diff --git a/kernel/src/lib.rs b/kernel/src/lib.rs
index b3e640ad2..abf1d26ec 100644
--- a/kernel/src/lib.rs
+++ b/kernel/src/lib.rs
@@ -115,7 +115,7 @@ fn ap_init() -> ! {
         let preempt_guard = ostd::task::disable_preempt();
         let cpu_id = preempt_guard.current_cpu();
         drop(preempt_guard);
-        log::info!("Kernel idle thread for CPU #{} started.", cpu_id);
+        log::info!("Kernel idle thread for CPU #{} started.", cpu_id.as_usize());
         loop {
             Thread::yield_now();
         }
diff --git a/kernel/src/process/posix_thread/futex.rs b/kernel/src/process/posix_thread/futex.rs
index a232baea2..c68d9752e 100644
--- a/kernel/src/process/posix_thread/futex.rs
+++ b/kernel/src/process/posix_thread/futex.rs
@@ -164,7 +164,7 @@ static FUTEX_BUCKETS: Once<FutexBucketVec> = Once::new();
 /// This number is calculated the same way as Linux's:
 /// <https://github.com/torvalds/linux/blob/master/kernel/futex/core.c>
 fn get_bucket_count() -> usize {
-    ((1 << 8) * num_cpus()).next_power_of_two() as usize
+    ((1 << 8) * num_cpus()).next_power_of_two()
 }
 
 fn get_futex_bucket(key: FutexKey) -> (usize, FutexBucketRef) {
diff --git a/kernel/src/sched/priority_scheduler.rs b/kernel/src/sched/priority_scheduler.rs
index 289d85945..8c2ff507f 100644
--- a/kernel/src/sched/priority_scheduler.rs
+++ b/kernel/src/sched/priority_scheduler.rs
@@ -1,7 +1,7 @@
 // SPDX-License-Identifier: MPL-2.0
 
 use ostd::{
-    cpu::{num_cpus, CpuSet, PinCurrentCpu},
+    cpu::{num_cpus, CpuId, CpuSet, PinCurrentCpu},
     sync::PreemptDisabled,
     task::{
         scheduler::{
@@ -33,8 +33,8 @@ struct PreemptScheduler<T: PreemptSchedInfo + FromTask<U>, U: CommonSchedInfo> {
 }
 
 impl<T: PreemptSchedInfo + FromTask<U>, U: CommonSchedInfo> PreemptScheduler<T, U> {
-    fn new(nr_cpus: u32) -> Self {
-        let mut rq = Vec::with_capacity(nr_cpus as usize);
+    fn new(nr_cpus: usize) -> Self {
+        let mut rq = Vec::with_capacity(nr_cpus);
         for _ in 0..nr_cpus {
             rq.push(SpinLock::new(PreemptRunQueue::new()));
         }
@@ -42,7 +42,7 @@ impl<T: PreemptSchedInfo + FromTask<U>, U: CommonSchedInfo> PreemptScheduler<T,
     }
 
     /// Selects a CPU for task to run on for the first time.
-    fn select_cpu(&self, entity: &PreemptSchedEntity<T, U>) -> u32 {
+    fn select_cpu(&self, entity: &PreemptSchedEntity<T, U>) -> CpuId {
         // If the CPU of a runnable task has been set before, keep scheduling
         // the task to that one.
         // TODO: Consider migrating tasks between CPUs for load balancing.
@@ -55,7 +55,7 @@ impl<T: PreemptSchedInfo + FromTask<U>, U: CommonSchedInfo> PreemptScheduler<T,
         let mut minimum_load = usize::MAX;
 
         for candidate in entity.thread.cpu_affinity().iter() {
-            let rq = self.rq[candidate as usize].lock();
+            let rq = self.rq[candidate.as_usize()].lock();
             // A wild guess measuring the load of a runqueue. We assume that
             // real-time tasks are 4-times as important as normal tasks.
             let load = rq.real_time_entities.len() * 8
@@ -74,7 +74,7 @@ impl<T: PreemptSchedInfo + FromTask<U>, U: CommonSchedInfo> PreemptScheduler<T,
 impl<T: Sync + Send + PreemptSchedInfo + FromTask<U>, U: Sync + Send + CommonSchedInfo> Scheduler<U>
     for PreemptScheduler<T, U>
 {
-    fn enqueue(&self, task: Arc<U>, flags: EnqueueFlags) -> Option<u32> {
+    fn enqueue(&self, task: Arc<U>, flags: EnqueueFlags) -> Option<CpuId> {
         let entity = PreemptSchedEntity::new(task);
         let mut still_in_rq = false;
         let target_cpu = {
@@ -88,7 +88,7 @@ impl<T: Sync + Send + PreemptSchedInfo + FromTask<U>, U: Sync + Send + CommonSch
             cpu_id
         };
 
-        let mut rq = self.rq[target_cpu as usize].disable_irq().lock();
+        let mut rq = self.rq[target_cpu.as_usize()].disable_irq().lock();
         if still_in_rq && let Err(_) = entity.task.cpu().set_if_is_none(target_cpu) {
             return None;
         }
@@ -105,14 +105,14 @@ impl<T: Sync + Send + PreemptSchedInfo + FromTask<U>, U: Sync + Send + CommonSch
 
     fn local_rq_with(&self, f: &mut dyn FnMut(&dyn LocalRunQueue<U>)) {
         let irq_guard = disable_local();
-        let local_rq: &PreemptRunQueue<T, U> = &self.rq[irq_guard.current_cpu() as usize].lock();
+        let local_rq: &PreemptRunQueue<T, U> = &self.rq[irq_guard.current_cpu().as_usize()].lock();
         f(local_rq);
     }
 
     fn local_mut_rq_with(&self, f: &mut dyn FnMut(&mut dyn LocalRunQueue<U>)) {
         let irq_guard = disable_local();
         let local_rq: &mut PreemptRunQueue<T, U> =
-            &mut self.rq[irq_guard.current_cpu() as usize].lock();
+            &mut self.rq[irq_guard.current_cpu().as_usize()].lock();
         f(local_rq);
     }
 }
diff --git a/kernel/src/thread/work_queue/mod.rs b/kernel/src/thread/work_queue/mod.rs
index 080d69ca6..284f97239 100644
--- a/kernel/src/thread/work_queue/mod.rs
+++ b/kernel/src/thread/work_queue/mod.rs
@@ -64,7 +64,7 @@
 //!
 //! ```
 
-use ostd::cpu::CpuSet;
+use ostd::cpu::{CpuId, CpuSet};
 use spin::Once;
 use work_item::WorkItem;
 use worker_pool::WorkerPool;
@@ -150,7 +150,7 @@ impl WorkQueue {
 
     /// Request a pending work item. The `request_cpu` indicates the CPU where
     /// the calling worker is located.
-    fn dequeue(&self, request_cpu: u32) -> Option<Arc<WorkItem>> {
+    fn dequeue(&self, request_cpu: CpuId) -> Option<Arc<WorkItem>> {
         let mut inner = self.inner.disable_irq().lock();
         let index = inner
             .pending_work_items
@@ -160,7 +160,7 @@ impl WorkQueue {
         Some(item)
     }
 
-    fn has_pending_work_items(&self, request_cpu: u32) -> bool {
+    fn has_pending_work_items(&self, request_cpu: CpuId) -> bool {
         self.inner
             .disable_irq()
             .lock()
diff --git a/kernel/src/thread/work_queue/work_item.rs b/kernel/src/thread/work_queue/work_item.rs
index ec083e4a8..c561343da 100644
--- a/kernel/src/thread/work_queue/work_item.rs
+++ b/kernel/src/thread/work_queue/work_item.rs
@@ -4,7 +4,7 @@
 
 use core::sync::atomic::{AtomicBool, Ordering};
 
-use ostd::cpu::CpuSet;
+use ostd::cpu::{CpuId, CpuSet};
 
 use crate::prelude::*;
 
@@ -33,7 +33,7 @@ impl WorkItem {
         &mut self.cpu_affinity
     }
 
-    pub(super) fn is_valid_cpu(&self, cpu_id: u32) -> bool {
+    pub(super) fn is_valid_cpu(&self, cpu_id: CpuId) -> bool {
         self.cpu_affinity.contains(cpu_id)
     }
 
diff --git a/kernel/src/thread/work_queue/worker.rs b/kernel/src/thread/work_queue/worker.rs
index ba1403632..685a68f94 100644
--- a/kernel/src/thread/work_queue/worker.rs
+++ b/kernel/src/thread/work_queue/worker.rs
@@ -2,7 +2,10 @@
 
 #![allow(dead_code)]
 
-use ostd::{cpu::CpuSet, task::Task};
+use ostd::{
+    cpu::{CpuId, CpuSet},
+    task::Task,
+};
 
 use super::worker_pool::WorkerPool;
 use crate::{
@@ -19,7 +22,7 @@ use crate::{
 pub(super) struct Worker {
     worker_pool: Weak<WorkerPool>,
     bound_task: Arc<Task>,
-    bound_cpu: u32,
+    bound_cpu: CpuId,
     inner: SpinLock<WorkerInner>,
 }
 
@@ -39,7 +42,7 @@ enum WorkerStatus {
 
 impl Worker {
     /// Creates a new `Worker` to the given `worker_pool`.
-    pub(super) fn new(worker_pool: Weak<WorkerPool>, bound_cpu: u32) -> Arc<Self> {
+    pub(super) fn new(worker_pool: Weak<WorkerPool>, bound_cpu: CpuId) -> Arc<Self> {
         Arc::new_cyclic(|worker_ref| {
             let weal_worker = worker_ref.clone();
             let task_fn = Box::new(move || {
diff --git a/kernel/src/thread/work_queue/worker_pool.rs b/kernel/src/thread/work_queue/worker_pool.rs
index b380c0f4a..6476e7281 100644
--- a/kernel/src/thread/work_queue/worker_pool.rs
+++ b/kernel/src/thread/work_queue/worker_pool.rs
@@ -7,7 +7,11 @@ use core::{
     time::Duration,
 };
 
-use ostd::{cpu::CpuSet, sync::WaitQueue, task::Task};
+use ostd::{
+    cpu::{CpuId, CpuSet},
+    sync::WaitQueue,
+    task::Task,
+};
 
 use super::{simple_scheduler::SimpleScheduler, worker::Worker, WorkItem, WorkPriority, WorkQueue};
 use crate::{
@@ -34,7 +38,7 @@ pub struct WorkerPool {
 
 /// A set of workers for a specific CPU.
 pub struct LocalWorkerPool {
-    cpu_id: u32,
+    cpu_id: CpuId,
     idle_wait_queue: WaitQueue,
     parent: Weak<WorkerPool>,
     /// A liveness check for LocalWorkerPool. The monitor periodically clears heartbeat,
@@ -66,7 +70,7 @@ pub struct Monitor {
 }
 
 impl LocalWorkerPool {
-    fn new(worker_pool: Weak<WorkerPool>, cpu_id: u32) -> Self {
+    fn new(worker_pool: Weak<WorkerPool>, cpu_id: CpuId) -> Self {
         LocalWorkerPool {
             cpu_id,
             idle_wait_queue: WaitQueue::new(),
@@ -151,7 +155,7 @@ impl WorkerPool {
         self.work_queues.disable_irq().lock().push(work_queue);
     }
 
-    pub fn has_pending_work_items(&self, request_cpu: u32) -> bool {
+    pub fn has_pending_work_items(&self, request_cpu: CpuId) -> bool {
         self.work_queues
             .disable_irq()
             .lock()
@@ -163,7 +167,7 @@ impl WorkerPool {
         self.scheduler.schedule();
     }
 
-    pub fn num_workers(&self, cpu_id: u32) -> u16 {
+    pub fn num_workers(&self, cpu_id: CpuId) -> u16 {
         self.local_pool(cpu_id).workers.disable_irq().lock().len() as u16
     }
 
@@ -171,7 +175,7 @@ impl WorkerPool {
         &self.cpu_set
     }
 
-    pub(super) fn fetch_pending_work_item(&self, request_cpu: u32) -> Option<Arc<WorkItem>> {
+    pub(super) fn fetch_pending_work_item(&self, request_cpu: CpuId) -> Option<Arc<WorkItem>> {
         for work_queue in self.work_queues.disable_irq().lock().iter() {
             let item = work_queue.dequeue(request_cpu);
             if item.is_some() {
@@ -181,22 +185,22 @@ impl WorkerPool {
         None
     }
 
-    fn local_pool(&self, cpu_id: u32) -> &Arc<LocalWorkerPool> {
+    fn local_pool(&self, cpu_id: CpuId) -> &Arc<LocalWorkerPool> {
         self.local_pools
             .iter()
             .find(|local_pool: &&Arc<LocalWorkerPool>| local_pool.cpu_id == cpu_id)
             .unwrap()
     }
 
-    pub(super) fn wake_worker(&self, cpu_id: u32) -> bool {
+    pub(super) fn wake_worker(&self, cpu_id: CpuId) -> bool {
         self.local_pool(cpu_id).wake_worker()
     }
 
-    pub(super) fn add_worker(&self, cpu_id: u32) {
+    pub(super) fn add_worker(&self, cpu_id: CpuId) {
         self.local_pool(cpu_id).add_worker();
     }
 
-    pub(super) fn remove_worker(&self, cpu_id: u32) {
+    pub(super) fn remove_worker(&self, cpu_id: CpuId) {
         self.local_pool(cpu_id).remove_worker();
     }
 
@@ -204,15 +208,15 @@ impl WorkerPool {
         self.priority == WorkPriority::High
     }
 
-    pub(super) fn heartbeat(&self, cpu_id: u32) -> bool {
+    pub(super) fn heartbeat(&self, cpu_id: CpuId) -> bool {
         self.local_pool(cpu_id).heartbeat()
     }
 
-    pub(super) fn set_heartbeat(&self, cpu_id: u32, heartbeat: bool) {
+    pub(super) fn set_heartbeat(&self, cpu_id: CpuId, heartbeat: bool) {
         self.local_pool(cpu_id).set_heartbeat(heartbeat)
     }
 
-    pub(super) fn idle_current_worker(&self, cpu_id: u32, worker: Arc<Worker>) {
+    pub(super) fn idle_current_worker(&self, cpu_id: CpuId, worker: Arc<Worker>) {
         self.local_pool(cpu_id).idle_current_worker(worker);
     }
 }
diff --git a/kernel/src/time/clocks/system_wide.rs b/kernel/src/time/clocks/system_wide.rs
index 12c6ef92d..a757030bf 100644
--- a/kernel/src/time/clocks/system_wide.rs
+++ b/kernel/src/time/clocks/system_wide.rs
@@ -4,13 +4,7 @@ use alloc::sync::Arc;
 use core::time::Duration;
 
 use aster_time::read_monotonic_time;
-use ostd::{
-    cpu::{num_cpus, PinCurrentCpu},
-    cpu_local,
-    sync::SpinLock,
-    task::disable_preempt,
-    timer::Jiffies,
-};
+use ostd::{cpu::PinCurrentCpu, cpu_local, sync::SpinLock, task::disable_preempt, timer::Jiffies};
 use paste::paste;
 use spin::Once;
 
@@ -232,7 +226,7 @@ macro_rules! define_timer_managers {
             $(
                 let clock = paste! {[<$clock_id _INSTANCE>].get().unwrap().clone()};
                 let clock_manager = TimerManager::new(clock);
-                for cpu in 0..num_cpus() {
+                for cpu in ostd::cpu::all_cpus() {
                     paste! {
                         [<$clock_id _MANAGER>].get_on_cpu(cpu).call_once(|| clock_manager.clone());
                     }
@@ -307,7 +301,7 @@ pub(super) fn init() {
 /// to avoid functions like this one.
 pub fn init_for_ktest() {
     // If `spin::Once` has initialized, this closure will not be executed.
-    for cpu in 0..num_cpus() {
+    for cpu in ostd::cpu::all_cpus() {
         CLOCK_REALTIME_MANAGER.get_on_cpu(cpu).call_once(|| {
             let clock = RealTimeClock { _private: () };
             TimerManager::new(Arc::new(clock))
diff --git a/ostd/src/arch/riscv/irq.rs b/ostd/src/arch/riscv/irq.rs
index 117257bb0..b476c59a4 100644
--- a/ostd/src/arch/riscv/irq.rs
+++ b/ostd/src/arch/riscv/irq.rs
@@ -8,6 +8,7 @@ use id_alloc::IdAlloc;
 use spin::Once;
 
 use crate::{
+    cpu::CpuId,
     sync::{Mutex, PreemptDisabled, SpinLock, SpinLockGuard},
     trap::TrapFrame,
 };
@@ -145,6 +146,6 @@ impl Drop for IrqCallbackHandle {
 ///
 /// The caller must ensure that the CPU ID and the interrupt number corresponds
 /// to a safe function to call.
-pub(crate) unsafe fn send_ipi(cpu_id: u32, irq_num: u8) {
+pub(crate) unsafe fn send_ipi(cpu_id: CpuId, irq_num: u8) {
     unimplemented!()
 }
diff --git a/ostd/src/arch/x86/irq.rs b/ostd/src/arch/x86/irq.rs
index 2d3f9a2c4..073c5d2aa 100644
--- a/ostd/src/arch/x86/irq.rs
+++ b/ostd/src/arch/x86/irq.rs
@@ -11,6 +11,7 @@ use spin::Once;
 use x86_64::registers::rflags::{self, RFlags};
 
 use crate::{
+    cpu::CpuId,
     sync::{Mutex, RwLock, RwLockReadGuard, SpinLock},
     trap::TrapFrame,
 };
@@ -160,11 +161,11 @@ impl Drop for IrqCallbackHandle {
 ///
 /// The caller must ensure that the CPU ID and the interrupt number corresponds
 /// to a safe function to call.
-pub(crate) unsafe fn send_ipi(cpu_id: u32, irq_num: u8) {
+pub(crate) unsafe fn send_ipi(cpu_id: CpuId, irq_num: u8) {
     use crate::arch::kernel::apic::{self, Icr};
 
     let icr = Icr::new(
-        apic::ApicId::from(cpu_id),
+        apic::ApicId::from(cpu_id.as_usize() as u32),
         apic::DestinationShorthand::NoShorthand,
         apic::TriggerMode::Edge,
         apic::Level::Assert,
diff --git a/ostd/src/cpu/local/cpu_local.rs b/ostd/src/cpu/local/cpu_local.rs
index f8548b451..98f9b0a84 100644
--- a/ostd/src/cpu/local/cpu_local.rs
+++ b/ostd/src/cpu/local/cpu_local.rs
@@ -5,7 +5,7 @@
 use core::{marker::Sync, ops::Deref};
 
 use super::{__cpu_local_end, __cpu_local_start};
-use crate::{arch, trap::DisabledLocalIrqGuard};
+use crate::{arch, cpu::CpuId, trap::DisabledLocalIrqGuard};
 
 /// Defines a CPU-local variable.
 ///
@@ -139,9 +139,11 @@ impl<T: 'static + Sync> CpuLocal<T> {
     /// # Panics
     ///
     /// Panics if the CPU ID is out of range.
-    pub fn get_on_cpu(&'static self, cpu_id: u32) -> &'static T {
+    pub fn get_on_cpu(&'static self, cpu_id: CpuId) -> &'static T {
         super::has_init::assert_true();
 
+        let cpu_id = cpu_id.as_usize();
+
         // If on the BSP, just use the statically linked storage.
         if cpu_id == 0 {
             return &self.0;
@@ -153,7 +155,7 @@ impl<T: 'static + Sync> CpuLocal<T> {
         let base = unsafe {
             super::CPU_LOCAL_STORAGES
                 .get_unchecked()
-                .get(cpu_id as usize - 1)
+                .get(cpu_id - 1)
                 .unwrap()
                 .start_paddr()
         };
diff --git a/ostd/src/cpu/local/mod.rs b/ostd/src/cpu/local/mod.rs
index 5ff528b8b..c91f8cd78 100644
--- a/ostd/src/cpu/local/mod.rs
+++ b/ostd/src/cpu/local/mod.rs
@@ -95,7 +95,7 @@ pub unsafe fn init_on_bsp() {
 
     let num_cpus = super::num_cpus();
 
-    let mut cpu_local_storages = Vec::with_capacity(num_cpus as usize - 1);
+    let mut cpu_local_storages = Vec::with_capacity(num_cpus - 1);
     for _ in 1..num_cpus {
         let ap_pages = {
             let nbytes = (bsp_end_va - bsp_base_va).align_up(PAGE_SIZE);
diff --git a/ostd/src/cpu/mod.rs b/ostd/src/cpu/mod.rs
index 1c50d15ad..d9be816ba 100644
--- a/ostd/src/cpu/mod.rs
+++ b/ostd/src/cpu/mod.rs
@@ -21,6 +21,37 @@ use crate::{
     trap::DisabledLocalIrqGuard,
 };
 
+/// The ID of a CPU in the system.
+///
+/// If converting from/to an integer, the integer must start from 0 and be less
+/// than the number of CPUs.
+#[derive(Debug, Clone, Copy, PartialEq, Eq)]
+pub struct CpuId(u32);
+
+impl CpuId {
+    /// Returns the CPU ID of the bootstrap processor (BSP).
+    pub const fn bsp() -> Self {
+        CpuId(0)
+    }
+
+    /// Converts the CPU ID to an `usize`.
+    pub const fn as_usize(self) -> usize {
+        self.0 as usize
+    }
+}
+
+impl TryFrom<usize> for CpuId {
+    type Error = &'static str;
+
+    fn try_from(value: usize) -> Result<Self, Self::Error> {
+        if value < num_cpus() {
+            Ok(CpuId(value as u32))
+        } else {
+            Err("The given CPU ID is out of range")
+        }
+    }
+}
+
 /// The number of CPUs.
 static NUM_CPUS: Once<u32> = Once::new();
 
@@ -46,14 +77,20 @@ pub(crate) unsafe fn set_this_cpu_id(id: u32) {
 }
 
 /// Returns the number of CPUs.
-pub fn num_cpus() -> u32 {
+pub fn num_cpus() -> usize {
     debug_assert!(
         NUM_CPUS.get().is_some(),
         "The number of CPUs is not initialized"
     );
     // SAFETY: The number of CPUs is initialized. The unsafe version is used
     // to avoid the overhead of the check.
-    unsafe { *NUM_CPUS.get_unchecked() }
+    let num = unsafe { *NUM_CPUS.get_unchecked() };
+    num as usize
+}
+
+/// Returns an iterator over all CPUs.
+pub fn all_cpus() -> impl Iterator<Item = CpuId> {
+    (0..num_cpus()).map(|id| CpuId(id as u32))
 }
 
 /// A marker trait for guard types that can "pin" the current task to the
@@ -70,10 +107,10 @@ pub fn num_cpus() -> u32 {
 /// CPU while any one of the instances of the implemented structure exists.
 pub unsafe trait PinCurrentCpu {
     /// Returns the number of the current CPU.
-    fn current_cpu(&self) -> u32 {
+    fn current_cpu(&self) -> CpuId {
         let id = CURRENT_CPU.load();
         debug_assert_ne!(id, u32::MAX, "This CPU is not initialized");
-        id
+        CpuId(id)
     }
 }
 
diff --git a/ostd/src/cpu/set.rs b/ostd/src/cpu/set.rs
index 01db7bb21..1865241b9 100644
--- a/ostd/src/cpu/set.rs
+++ b/ostd/src/cpu/set.rs
@@ -7,7 +7,7 @@ use core::sync::atomic::{AtomicU64, Ordering};
 use smallvec::SmallVec;
 use static_assertions::const_assert_eq;
 
-use super::num_cpus;
+use super::{num_cpus, CpuId};
 
 /// A subset of all CPUs in the system.
 #[derive(Clone, Debug, Default)]
@@ -21,12 +21,12 @@ type InnerPart = u64;
 const BITS_PER_PART: usize = core::mem::size_of::<InnerPart>() * 8;
 const NR_PARTS_NO_ALLOC: usize = 2;
 
-const fn part_idx(cpu_id: u32) -> usize {
-    cpu_id as usize / BITS_PER_PART
+const fn part_idx(cpu_id: CpuId) -> usize {
+    cpu_id.as_usize() / BITS_PER_PART
 }
 
-const fn bit_idx(cpu_id: u32) -> usize {
-    cpu_id as usize % BITS_PER_PART
+const fn bit_idx(cpu_id: CpuId) -> usize {
+    cpu_id.as_usize() % BITS_PER_PART
 }
 
 const fn parts_for_cpus(num_cpus: usize) -> usize {
@@ -36,18 +36,18 @@ const fn parts_for_cpus(num_cpus: usize) -> usize {
 impl CpuSet {
     /// Creates a new `CpuSet` with all CPUs in the system.
     pub fn new_full() -> Self {
-        let mut ret = Self::with_capacity_val(num_cpus() as usize, !0);
+        let mut ret = Self::with_capacity_val(num_cpus(), !0);
         ret.clear_nonexistent_cpu_bits();
         ret
     }
 
     /// Creates a new `CpuSet` with no CPUs in the system.
     pub fn new_empty() -> Self {
-        Self::with_capacity_val(num_cpus() as usize, 0)
+        Self::with_capacity_val(num_cpus(), 0)
     }
 
     /// Adds a CPU to the set.
-    pub fn add(&mut self, cpu_id: u32) {
+    pub fn add(&mut self, cpu_id: CpuId) {
         let part_idx = part_idx(cpu_id);
         let bit_idx = bit_idx(cpu_id);
         if part_idx >= self.bits.len() {
@@ -57,7 +57,7 @@ impl CpuSet {
     }
 
     /// Removes a CPU from the set.
-    pub fn remove(&mut self, cpu_id: u32) {
+    pub fn remove(&mut self, cpu_id: CpuId) {
         let part_idx = part_idx(cpu_id);
         let bit_idx = bit_idx(cpu_id);
         if part_idx < self.bits.len() {
@@ -66,7 +66,7 @@ impl CpuSet {
     }
 
     /// Returns true if the set contains the specified CPU.
-    pub fn contains(&self, cpu_id: u32) -> bool {
+    pub fn contains(&self, cpu_id: CpuId) -> bool {
         let part_idx = part_idx(cpu_id);
         let bit_idx = bit_idx(cpu_id);
         part_idx < self.bits.len() && (self.bits[part_idx] & (1 << bit_idx)) != 0
@@ -92,11 +92,12 @@ impl CpuSet {
     }
 
     /// Iterates over the CPUs in the set.
-    pub fn iter(&self) -> impl Iterator<Item = u32> + '_ {
+    pub fn iter(&self) -> impl Iterator<Item = CpuId> + '_ {
         self.bits.iter().enumerate().flat_map(|(part_idx, &part)| {
             (0..BITS_PER_PART).filter_map(move |bit_idx| {
                 if (part & (1 << bit_idx)) != 0 {
-                    Some((part_idx * BITS_PER_PART + bit_idx) as u32)
+                    let id = part_idx * BITS_PER_PART + bit_idx;
+                    Some(CpuId(id as u32))
                 } else {
                     None
                 }
@@ -113,7 +114,7 @@ impl CpuSet {
     }
 
     fn clear_nonexistent_cpu_bits(&mut self) {
-        let num_cpus = num_cpus() as usize;
+        let num_cpus = num_cpus();
         if num_cpus % BITS_PER_PART != 0 {
             let num_parts = parts_for_cpus(num_cpus);
             self.bits[num_parts - 1] &= (1 << (num_cpus % BITS_PER_PART)) - 1;
@@ -121,8 +122,8 @@ impl CpuSet {
     }
 }
 
-impl From<u32> for CpuSet {
-    fn from(cpu_id: u32) -> Self {
+impl From<CpuId> for CpuSet {
+    fn from(cpu_id: CpuId) -> Self {
         let mut set = Self::new_empty();
         set.add(cpu_id);
         set
@@ -171,7 +172,7 @@ impl AtomicCpuSet {
     }
 
     /// Atomically adds a CPU with the given ordering.
-    pub fn add(&self, cpu_id: u32, ordering: Ordering) {
+    pub fn add(&self, cpu_id: CpuId, ordering: Ordering) {
         let part_idx = part_idx(cpu_id);
         let bit_idx = bit_idx(cpu_id);
         if part_idx < self.bits.len() {
@@ -180,7 +181,7 @@ impl AtomicCpuSet {
     }
 
     /// Atomically removes a CPU with the given ordering.
-    pub fn remove(&self, cpu_id: u32, ordering: Ordering) {
+    pub fn remove(&self, cpu_id: CpuId, ordering: Ordering) {
         let part_idx = part_idx(cpu_id);
         let bit_idx = bit_idx(cpu_id);
         if part_idx < self.bits.len() {
@@ -189,7 +190,7 @@ impl AtomicCpuSet {
     }
 
     /// Atomically checks if the set contains the specified CPU.
-    pub fn contains(&self, cpu_id: u32, ordering: Ordering) -> bool {
+    pub fn contains(&self, cpu_id: CpuId, ordering: Ordering) -> bool {
         let part_idx = part_idx(cpu_id);
         let bit_idx = bit_idx(cpu_id);
         part_idx < self.bits.len() && (self.bits[part_idx].load(ordering) & (1 << bit_idx)) != 0
@@ -199,24 +200,23 @@ impl AtomicCpuSet {
 #[cfg(ktest)]
 mod test {
     use super::*;
-    use crate::prelude::*;
+    use crate::{cpu::all_cpus, prelude::*};
 
     #[ktest]
     fn test_full_cpu_set_iter_is_all() {
         let set = CpuSet::new_full();
         let num_cpus = num_cpus();
-        let all_cpus = (0..num_cpus).collect::<Vec<_>>();
+        let all_cpus = all_cpus().collect::<Vec<_>>();
         let set_cpus = set.iter().collect::<Vec<_>>();
 
-        assert!(set_cpus.len() == num_cpus as usize);
+        assert!(set_cpus.len() == num_cpus);
         assert_eq!(set_cpus, all_cpus);
     }
 
     #[ktest]
     fn test_full_cpu_set_contains_all() {
         let set = CpuSet::new_full();
-        let num_cpus = num_cpus();
-        for cpu_id in 0..num_cpus {
+        for cpu_id in all_cpus() {
             assert!(set.contains(cpu_id));
         }
     }
@@ -231,28 +231,29 @@ mod test {
     #[ktest]
     fn test_empty_cpu_set_contains_none() {
         let set = CpuSet::new_empty();
-        let num_cpus = num_cpus();
-        for cpu_id in 0..num_cpus {
+        for cpu_id in all_cpus() {
             assert!(!set.contains(cpu_id));
         }
     }
 
     #[ktest]
     fn test_atomic_cpu_set_multiple_sizes() {
-        for test_num_cpus in [1, 3, 12, 64, 96, 99, 128, 256, 288, 1024] {
+        for test_num_cpus in [1usize, 3, 12, 64, 96, 99, 128, 256, 288, 1024] {
+            let test_all_iter = || (0..test_num_cpus).map(|id| CpuId(id as u32));
+
             let set = CpuSet::with_capacity_val(test_num_cpus, 0);
             let atomic_set = AtomicCpuSet::new(set);
 
-            for cpu_id in 0..test_num_cpus as u32 {
+            for cpu_id in test_all_iter() {
                 assert!(!atomic_set.contains(cpu_id, Ordering::Relaxed));
-                if cpu_id % 3 == 0 {
+                if cpu_id.as_usize() % 3 == 0 {
                     atomic_set.add(cpu_id, Ordering::Relaxed);
                 }
             }
 
             let loaded = atomic_set.load();
             for cpu_id in loaded.iter() {
-                if cpu_id % 3 == 0 {
+                if cpu_id.as_usize() % 3 == 0 {
                     assert!(loaded.contains(cpu_id));
                 } else {
                     assert!(!loaded.contains(cpu_id));
@@ -261,7 +262,7 @@ mod test {
 
             atomic_set.store(CpuSet::with_capacity_val(test_num_cpus, 0));
 
-            for cpu_id in 0..test_num_cpus as u32 {
+            for cpu_id in test_all_iter() {
                 assert!(!atomic_set.contains(cpu_id, Ordering::Relaxed));
                 atomic_set.add(cpu_id, Ordering::Relaxed);
             }
diff --git a/ostd/src/mm/page_table/boot_pt.rs b/ostd/src/mm/page_table/boot_pt.rs
index 6cd36dfa3..39f01001a 100644
--- a/ostd/src/mm/page_table/boot_pt.rs
+++ b/ostd/src/mm/page_table/boot_pt.rs
@@ -68,7 +68,7 @@ where
 ///    another page table and before this dismissal.
 pub(crate) unsafe fn dismiss() {
     IS_DISMISSED.store(true);
-    if DISMISS_COUNT.fetch_add(1, Ordering::SeqCst) == num_cpus() - 1 {
+    if DISMISS_COUNT.fetch_add(1, Ordering::SeqCst) as usize == num_cpus() - 1 {
         BOOT_PAGE_TABLE.lock().take();
     }
 }
diff --git a/ostd/src/mm/vm_space.rs b/ostd/src/mm/vm_space.rs
index ccd211eb3..1bcf8179c 100644
--- a/ostd/src/mm/vm_space.rs
+++ b/ostd/src/mm/vm_space.rs
@@ -16,7 +16,7 @@ use core::{
 
 use crate::{
     arch::mm::{current_page_table_paddr, PageTableEntry, PagingConsts},
-    cpu::{num_cpus, CpuExceptionInfo, CpuSet, PinCurrentCpu},
+    cpu::{all_cpus, CpuExceptionInfo, CpuSet, PinCurrentCpu},
     cpu_local,
     mm::{
         io::Fallible,
@@ -93,7 +93,7 @@ impl VmSpace {
 
             let mut activated_cpus = CpuSet::new_empty();
 
-            for cpu in 0..num_cpus() {
+            for cpu in all_cpus() {
                 // The activation lock is held; other CPUs cannot activate this `VmSpace`.
                 let ptr =
                     ACTIVATED_VM_SPACE.get_on_cpu(cpu).load(Ordering::Relaxed) as *const VmSpace;
diff --git a/ostd/src/smp.rs b/ostd/src/smp.rs
index 910bb564a..7e3913ec9 100644
--- a/ostd/src/smp.rs
+++ b/ostd/src/smp.rs
@@ -72,7 +72,7 @@ fn do_inter_processor_call(_trapframe: &TrapFrame) {
     let mut queue = CALL_QUEUES.get_on_cpu(cur_cpu).lock();
     while let Some(f) = queue.pop_front() {
         log::trace!(
-            "Performing inter-processor call to {:#?} on CPU {}",
+            "Performing inter-processor call to {:#?} on CPU {:#?}",
             f,
             cur_cpu
         );
diff --git a/ostd/src/task/scheduler/fifo_scheduler.rs b/ostd/src/task/scheduler/fifo_scheduler.rs
index 28121e69c..78e0c45da 100644
--- a/ostd/src/task/scheduler/fifo_scheduler.rs
+++ b/ostd/src/task/scheduler/fifo_scheduler.rs
@@ -6,7 +6,7 @@ use super::{
     info::CommonSchedInfo, inject_scheduler, EnqueueFlags, LocalRunQueue, Scheduler, UpdateFlags,
 };
 use crate::{
-    cpu::{num_cpus, PinCurrentCpu},
+    cpu::{num_cpus, CpuId, PinCurrentCpu},
     sync::SpinLock,
     task::{disable_preempt, Task},
 };
@@ -25,7 +25,7 @@ struct FifoScheduler<T: CommonSchedInfo> {
 
 impl<T: CommonSchedInfo> FifoScheduler<T> {
     /// Creates a new instance of `FifoScheduler`.
-    fn new(nr_cpus: u32) -> Self {
+    fn new(nr_cpus: usize) -> Self {
         let mut rq = Vec::new();
         for _ in 0..nr_cpus {
             rq.push(SpinLock::new(FifoRunQueue::new()));
@@ -33,14 +33,14 @@ impl<T: CommonSchedInfo> FifoScheduler<T> {
         Self { rq }
     }
 
-    fn select_cpu(&self) -> u32 {
+    fn select_cpu(&self) -> CpuId {
         // FIXME: adopt more reasonable policy once we fully enable SMP.
-        0
+        CpuId::bsp()
     }
 }
 
 impl<T: CommonSchedInfo + Send + Sync> Scheduler<T> for FifoScheduler<T> {
-    fn enqueue(&self, runnable: Arc<T>, flags: EnqueueFlags) -> Option<u32> {
+    fn enqueue(&self, runnable: Arc<T>, flags: EnqueueFlags) -> Option<CpuId> {
         let mut still_in_rq = false;
         let target_cpu = {
             let mut cpu_id = self.select_cpu();
@@ -53,7 +53,7 @@ impl<T: CommonSchedInfo + Send + Sync> Scheduler<T> for FifoScheduler<T> {
             cpu_id
         };
 
-        let mut rq = self.rq[target_cpu as usize].disable_irq().lock();
+        let mut rq = self.rq[target_cpu.as_usize()].disable_irq().lock();
         if still_in_rq && let Err(_) = runnable.cpu().set_if_is_none(target_cpu) {
             return None;
         }
@@ -64,7 +64,7 @@ impl<T: CommonSchedInfo + Send + Sync> Scheduler<T> for FifoScheduler<T> {
 
     fn local_rq_with(&self, f: &mut dyn FnMut(&dyn LocalRunQueue<T>)) {
         let preempt_guard = disable_preempt();
-        let local_rq: &FifoRunQueue<T> = &self.rq[preempt_guard.current_cpu() as usize]
+        let local_rq: &FifoRunQueue<T> = &self.rq[preempt_guard.current_cpu().as_usize()]
             .disable_irq()
             .lock();
         f(local_rq);
@@ -72,7 +72,7 @@ impl<T: CommonSchedInfo + Send + Sync> Scheduler<T> for FifoScheduler<T> {
 
     fn local_mut_rq_with(&self, f: &mut dyn FnMut(&mut dyn LocalRunQueue<T>)) {
         let preempt_guard = disable_preempt();
-        let local_rq: &mut FifoRunQueue<T> = &mut self.rq[preempt_guard.current_cpu() as usize]
+        let local_rq: &mut FifoRunQueue<T> = &mut self.rq[preempt_guard.current_cpu().as_usize()]
             .disable_irq()
             .lock();
         f(local_rq);
diff --git a/ostd/src/task/scheduler/info.rs b/ostd/src/task/scheduler/info.rs
index f804613b8..dd007bf39 100644
--- a/ostd/src/task/scheduler/info.rs
+++ b/ostd/src/task/scheduler/info.rs
@@ -4,7 +4,7 @@
 
 use core::sync::atomic::{AtomicU32, Ordering};
 
-use crate::task::Task;
+use crate::{cpu::CpuId, task::Task};
 
 /// Fields of a task that OSTD will never touch.
 ///
@@ -28,17 +28,22 @@ impl AtomicCpuId {
     /// An `AtomicCpuId` with `AtomicCpuId::NONE` as its inner value is empty.
     const NONE: u32 = u32::MAX;
 
-    fn new(cpu_id: u32) -> Self {
-        Self(AtomicU32::new(cpu_id))
-    }
-
     /// Sets the inner value of an `AtomicCpuId` if it's empty.
     ///
     /// The return value is a result indicating whether the new value was written
-    /// and containing the previous value.
-    pub fn set_if_is_none(&self, cpu_id: u32) -> core::result::Result<u32, u32> {
+    /// and containing the previous value. If the previous value is empty, it returns
+    /// `Ok(())`. Otherwise, it returns `Err(previous_value)` which the previous
+    /// value is a valid CPU ID.
+    pub fn set_if_is_none(&self, cpu_id: CpuId) -> core::result::Result<(), CpuId> {
         self.0
-            .compare_exchange(Self::NONE, cpu_id, Ordering::Relaxed, Ordering::Relaxed)
+            .compare_exchange(
+                Self::NONE,
+                cpu_id.as_usize() as u32,
+                Ordering::Relaxed,
+                Ordering::Relaxed,
+            )
+            .map(|_| ())
+            .map_err(|prev| (prev as usize).try_into().unwrap())
     }
 
     /// Sets the inner value of an `AtomicCpuId` to `AtomicCpuId::NONE`, i.e. makes
@@ -48,19 +53,19 @@ impl AtomicCpuId {
     }
 
     /// Gets the inner value of an `AtomicCpuId`.
-    pub fn get(&self) -> Option<u32> {
+    pub fn get(&self) -> Option<CpuId> {
         let val = self.0.load(Ordering::Relaxed);
         if val == Self::NONE {
             None
         } else {
-            Some(val)
+            Some((val as usize).try_into().ok()?)
         }
     }
 }
 
 impl Default for AtomicCpuId {
     fn default() -> Self {
-        Self::new(Self::NONE)
+        Self(AtomicU32::new(Self::NONE))
     }
 }
 
diff --git a/ostd/src/task/scheduler/mod.rs b/ostd/src/task/scheduler/mod.rs
index 016d48532..2ac389b3b 100644
--- a/ostd/src/task/scheduler/mod.rs
+++ b/ostd/src/task/scheduler/mod.rs
@@ -13,7 +13,12 @@ use core::sync::atomic::{AtomicBool, Ordering};
 use spin::Once;
 
 use super::{preempt::cpu_local, processor, Task};
-use crate::{cpu::PinCurrentCpu, prelude::*, task::disable_preempt, timer};
+use crate::{
+    cpu::{CpuId, PinCurrentCpu},
+    prelude::*,
+    task::disable_preempt,
+    timer,
+};
 
 /// Injects a scheduler implementation into framework.
 ///
@@ -40,7 +45,7 @@ pub trait Scheduler<T = Task>: Sync + Send {
     ///
     /// If the `current` of a CPU needs to be preempted, this method returns the id of
     /// that CPU.
-    fn enqueue(&self, runnable: Arc<T>, flags: EnqueueFlags) -> Option<u32>;
+    fn enqueue(&self, runnable: Arc<T>, flags: EnqueueFlags) -> Option<CpuId>;
 
     /// Gets an immutable access to the local runqueue of the current CPU core.
     fn local_rq_with(&self, f: &mut dyn FnMut(&dyn LocalRunQueue<T>));