diff --git a/kernel/Kconfig.power_mgmt b/kernel/Kconfig.power_mgmt
index 73baa7be807..8e6c15e5483 100644
--- a/kernel/Kconfig.power_mgmt
+++ b/kernel/Kconfig.power_mgmt
@@ -14,6 +14,29 @@ menuconfig SYS_POWER_MANAGEMENT
 	  timer is due to expire.
 
 if SYS_POWER_MANAGEMENT
+choice POWER_MANAGEMENT_CONTROL
+	prompt "Power Management Control"
+	default PM_CONTROL_APP
+	help
+	  Select the Application managed or OS managed power saving
+	  mechanism.
+
+config PM_CONTROL_APP
+	bool
+	prompt "Handled at Application level"
+	help
+	  This option enbles the Application to handle all the Power
+	  Management flows for the platform.
+
+config PM_CONTROL_OS
+	bool
+	prompt "Handle at OS level"
+	help
+	  This option allows the OS to handle all the Power
+	  Management flows for the platform.
+
+endchoice # POWER_MANAGEMENT_CONTROL
+
 config SYS_POWER_LOW_POWER_STATE
 	bool "Low power state"
 	depends on SYS_POWER_LOW_POWER_STATE_SUPPORTED
diff --git a/subsys/CMakeLists.txt b/subsys/CMakeLists.txt
index 3e71bac6f69..50b74b66975 100644
--- a/subsys/CMakeLists.txt
+++ b/subsys/CMakeLists.txt
@@ -14,3 +14,4 @@ add_subdirectory_ifdef(CONFIG_USB                  usb)
 add_subdirectory(random)
 add_subdirectory(storage)
 add_subdirectory_ifdef(CONFIG_SETTINGS             settings)
+add_subdirectory_ifdef(CONFIG_PM_CONTROL_OS        power)
diff --git a/subsys/Kconfig b/subsys/Kconfig
index 0630d7d6751..f3deb779855 100644
--- a/subsys/Kconfig
+++ b/subsys/Kconfig
@@ -34,3 +34,5 @@ source "subsys/storage/Kconfig"
 source "subsys/settings/Kconfig"
 
 source "subsys/app_memory/Kconfig"
+
+source "subsys/power/Kconfig"
diff --git a/subsys/power/CMakeLists.txt b/subsys/power/CMakeLists.txt
new file mode 100644
index 00000000000..fc6ed85e682
--- /dev/null
+++ b/subsys/power/CMakeLists.txt
@@ -0,0 +1,5 @@
+zephyr_sources(
+  power.c
+  policy.c
+  device.c
+  )
diff --git a/subsys/power/Kconfig b/subsys/power/Kconfig
new file mode 100644
index 00000000000..d88499e5b2a
--- /dev/null
+++ b/subsys/power/Kconfig
@@ -0,0 +1,88 @@
+
+if PM_CONTROL_OS
+menu "OS Power Management"
+
+if SYS_POWER_LOW_POWER_STATE
+config PM_CONTROL_OS_LPS
+	bool "Platform supports LPS"
+	help
+	  Select this option if SoC support LPS state.
+
+if PM_CONTROL_OS_LPS
+config PM_LPS_MIN_RES
+	int "LPS minimum residency"
+	default 5
+	help
+	  Minimum residency in ticks to enter LPS state.
+endif
+
+config PM_CONTROL_OS_LPS_1
+	bool "Platform supports LPS_1"
+	help
+	  Select this option if SoC support LPS_1 state.
+
+if PM_CONTROL_OS_LPS_1
+config PM_LPS_1_MIN_RES
+	int "LPS_1 minimum residency"
+	default 10
+	help
+	  Minimum residency in ticks to enter LPS_1 state.
+endif
+
+config PM_CONTROL_OS_LPS_2
+	bool "Platform supports LPS_2"
+	help
+	  Select this option if SoC support LPS_2 state.
+
+if PM_CONTROL_OS_LPS_2
+config PM_LPS_2_MIN_RES
+	int "LPS_2 minimum residency"
+	default 30
+	help
+	  Minimum residency in ticks to enter LPS_2 state.
+endif
+endif # SYS_POWER_LOW_POWER_STATE
+
+if SYS_POWER_DEEP_SLEEP
+config PM_CONTROL_OS_DEEP_SLEEP
+	bool "Platform supports DEEP_SLEEP"
+	help
+	  Select this option if SoC support DEEP_SLEEP state.
+
+if PM_CONTROL_OS_DEEP_SLEEP
+config PM_DEEP_SLEEP_MIN_RES
+	int "DEEP_SLEEP minimum residency"
+	default 60
+	help
+	  Minimum residency in ticks to enter DEEP_SLEEP state.
+endif
+
+config PM_CONTROL_OS_DEEP_SLEEP_1
+	bool "Platform supports DEEP_SLEEP_1"
+	help
+	  Select this option if SoC support DEEP_SLEEP_1 state.
+
+if PM_CONTROL_OS_DEEP_SLEEP_1
+config PM_DEEP_SLEEP_1_MIN_RES
+	int "DEEP_SLEEP_1 minimum residency"
+	default 90
+	help
+	  Minimum residency in ticks to enter DEEP_SLEEP_1 state.
+endif
+
+config PM_CONTROL_OS_DEEP_SLEEP_2
+	bool "Platform supports DEEP_SLEEP_2"
+	help
+	  Select this option if SoC support DEEP_SLEEP_2 state.
+
+if PM_CONTROL_OS_DEEP_SLEEP_2
+config PM_DEEP_SLEEP_2_MIN_RES
+	int "DEEP_SLEEP_2 minimum residency"
+	default 120
+	help
+	  Minimum residency in ticks to enter DEEP_SLEEP_2 state.
+endif
+endif # PM_CONTROL_OS_DEEP_SLEEP
+
+endmenu
+endif # PM_CONTROL_OS
diff --git a/subsys/power/device.c b/subsys/power/device.c
new file mode 100644
index 00000000000..2f411183544
--- /dev/null
+++ b/subsys/power/device.c
@@ -0,0 +1,115 @@
+/*
+ * Copyright (c) 2018 Intel Corporation.
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
+#include <zephyr.h>
+#include <kernel.h>
+#include <string.h>
+#include <soc.h>
+#include <device.h>
+#include "pm_policy.h"
+
+#define LOG_MODULE_NAME power
+#define LOG_LEVEL CONFIG_PM_LOG_LEVEL /* From power module Kconfig */
+#include <logging/log.h>
+LOG_MODULE_DECLARE();
+
+/*
+ * FIXME: Remove the conditional inclusion of
+ * core_devices array once we enble the capability
+ * to build the device list based on devices power
+ * and clock domain dependencies.
+ */
+#ifdef CONFIG_SOC_SERIES_NRF52X
+#define MAX_PM_DEVICES	15
+#define NUM_CORE_DEVICES	4
+#define MAX_DEV_NAME_LEN	16
+static const char core_devices[NUM_CORE_DEVICES][MAX_DEV_NAME_LEN] = {
+	"clk_k32src",
+	"clk_m16src",
+	"sys_clock",
+	"UART_0",
+};
+#else
+#error "Add SoC's core devices list for PM"
+#endif
+
+/*
+ * Ordered list to store devices on which
+ * device power policies would be executed.
+ */
+static int device_ordered_list[MAX_PM_DEVICES];
+static int device_retval[MAX_PM_DEVICES];
+static struct device *pm_device_list;
+static int device_count;
+
+int sys_pm_suspend_devices(void)
+{
+	for (int i = device_count - 1; i >= 0; i--) {
+		int idx = device_ordered_list[i];
+
+		/* TODO: Improve the logic by checking device status
+		 * and set the device states accordingly.
+		 */
+		device_retval[i] = device_set_power_state(&pm_device_list[idx],
+						DEVICE_PM_SUSPEND_STATE);
+		if (device_retval[i]) {
+			LOG_ERR("%s suspend operation failed\n",
+					pm_device_list[idx].config->name);
+			return device_retval[i];
+		}
+	}
+
+	return 0;
+}
+
+void sys_pm_resume_devices(void)
+{
+	int i;
+
+	for (i = 0; i < device_count; i++) {
+		if (!device_retval[i]) {
+			int idx = device_ordered_list[i];
+
+			device_set_power_state(&pm_device_list[idx],
+						DEVICE_PM_ACTIVE_STATE);
+		}
+	}
+}
+
+void sys_pm_create_device_list(void)
+{
+	int count;
+	int i, j;
+	bool is_core_dev;
+
+	/*
+	 * Create an ordered list of devices that will be suspended.
+	 * Ordering should be done based on dependencies. Devices
+	 * in the beginning of the list will be resumed first.
+	 */
+	device_list_get(&pm_device_list, &count);
+
+	/* Reserve for 32KHz, 16MHz, system clock, etc... */
+	device_count = NUM_CORE_DEVICES;
+
+	for (i = 0; (i < count) && (device_count < MAX_PM_DEVICES); i++) {
+
+		/* Check if the device is core device */
+		for (j = 0, is_core_dev = false; j < NUM_CORE_DEVICES; j++) {
+			if (!strcmp(pm_device_list[i].config->name,
+						&core_devices[j][0])) {
+				is_core_dev = true;
+				break;
+			}
+		}
+
+		if (is_core_dev) {
+			device_ordered_list[j] = i;
+		} else {
+			device_ordered_list[device_count++] = i;
+		}
+	}
+}
diff --git a/subsys/power/pm_policy.h b/subsys/power/pm_policy.h
new file mode 100644
index 00000000000..0fb5a99d039
--- /dev/null
+++ b/subsys/power/pm_policy.h
@@ -0,0 +1,57 @@
+/*
+ * Copyright (c) 2018 Intel Corporation.
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
+#ifndef _PM_POLICY_H_
+#define _PM_POLICY_H_
+
+#include <power.h>
+#include <soc_power.h>
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+/**
+ * @brief Function to create device PM list
+ */
+extern  void sys_pm_create_device_list(void);
+
+/**
+ * @brief Function to suspend the devices in PM device list
+ */
+extern int sys_pm_suspend_devices(void);
+
+/**
+ * @brief Function to resume the devices in PM device list
+ */
+extern void sys_pm_resume_devices(void);
+
+/**
+ * @brief Function to get the next PM state based on the ticks
+ */
+extern int sys_pm_policy_next_state(s32_t ticks, enum power_states *state);
+
+/**
+ * @brief Application defined function for Lower Power entry
+ *
+ * Application defined function for doing any target specific operations
+ * for low power entry.
+ */
+extern void sys_pm_notify_lps_entry(enum power_states state);
+
+/**
+ * @brief Application defined function for Lower Power exit
+ *
+ * Application defined function for doing any target specific operations
+ * for low power exit.
+ */
+extern void sys_pm_notify_lps_exit(enum power_states state);
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif /* _PM_POLICY_H_ */
diff --git a/subsys/power/policy.c b/subsys/power/policy.c
new file mode 100644
index 00000000000..ef204e1dde5
--- /dev/null
+++ b/subsys/power/policy.c
@@ -0,0 +1,98 @@
+/*
+ * Copyright (c) 2018 Intel Corporation.
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
+#include <zephyr.h>
+#include <kernel.h>
+#include <soc.h>
+#include "pm_policy.h"
+
+#define LOG_MODULE_NAME power
+#define LOG_LEVEL CONFIG_PM_LOG_LEVEL /* From power module Kconfig */
+#include <logging/log.h>
+LOG_MODULE_DECLARE();
+
+#ifdef CONFIG_TICKLESS_KERNEL
+#define SECS_TO_TICKS		CONFIG_TICKLESS_KERNEL_TIME_UNIT_IN_MICRO_SECS
+#else
+#define SECS_TO_TICKS		CONFIG_SYS_CLOCK_TICKS_PER_SEC
+#endif
+
+#if !(defined(CONFIG_PM_CONTROL_OS_LPS) || \
+		defined(CONFIG_PM_CONTROL_OS_LPS_1) || \
+		defined(CONFIG_PM_CONTROL_OS_LPS_2) || \
+		defined(CONFIG_PM_CONTROL_OS_DEEP_SLEEP) || \
+		defined(CONFIG_PM_CONTROL_OS_DEEP_SLEEP_1) || \
+		defined(CONFIG_PM_CONTROL_OS_DEEP_SLEEP_2))
+#error "Low Power states not selected by policy"
+#endif
+
+struct sys_soc_pm_policy {
+	enum power_states pm_state;
+	int sys_state;
+	int min_residency;
+};
+
+/* PM Policy based on SoC/Platform residency requirements */
+static struct sys_soc_pm_policy pm_policy[] = {
+#ifdef CONFIG_PM_CONTROL_OS_LPS
+	{SYS_POWER_STATE_CPU_LPS, SYS_PM_LOW_POWER_STATE,
+			CONFIG_PM_LPS_MIN_RES * SECS_TO_TICKS},
+#endif
+
+#ifdef CONFIG_PM_CONTROL_OS_LPS_1
+	{SYS_POWER_STATE_CPU_LPS_1, SYS_PM_LOW_POWER_STATE,
+			CONFIG_PM_LPS_1_MIN_RES * SECS_TO_TICKS},
+#endif
+
+#ifdef CONFIG_PM_CONTROL_OS_LPS_2
+	{SYS_POWER_STATE_CPU_LPS_2, SYS_PM_LOW_POWER_STATE,
+			CONFIG_PM_LPS_2_MIN_RES * SECS_TO_TICKS},
+#endif
+
+#ifdef CONFIG_PM_CONTROL_OS_DEEP_SLEEP
+	{SYS_POWER_STATE_DEEP_SLEEP, SYS_PM_DEEP_SLEEP,
+			CONFIG_PM_DEEP_SLEEP_MIN_RES * SECS_TO_TICKS},
+#endif
+
+#ifdef CONFIG_PM_CONTROL_OS_DEEP_SLEEP_1
+	{SYS_POWER_STATE_DEEP_SLEEP_1, SYS_PM_DEEP_SLEEP,
+			CONFIG_PM_DEEP_SLEEP_1_MIN_RES * SECS_TO_TICKS},
+#endif
+
+#ifdef CONFIG_PM_CONTROL_OS_DEEP_SLEEP_2
+	{SYS_POWER_STATE_DEEP_SLEEP_2, SYS_PM_DEEP_SLEEP,
+			CONFIG_PM_DEEP_SLEEP_2_MIN_RES * SECS_TO_TICKS},
+#endif
+};
+
+int sys_pm_policy_next_state(s32_t ticks, enum power_states *pm_state)
+{
+	int i;
+
+	if ((ticks != K_FOREVER) && (ticks < pm_policy[0].min_residency)) {
+		LOG_ERR("Not enough time for PM operations: %d\n", ticks);
+		return SYS_PM_NOT_HANDLED;
+	}
+
+	for (i = 0; i < (ARRAY_SIZE(pm_policy) - 1); i++) {
+		if ((ticks >= pm_policy[i].min_residency) &&
+			       (ticks < pm_policy[i + 1].min_residency)) {
+			break;
+		}
+	}
+
+	if (!_sys_soc_is_valid_power_state(pm_policy[i].pm_state)) {
+		LOG_ERR("pm_state(%d) not supported by SoC\n",
+						pm_policy[i].pm_state);
+		return SYS_PM_NOT_HANDLED;
+	}
+
+	*pm_state = pm_policy[i].pm_state;
+	LOG_DBG("pm_state: %d, min_residency: %d, idx: %d\n",
+				*pm_state, pm_policy[i].min_residency, i);
+
+	return pm_policy[i].sys_state;
+}
diff --git a/subsys/power/power.c b/subsys/power/power.c
new file mode 100644
index 00000000000..fa039836a88
--- /dev/null
+++ b/subsys/power/power.c
@@ -0,0 +1,107 @@
+/*
+ * Copyright (c) 2018 Intel Corporation.
+ *
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
+#include <zephyr.h>
+#include <kernel.h>
+#include <init.h>
+#include <string.h>
+#include <soc.h>
+#include "pm_policy.h"
+
+#define LOG_MODULE_NAME power
+#define LOG_LEVEL CONFIG_PM_LOG_LEVEL /* From power module Kconfig */
+#include <logging/log.h>
+LOG_MODULE_REGISTER();
+
+static int post_ops_done = 1;
+static enum power_states pm_state;
+
+int _sys_soc_suspend(s32_t ticks)
+{
+	int sys_state;
+
+	post_ops_done = 0;
+
+	sys_state = sys_pm_policy_next_state(ticks, &pm_state);
+
+	switch (sys_state) {
+	case SYS_PM_LOW_POWER_STATE:
+		sys_pm_notify_lps_entry(pm_state);
+		/* Do CPU LPS operations */
+		_sys_soc_set_power_state(pm_state);
+		break;
+	case SYS_PM_DEEP_SLEEP:
+		/* Don't need pm idle exit event notification */
+		_sys_soc_pm_idle_exit_notification_disable();
+
+		sys_pm_notify_lps_entry(pm_state);
+
+		/* Save device states and turn off peripherals as necessary */
+		if (sys_pm_suspend_devices()) {
+			LOG_ERR("System level device suspend failed\n");
+			break;
+		}
+
+		/* Enter CPU deep sleep state */
+		_sys_soc_set_power_state(pm_state);
+
+		/* Turn on peripherals and restore device states as necessary */
+		sys_pm_resume_devices();
+		break;
+	default:
+		/* No PM operations */
+		LOG_DBG("\nNo PM operations done\n");
+		break;
+	}
+
+	if (sys_state != SYS_PM_NOT_HANDLED) {
+		/*
+		 * Do any arch or soc specific post operations specific to the
+		 * power state.
+		 */
+		if (!post_ops_done) {
+			post_ops_done = 1;
+			sys_pm_notify_lps_exit(pm_state);
+			_sys_soc_power_state_post_ops(pm_state);
+		}
+	}
+
+	return sys_state;
+}
+
+void _sys_soc_resume(void)
+{
+	/*
+	 * This notification is called from the ISR of the event
+	 * that caused exit from kernel idling after PM operations.
+	 *
+	 * Some CPU low power states require enabling of interrupts
+	 * atomically when entering those states. The wake up from
+	 * such a state first executes code in the ISR of the interrupt
+	 * that caused the wake. This hook will be called from the ISR.
+	 * For such CPU LPS states, do post operations and restores here.
+	 * The kernel scheduler will get control after the ISR finishes
+	 * and it may schedule another thread.
+	 *
+	 * Call _sys_soc_pm_idle_exit_notification_disable() if this
+	 * notification is not required.
+	 */
+	if (!post_ops_done) {
+		post_ops_done = 1;
+		sys_pm_notify_lps_exit(pm_state);
+		_sys_soc_power_state_post_ops(pm_state);
+	}
+}
+
+static int sys_pm_init(struct device *dev)
+{
+	ARG_UNUSED(dev);
+
+	sys_pm_create_device_list();
+	return 0;
+}
+
+SYS_INIT(sys_pm_init, APPLICATION, CONFIG_KERNEL_INIT_PRIORITY_DEFAULT);