diff --git a/misc/h2o/h2o.jl b/misc/h2o/h2o.jl
new file mode 100644
index 0000000..ea7520d
--- /dev/null
+++ b/misc/h2o/h2o.jl
@@ -0,0 +1,32 @@
+ENV["JULIA_MEMPOOL_EXPERIMENTAL_FANCY_ALLOCATOR"] = "true"
+ENV["JULIA_MEMPOOL_EXPERIMENTAL_MEMORY_BOUND"] = string(8 * (2^30)) 
+
+using CSV
+using DTables
+using DataFrames
+using Tables
+using OnlineStats
+
+d = DTable(CSV.Chunks("misc/h2o/data/G1_1e8_1e2_0_0.csv",pool=true,ntasks=100))
+
+gb = GroupBy(String7, Sum())
+f = r -> (r.id1, r.v1)
+@time a3 = fetch(DTables.mapreduce(f, fit!, d, init=gb))
+
+# @time a1 = fetch(
+#     reduce(
+#         fit!, 
+#         DTables.groupby(d,:id1),
+#         cols=[:v1];
+#         init=Sum()
+#     )
+# )
+
+# @time a2 = fetch(
+#     reduce(
+#         fit!, 
+#         DTables.groupby(d,:id1),
+#         cols=[:v1];
+#         init=Sum()
+#     )
+# )
\ No newline at end of file