Link to SS3 benchmarks

README.md

@@ -2,11 +2,11 @@
 
 Two modern laptops, Lenovo P15 Gen 1 (i7-10875H) and Dell Precision 3560
 (i7-1165G7), each with a number of VMs. ADMB built from the current
-[admb-13.0](https://github.com/admb-project/admb/tree/admb-13.0) branch
-(8820af06), compiling models using `admb -f`. The tests (catage and vol) are
-from
-[admb/examples](https://github.com/admb-project/admb/tree/main/examples/admb):
-`catage -mcmc 1e6 -mcsave 1e3` and `nm2vol`.
+[admb-13.0](https://github.com/admb-project/admb/tree/admb-13.0) branch (commit
+8820af06) with `make -j8`. The
+[example](https://github.com/admb-project/admb/tree/main/examples/admb) models
+(catage and vol) are built using `admb -f` and then run as `catage -mcmc 1e6
+-mcsave 1e3` and `nm2vol`.
 
 The [analysis](analysis.R) of the [results](results.csv) indicates that for
 these examples, ADMB 13.0-beta is consistently and considerably slower than ADMB
@@ -17,3 +17,10 @@ One would hope that ADMB 13.0 will be equally fast as previous versions - or at
 least not much slower.
 
 This seems worth looking into before releasing 13.0.
+
+---
+
+*Edit:*
+
+Link to recent [SS3
+benchmarks](https://github.com/nmfs-stock-synthesis/stock-synthesis/issues/257#issuecomment-1125039928).