diff --git a/rpcs3/Emu/Cell/SPUThread.cpp b/rpcs3/Emu/Cell/SPUThread.cpp
index 14129e731d..36bd0143b6 100644
--- a/rpcs3/Emu/Cell/SPUThread.cpp
+++ b/rpcs3/Emu/Cell/SPUThread.cpp
@@ -1662,7 +1662,7 @@ void spu_thread::cpu_work()
 	// If either MFC size exceeds limit or timeout has been reached execute pending MFC commands
 	if (mfc_size > g_cfg.core.mfc_transfers_shuffling || (timeout && get_system_time() - mfc_last_timestamp >= timeout))
 	{
-		do_mfc(false);
+		do_mfc(false, false);
 		check_mfc_interrupts(pc + 4);
 	}
 }
@@ -2960,7 +2960,7 @@ void spu_thread::do_putlluc(const spu_mfc_cmd& args)
 	vm::reservation_notifier(addr).notify_all(-128);
 }
 
-void spu_thread::do_mfc(bool can_escape)
+void spu_thread::do_mfc(bool can_escape, bool must_finish)
 {
 	u32 removed = 0;
 	u32 barrier = 0;
@@ -3106,6 +3106,14 @@ void spu_thread::do_mfc(bool can_escape)
 		{
 			break;
 		}
+
+		if (!must_finish && g_cfg.core.mfc_shuffling_in_steps)
+		{
+			// Exit early, not all pending commands have to be executed at a single iteration
+			// Update last timestamp so the next MFC timeout check will use the current time
+			mfc_last_timestamp = get_system_time();
+			return;
+		}
 	}
 
 	if (state & cpu_flag::pending)
@@ -3171,12 +3179,14 @@ bool spu_thread::process_mfc_cmd()
 		// Reset MFC timestamp in the case of full queue
 		mfc_last_timestamp = 0;
 
-		// Process MFC commands
 		if (test_stopped())
 		{
 			return false;
 		}
 
+		// Process MFC commands
+		do_mfc();
+
 		auto old = state.add_fetch(cpu_flag::wait);
 
 		if (is_stopped(old))
diff --git a/rpcs3/Emu/Cell/SPUThread.h b/rpcs3/Emu/Cell/SPUThread.h
index 5047676b05..8ef8a7cec5 100644
--- a/rpcs3/Emu/Cell/SPUThread.h
+++ b/rpcs3/Emu/Cell/SPUThread.h
@@ -791,7 +791,7 @@ public:
 	bool do_list_transfer(spu_mfc_cmd& args);
 	void do_putlluc(const spu_mfc_cmd& args);
 	bool do_putllc(const spu_mfc_cmd& args);
-	void do_mfc(bool can_escape = true);
+	void do_mfc(bool can_escape = true, bool must_finish = true);
 	u32 get_mfc_completed() const;
 
 	bool process_mfc_cmd();
diff --git a/rpcs3/Emu/system_config.h b/rpcs3/Emu/system_config.h
index 7cffb9aefb..3c76c271e9 100644
--- a/rpcs3/Emu/system_config.h
+++ b/rpcs3/Emu/system_config.h
@@ -48,7 +48,8 @@ struct cfg_root : cfg::node
 		cfg::_bool spu_cache{ this, "SPU Cache", true };
 		cfg::_bool spu_prof{ this, "SPU Profiler", false };
 		cfg::uint<0, 16> mfc_transfers_shuffling{ this, "MFC Commands Shuffling Limit", 0 };
-		cfg::uint<0, 10000> mfc_transfers_timeout{ this, "MFC Commands Timeout", 0, true};
+		cfg::uint<0, 10000> mfc_transfers_timeout{ this, "MFC Commands Timeout", 0, true };
+		cfg::_bool mfc_shuffling_in_steps{ this, "MFC Commands Shuffling In Steps", false, true };
 		cfg::_enum<tsx_usage> enable_TSX{ this, "Enable TSX", has_rtm() ? tsx_usage::enabled : tsx_usage::disabled }; // Enable TSX. Forcing this on Haswell/Broadwell CPUs should be used carefully
 		cfg::_bool spu_accurate_xfloat{ this, "Accurate xfloat", false };
 		cfg::_bool spu_approx_xfloat{ this, "Approximate xfloat", true };