Skip to content

Commit

Permalink
Add updated benchmark results.
Browse files Browse the repository at this point in the history
Co-authored-by: David Cooke <[email protected]>
  • Loading branch information
iziamos and davidcooke2 committed Sep 13, 2024
1 parent bfc35ee commit 9b173ee
Showing 1 changed file with 38 additions and 1 deletion.
39 changes: 38 additions & 1 deletion src/docs/asciidoc/en/disruptor.adoc
Original file line number Diff line number Diff line change
Expand Up @@ -472,6 +472,43 @@ h| Disruptor
| 15,295,197
|===

.Comparative throughput updated for modern hardware (in ops per sec)
[cols=4*,options="header"]
|===
|
3+| AMD EPYC 9374F – Linux 5.4.277 – OpenJDK 11.0.24

|
h| ABQ
h| Disruptor 3
h| Disruptor 4

| Unicast: 1P – 1C
| 20,895,148
| 134,553,283
| 160,359,204

| Pipeline: 1P – 3C
| 5,216,647
| 76,068,766
| 101,317,122

| Sequencer: 3P – 1C
| 18,791,340
| 16,010,759
| 29,726,516

| Multicast: 1P – 3C
| 2,355,379
| 68,157,033
| 70,018,204

| Diamond: 1P – 3C
| 3,433,665
| 61,229,488
| 63,123,343
|===

== Latency Performance Testing
To measure latency we take the three stage pipeline and generate events at less than saturation.
This is achieved by waiting 1 microsecond after injecting an event before injecting the next and repeating 50 million times.
Expand Down Expand Up @@ -523,4 +560,4 @@ As load increases on the Disruptor, latency remains almost flat until saturation

We believe that the Disruptor establishes a new benchmark for high-performance computing and is very well placed to continue to take advantage of current trends in processor and computer design.

View the original PDF of this paper link:./files/Disruptor-1.0.pdf[here].
View the original PDF of this paper link:./files/Disruptor-1.0.pdf[here].

0 comments on commit 9b173ee

Please sign in to comment.