diff --git a/libraries/libmesosphere/source/arch/arm64/kern_cpu.cpp b/libraries/libmesosphere/source/arch/arm64/kern_cpu.cpp
index 43748f070..ec4097944 100644
--- a/libraries/libmesosphere/source/arch/arm64/kern_cpu.cpp
+++ b/libraries/libmesosphere/source/arch/arm64/kern_cpu.cpp
@@ -17,9 +17,14 @@
 
 namespace ams::kern::arm64::cpu {
 
+    /* Declare prototype to be implemented in asm. */
+    void SynchronizeAllCoresImpl(s32 *sync_var, s32 num_cores);
+
+
     namespace {
 
-        std::atomic<s32> g_all_core_sync_count;
+        /* Expose this as a global, for asm to use. */
+        s32 g_all_core_sync_count;
 
         void FlushEntireDataCacheImpl(int level) {
             /* Used in multiple locations. */
@@ -48,6 +53,14 @@ namespace ams::kern::arm64::cpu {
             }
         }
 
+        ALWAYS_INLINE void SetEventLocally() {
+            __asm__ __volatile__("sevl" ::: "memory");
+        }
+
+        ALWAYS_INLINE void WaitForEvent() {
+            __asm__ __volatile__("wfe" ::: "memory");
+        }
+
     }
 
     void FlushEntireDataCacheShared() {
@@ -70,19 +83,7 @@ namespace ams::kern::arm64::cpu {
     }
 
     NOINLINE void SynchronizeAllCores() {
-        /* Wait until the count can be read. */
-        while (!(g_all_core_sync_count < static_cast<s32>(cpu::NumCores))) { /* ... */ }
-
-        const s32 per_core_idx = g_all_core_sync_count.fetch_add(1);
-
-        /* Loop until it's our turn. This will act on each core in order. */
-        while (g_all_core_sync_count != per_core_idx + static_cast<s32>(cpu::NumCores)) { /* ... */ }
-
-        if (g_all_core_sync_count != 2 * static_cast<s32>(cpu::NumCores) - 1) {
-            g_all_core_sync_count++;
-        } else {
-            g_all_core_sync_count = 0;
-        }
+        SynchronizeAllCoresImpl(&g_all_core_sync_count, static_cast<s32>(cpu::NumCores));
     }
 
 }
diff --git a/libraries/libmesosphere/source/arch/arm64/kern_cpu_asm.s b/libraries/libmesosphere/source/arch/arm64/kern_cpu_asm.s
new file mode 100644
index 000000000..fbc000e8b
--- /dev/null
+++ b/libraries/libmesosphere/source/arch/arm64/kern_cpu_asm.s
@@ -0,0 +1,63 @@
+/*
+ * Copyright (c) 2018-2020 Atmosphère-NX
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+/* ams::kern::arm64::cpu::SynchronizeAllCoresImpl(int *sync_var, int num_cores) */
+.section    .text._ZN3ams4kern5arm643cpu23SynchronizeAllCoresImplEPii, "ax", %progbits
+.global     _ZN3ams4kern5arm643cpu23SynchronizeAllCoresImplEPii
+.type       _ZN3ams4kern5arm643cpu23SynchronizeAllCoresImplEPii, %function
+_ZN3ams4kern5arm643cpu23SynchronizeAllCoresImplEPii:
+    /* Loop until the sync var is less than num cores. */
+    sevl
+1:
+    wfe
+    ldaxr w2, [x0]
+    cmp   w2, w1
+    b.gt  1b
+
+    /* Increment the sync var. */
+2:
+    ldaxr w2, [x0]
+    add   w3, w2, #1
+    stlxr w4, w3, [x0]
+    cbnz  w4, 2b
+
+    /* Loop until the sync var matches our ticket. */
+    add   w3, w2, w1
+    sevl
+3:
+    wfe
+    ldaxr w2, [x0]
+    cmp   w2, w3
+    b.ne  3b
+
+    /* Check if the ticket is the last. */
+    sub w2, w1, #1
+    add w2, w2, w1
+    cmp   w3, w2
+    b.eq 5f
+
+    /* Our ticket is not the last one. Increment. */
+4:
+    ldaxr w2, [x0]
+    add   w3, w2, #1
+    stlxr w4, w3, [x0]
+    cbnz  w4, 4b
+    ret
+
+    /* Our ticket is the last one. */
+5:
+    stlr wzr, [x0]
+    ret