diff --git a/tests/tests_net_messages/test_benchmark.py b/tests/tests_net_messages/test_benchmark.py
index 9387970e..8ca36ffc 100644
--- a/tests/tests_net_messages/test_benchmark.py
+++ b/tests/tests_net_messages/test_benchmark.py
@@ -1,8 +1,10 @@
+import os
 import asyncio
 import random
 import string
 import time
 import threading
+import multiprocessing
 from unittest import IsolatedAsyncioTestCase, skip
 from lifeblood.logging import get_logger, set_default_loglevel
 from lifeblood.nethelpers import get_localhost
@@ -32,20 +34,40 @@ async def process_message(self, message: Message, client: MessageClient):
         self.test_messages_count += 1
 
 
-class ThreadedFoo(threading.Thread):
-    def __init__(self, server: TcpMessageProcessor):
+class FooRunner:
+    def start(self):
+        raise NotImplementedError()
+
+    def stop(self):
+        raise NotImplementedError()
+
+    def join(self):
+        raise NotImplementedError()
+
+    def get_message_count(self) -> int:
+        raise NotImplementedError()
+
+
+class ThreadedFoo(threading.Thread, FooRunner):
+    def __init__(self, server: NoopMessageServer):
         super().__init__()
-        self.__stop = False
+        self.__stop = threading.Event()
+        self.__ready = threading.Event()
         self.__server = server
 
     def run(self):
         asyncio.run(self.async_run())
 
+    def start(self):
+        super().start()
+        self.__ready.wait()
+
     async def async_run(self):
         await self.__server.start()
+        self.__ready.set()
         while True:
             await asyncio.sleep(1)
-            if self.__stop:
+            if self.__stop.is_set():
                 break
 
         self.__server.stop()
@@ -53,15 +75,72 @@ async def async_run(self):
 
     def stop(self):
         # crude crude crude
-        self.__stop = True
+        self.__stop.set()
+
+    def get_message_count(self) -> int:
+        return self.__server.test_messages_count
+
+
+class ProcessedFoo(FooRunner):
+    def __init__(self, server: NoopMessageServer):
+        super().__init__()
+        self.__server = server
+        ctx = multiprocessing.get_context('spawn')
+        self.__stop = ctx.Event()
+        self.__value = ctx.Value('i', -1)
+        self.__ready = ctx.Event()
+        self.__proc = ctx.Process(target=self.body)
+
+    def start(self):
+        self.__proc.start()
+        self.__ready.wait()
+
+    def body(self):
+        asyncio.run(self.async_run())
+
+    async def async_run(self):
+        print('another process started')
+        await self.__server.start()
+        self.__ready.set()
+        print('another process server started')
+        while True:
+            await asyncio.sleep(1)
+            if self.__stop.is_set():
+                break
+
+        self.__server.stop()
+        await self.__server.wait_till_stops()
+        self.__value.value = self.__server.test_messages_count
+
+    def stop(self):
+        self.__stop.set()
+
+    def join(self):
+        self.__proc.join()
+
+    def get_message_count(self) -> int:
+        return self.__value.value
 
 
 class TestBenchmarkSendReceive(IsolatedAsyncioTestCase):
-    async def test1(self):
+    @skip("no reason to benchmark on slow machines")
+    async def test_threaded(self):
+        await self.helper_test(ThreadedFoo)
+
+    @skip("no reason to benchmark on slow machines")
+    async def test_proc(self):
+        await self.helper_test(ProcessedFoo)
+
+    async def helper_test(self, foo_factory: Callable[[NoopMessageServer], FooRunner]):
+        """
+        runs 2 servers
+        starts X clients asyncio coroutines on one of the servers, each sends Y messages.
+        average per-message time is then calculated
+        """
         data = ''.join(random.choice(string.ascii_letters) for _ in range(16000)).encode('latin1')
         server1 = NoopMessageServer((get_localhost(), 28385))
         server2 = NoopMessageServer((get_localhost(), 28386))
-        server1_runner = ThreadedFoo(server1)
+        server1_runner = foo_factory(server1)
         server1_runner.start()
         await server2.start()
         pure_send_time = 0.0
@@ -90,5 +169,6 @@ async def test_foo():
         server1_runner.stop()
         await server2.wait_till_stops()
         server1_runner.join()
-        print(f'total go {server1.test_messages_count} in {total_time}s (pure send: {pure_send_time}s, avg {server1.test_messages_count/total_time} (pure: {server1.test_messages_count/pure_send_time}) msg/s')
-        self.assertEqual(total_clients * messages_per_client, server1.test_messages_count)
+        s1_message_count = server1_runner.get_message_count()
+        print(f'threaded total go {s1_message_count} in {total_time}s (pure send: {pure_send_time}s, avg {s1_message_count/total_time} (pure: {s1_message_count/pure_send_time}) msg/s')
+        self.assertEqual(total_clients * messages_per_client, s1_message_count)