diff --git a/libafl/src/bolts/cpu.rs b/libafl/src/bolts/cpu.rs
index 42d26448ea..eb2c4eea38 100644
--- a/libafl/src/bolts/cpu.rs
+++ b/libafl/src/bolts/cpu.rs
@@ -1,6 +1,9 @@
 //! Architecture agnostic processor features
 
-#[cfg(not(any(target_arch = "x86_64", target_arch = "x86")))]
+#[cfg(target_arch = "aarch64")]
+use core::arch::asm;
+
+#[cfg(not(any(target_arch = "x86_64", target_arch = "x86", target_arch = "aarch64")))]
 use crate::bolts::current_nanos;
 
 // TODO: Add more architectures, using C code, see
@@ -26,13 +29,28 @@ pub fn read_time_counter() -> u64 {
     }
 }
 
+/// Read a timestamp for measurements
+///
+/// Fetches the counter-virtual count register
+/// as we do not need to remove the cntvct_el2 offset.
+#[cfg(target_arch = "aarch64")]
+#[must_use]
+pub fn read_time_counter() -> u64 {
+    let v: u64 = 0;
+    unsafe {
+        // TODO pushing a change in core::arch::aarch64 ?
+        asm!("mrs {v}, cntvct_el0", v = out(reg) _);
+    }
+    v
+}
+
 /// Read a timestamp for measurements.
 ///
 /// This function is a wrapper around different ways to get a timestamp, fast.
 /// In this way, an experiment only has to
 /// change this implementation rather than every instead of [`read_time_counter`]
 /// On unsupported architectures, it's falling back to normal system time, in millis.
-#[cfg(not(any(target_arch = "x86_64", target_arch = "x86")))]
+#[cfg(not(any(target_arch = "x86_64", target_arch = "x86", target_arch = "aarch64")))]
 #[must_use]
 pub fn read_time_counter() -> u64 {
     current_nanos()