diff --git a/.github/scripts/performance-benchmarking/get-dolt-dolt-job-json.sh b/.github/scripts/performance-benchmarking/get-dolt-dolt-job-json.sh
index 2fc88e0877..f48cc878c9 100755
--- a/.github/scripts/performance-benchmarking/get-dolt-dolt-job-json.sh
+++ b/.github/scripts/performance-benchmarking/get-dolt-dolt-job-json.sh
@@ -3,7 +3,7 @@
 set -e
 
 if [ "$#" -lt 10 ]; then
-    echo  "Usage: ./get-job-json.sh <jobname> <fromServer> <fromVersion> <toServer> <toVersion> <timePrefix> <actorPrefix> <format> <issueNumber> <initBigRepo>"
+    echo  "Usage: ./get-job-json.sh <jobname> <fromServer> <fromVersion> <toServer> <toVersion> <timePrefix> <actorPrefix> <format> <issueNumber> <initBigRepo> <nomsBinFormat>"
     exit 1
 fi
 
@@ -17,6 +17,7 @@ actorPrefix="$7"
 format="$8"
 issueNumber="$9"
 initBigRepo="${10}"
+nomsBinFormat="${11}"
 tpccRegex="tpcc%"
 
 readTests="('oltp_read_only', 'oltp_point_select', 'select_random_points', 'select_random_ranges', 'covering_index_scan', 'index_scan', 'table_scan', 'groupby_scan')"
@@ -75,7 +76,8 @@ echo '
               "--sysbenchQueries='"$medianLatencyChangeWritesQuery"'",
               "--tpccQueries='"$tpccLatencyQuery"'",
               "--tpccQueries='"$tpccTpsQuery"'",
-              "--init-big-repo='"$initBigRepo"'""
+              "--init-big-repo='"$initBigRepo"'"",
+              "--noms-bin-format='"$nomsBinFormat"'""
             ]
           }
         ],
diff --git a/.github/scripts/performance-benchmarking/run-benchmarks.sh b/.github/scripts/performance-benchmarking/run-benchmarks.sh
index 9a90e38962..25d4a9c112 100755
--- a/.github/scripts/performance-benchmarking/run-benchmarks.sh
+++ b/.github/scripts/performance-benchmarking/run-benchmarks.sh
@@ -62,7 +62,8 @@ source \
   "$actorprefix"     \
   "$format"          \
   "$issuenumber"     \
-  "$INIT_BIG_REPO"   > job.json
+  "$INIT_BIG_REPO"   \
+  "$NOMS_BIN_FORMAT" > job.json
 
 out=$(KUBECONFIG="$KUBECONFIG" kubectl apply -f job.json || true)
 
diff --git a/.github/workflows/ci-performance-benchmarks.yaml b/.github/workflows/ci-performance-benchmarks.yaml
index 9c1d437d13..576facf798 100644
--- a/.github/workflows/ci-performance-benchmarks.yaml
+++ b/.github/workflows/ci-performance-benchmarks.yaml
@@ -42,6 +42,15 @@ jobs:
           echo "::set-output name=benchmark::true"
 
   performance:
+    strategy:
+      matrix:
+        biginit: ["true", "false"]
+        nbf: ["__LD_1__", "__DOLT_1__"]
+        exclude:
+          - biginit: "true"
+            nbf: "__DOLT_1__"
+          - bigint: "false"
+            nbf: "__LD_1__"
     runs-on: ubuntu-18.04
     needs: [validate-commentor, check-comments]
     if: ${{ needs.check-comments.outputs.benchmark == 'true' }}
@@ -95,5 +104,6 @@ jobs:
           ACTOR: ${{ github.actor }}
           REPO_ACCESS_TOKEN: ${{ secrets.REPO_ACCESS_TOKEN }}
           KUBECONFIG: "./kubeconfig"
-          INIT_BIG_REPO: "true"
+          INIT_BIG_REPO: ${{ matrix.biginit }}
+          NOMS_BIN_FORMAT: ${{ matrix.nbf }}
           TEMPLATE_SCRIPT: "./.github/scripts/performance-benchmarking/get-dolt-dolt-job-json.sh"
diff --git a/go/performance/utils/sysbench_runner/config.go b/go/performance/utils/sysbench_runner/config.go
index 75a35f98df..22acff9bf6 100644
--- a/go/performance/utils/sysbench_runner/config.go
+++ b/go/performance/utils/sysbench_runner/config.go
@@ -281,6 +281,8 @@ type Config struct {
 	ScriptDir string
 	// InitBigRepo downloads a database with existing chunks and commits
 	InitBigRepo bool
+	// NomsBinFormat specifies the NomsBinFormat
+	NomsBinFormat string
 }
 
 // NewConfig returns a new Config
diff --git a/go/performance/utils/sysbench_runner/dolt.go b/go/performance/utils/sysbench_runner/dolt.go
index b08ec21502..67aee65e8b 100644
--- a/go/performance/utils/sysbench_runner/dolt.go
+++ b/go/performance/utils/sysbench_runner/dolt.go
@@ -32,6 +32,7 @@ const (
 	dbName       = "test"
 	luaPath      = "?.lua"
 	bigEmptyRepo = "max-hoffman/big-empty"
+	nbfEnvVar    = "DOLT_DEFAULT_BIN_FORMAT"
 )
 
 var stampFunc = func() string { return time.Now().UTC().Format(stampFormat) }
@@ -50,7 +51,7 @@ func BenchmarkDolt(ctx context.Context, config *Config, serverConfig *ServerConf
 		return nil, err
 	}
 
-	testRepo, err := initDoltRepo(ctx, serverConfig, config.InitBigRepo)
+	testRepo, err := initDoltRepo(ctx, serverConfig, config.InitBigRepo, config.NomsBinFormat)
 	if err != nil {
 		return nil, err
 	}
@@ -133,7 +134,7 @@ func doltVersion(ctx context.Context, config *ServerConfig) error {
 }
 
 // initDoltRepo initializes a dolt repo and returns the repo path
-func initDoltRepo(ctx context.Context, config *ServerConfig, initBigRepo bool) (string, error) {
+func initDoltRepo(ctx context.Context, config *ServerConfig, initBigRepo bool, nbf string) (string, error) {
 	cwd, err := os.Getwd()
 	if err != nil {
 		return "", err
@@ -152,6 +153,12 @@ func initDoltRepo(ctx context.Context, config *ServerConfig, initBigRepo bool) (
 		return "", err
 	}
 
+	if nbf != "" {
+		if err = os.Setenv(nbfEnvVar, nbf); err != nil {
+			return "", err
+		}
+	}
+
 	doltInit := ExecCommand(ctx, config.ServerExec, "init")
 	doltInit.Dir = testRepo
 	err = doltInit.Run()