diff --git a/crates/xenia-app/src/main.rs b/crates/xenia-app/src/main.rs
index 5a2fa01..e4765db 100644
--- a/crates/xenia-app/src/main.rs
+++ b/crates/xenia-app/src/main.rs
@@ -2138,7 +2138,7 @@ fn coord_pre_round(
     // is the guest-cycle timebase, not host_ns. This runs in `coord_pre_round`
     // which both the lockstep and parallel outer loops call every round.
     loop {
-        let now = kernel.scheduler.ctx(0).timebase;
+        let now = kernel.now_basis_at(0);
         let Some((r, reason)) = kernel.scheduler.advance_to_next_wake_if_due(now)
         else {
             break;
@@ -3146,6 +3146,16 @@ fn run_execution_parallel(
                                     .and_then(|t| guard.scheduler.find_by_tid(t))
                                     .unwrap_or(thread_ref);
                                 *guard.scheduler.ctx_mut_ref(target_ref) = ctx_taken;
+                                // Advance the parallel-mode coherent clock by
+                                // the instructions this block retired. This is
+                                // the single authoritative "now" the kernel
+                                // deadline-arithmetic reads in parallel mode
+                                // (per-thread `ctx.timebase` is incoherent here
+                                // because peers extract/zero their slots) —
+                                // keeping it monotonic breaks the timebase-
+                                // desync livelock where a woken thread re-armed
+                                // the same constant deadline forever.
+                                guard.scheduler.advance_global_clock(executed);
                                 // worker_epilogue's exit_current path
                                 // expects scheduler.current to be set
                                 // to the running thread.
diff --git a/crates/xenia-cpu/src/scheduler.rs b/crates/xenia-cpu/src/scheduler.rs
index 4cf926f..dfa9aa4 100644
--- a/crates/xenia-cpu/src/scheduler.rs
+++ b/crates/xenia-cpu/src/scheduler.rs
@@ -351,6 +351,19 @@ pub struct Scheduler {
     /// Sorted by deadline ascending. Scheduler wakes the first entry via
     /// `advance_to_next_wake` when a round finds nothing runnable.
     timed_waits: Vec<(u64, ThreadRef)>,
+    /// Parallel-mode coherent monotonic clock. In `--parallel`, workers
+    /// extract their `PpcContext` (leaving a zeroed timebase in the slot)
+    /// and step unlocked, so `ctx(hw_id).timebase` is NOT a coherent "now"
+    /// — a coordinator that reads it can see a stale/zero basis decoupled
+    /// from the deadline it just advanced to, re-arming the same constant
+    /// deadline forever (timebase-desync livelock). This field is the
+    /// single authoritative "now" the parallel coordinator and kernel
+    /// deadline-arithmetic read instead. Advanced by `advance_global_clock`
+    /// (per-block retired-instruction count) on each parallel writeback and
+    /// floored up by `advance_all_timebases_to`. LOCKSTEP never reads it
+    /// (gated by `KernelState::parallel_active`), so it has zero effect on
+    /// the deterministic lockstep trace.
+    global_clock: u64,
     /// Global count of TLS slots allocated — `spawn` pre-sizes new threads'
     /// `tls_values` to this.
     tls_slot_count: usize,
@@ -389,6 +402,7 @@ impl Scheduler {
             order,
             rng_state,
             timed_waits: Vec::new(),
+            global_clock: 0,
             tls_slot_count: 0,
             non_empty_runnable: 0,
             rotation_cursor: 0,
@@ -1114,6 +1128,29 @@ impl Scheduler {
                 }
             }
         }
+        // Keep the parallel-mode coherent clock at least as far forward as
+        // any deadline we fast-forward to (idle/timer/wake advances). This
+        // only mutates the new `global_clock` field — lockstep never reads
+        // it — so it cannot perturb the deterministic lockstep trace.
+        self.global_clock = self.global_clock.max(deadline);
+    }
+
+    /// Parallel-mode coherent "now" (see [`Self::global_clock`] field doc).
+    /// Read by the kernel deadline-arithmetic ONLY when
+    /// `KernelState::parallel_active`; lockstep keeps reading per-thread
+    /// `ctx(hw_id).timebase`.
+    #[inline]
+    pub fn global_clock(&self) -> u64 {
+        self.global_clock
+    }
+
+    /// Advance the parallel-mode coherent clock by `n` retired instructions.
+    /// Called from the parallel worker writeback with the block's executed
+    /// count so "now" tracks aggregate guest progress. Never called in
+    /// lockstep (the clock stays 0 and unread there).
+    #[inline]
+    pub fn advance_global_clock(&mut self, n: u64) {
+        self.global_clock = self.global_clock.saturating_add(n);
     }
 
     /// Fast-forward the timebase to the earliest pending timed wait and
diff --git a/crates/xenia-kernel/src/exports.rs b/crates/xenia-kernel/src/exports.rs
index 5243688..392401e 100644
--- a/crates/xenia-kernel/src/exports.rs
+++ b/crates/xenia-kernel/src/exports.rs
@@ -2165,7 +2165,7 @@ fn nt_set_timer_ex(ctx: &mut PpcContext, mem: &GuestMemory, state: &mut KernelSt
     // timebase separately (immutable borrow) before any mutation of the
     // object to keep the borrow-checker happy.
     let hw_id = state.scheduler.current_hw_id().unwrap_or(0);
-    let now = state.scheduler.ctx(hw_id).timebase;
+    let now = state.now_basis_at(hw_id);
 
     // Read signed i64 due_time (big-endian hi/lo — same pattern as
     // parse_timeout). Negative = relative-from-now, positive = absolute
@@ -3589,7 +3589,7 @@ pub(crate) fn parse_timeout(state: &KernelState, timeout_ptr: u32, mem: &GuestMe
         return Some(Some(0)); // poll
     }
     let hw_id = state.scheduler.current_hw_id().unwrap_or(0);
-    let now = state.scheduler.ctx(hw_id).timebase;
+    let now = state.now_basis_at(hw_id);
     // Negative = relative, positive = absolute wall-clock. Our timebase is a
     // plain instruction counter, so we treat all timeouts as "time-units
     // after now" regardless of sign, using the magnitude.
diff --git a/crates/xenia-kernel/src/state.rs b/crates/xenia-kernel/src/state.rs
index 290818c..9ccf0ed 100644
--- a/crates/xenia-kernel/src/state.rs
+++ b/crates/xenia-kernel/src/state.rs
@@ -1251,6 +1251,26 @@ impl KernelState {
         self.pending_timer_fires.first().map(|&(d, _)| d)
     }
 
+    /// Coherent "now" basis for deadline arithmetic, gated on execution mode.
+    ///
+    /// In **lockstep** (`parallel_active == false`) this returns exactly the
+    /// pre-existing per-thread `ctx(hw_id).timebase` each call site read
+    /// before, so the deterministic lockstep trace is byte-identical (no
+    /// golden re-baseline). In **parallel** (`parallel_active == true`) the
+    /// per-thread timebases are incoherent (workers extract/zero their slots
+    /// while stepping unlocked), so we return the scheduler's single
+    /// monotonic `global_clock` instead — the basis that breaks the
+    /// timebase-desync livelock. Callers pass the `hw_id` they would have
+    /// used for the lockstep `ctx()` read (slot 0 for coordinator-side
+    /// drains, the current thread's slot for in-guest waits).
+    pub fn now_basis_at(&self, hw_id: u8) -> u64 {
+        if self.parallel_active {
+            self.scheduler.global_clock()
+        } else {
+            self.scheduler.ctx(hw_id).timebase
+        }
+    }
+
     /// Fire every timer whose deadline is `<= now` (derived from slot 0's
     /// timebase, matching `parse_timeout`'s "current thread" fallback).
     /// For each fire: mark the timer `signaled=true`, clear its
@@ -1259,7 +1279,7 @@ impl KernelState {
     /// fired — the caller uses this to decide whether the scheduler round
     /// needs a follow-up `advance_to_next_wake_if_due` step.
     pub fn fire_due_timers(&mut self) -> bool {
-        let now = self.scheduler.ctx(0).timebase;
+        let now = self.now_basis_at(0);
         let mut fired = false;
         loop {
             let Some(&(deadline, handle)) = self.pending_timer_fires.first() else {