmark.benchmarkOldVarIntSmall       thrpt    5  2810652.462 ± 
66371.139  ops/s
   ```


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to