diff --git a/OVMF/OVMF_VARS_4M.fd b/OVMF/OVMF_VARS_4M.fd
index 3bad6f6..0268d2e 100644
Binary files a/OVMF/OVMF_VARS_4M.fd and b/OVMF/OVMF_VARS_4M.fd differ
diff --git a/fatroot/EFI/Boot/BootX64.efi b/fatroot/EFI/Boot/BootX64.efi
index 9d0d580..9b2dadd 100644
Binary files a/fatroot/EFI/Boot/BootX64.efi and b/fatroot/EFI/Boot/BootX64.efi differ
diff --git a/fatroot/video.bk2 b/fatroot/video.bk2
new file mode 100644
index 0000000..2503128
Binary files /dev/null and b/fatroot/video.bk2 differ
diff --git a/kernel32.cpp b/kernel32.cpp
index 143c97f..07d19bf 100644
--- a/kernel32.cpp
+++ b/kernel32.cpp
@@ -191,13 +191,27 @@ BOOL HeapFree(HANDLE, DWORD, LPVOID ptr) {
     return TRUE;
 }
 
+static void *memcpy(void *dest, const void *src, size_t n) {
+    // Cast the void pointers to char pointers so we can perform byte-wise arithmetic
+    unsigned char *d = (unsigned char *)dest;
+    const unsigned char *s = (const unsigned char *)src;
+
+    // Copy each byte from src to dest
+    for (size_t i = 0; i < n; i++) {
+        d[i] = s[i];
+    }
+
+    // Return the original destination pointer
+    return dest;
+}
+
 LPVOID HeapReAlloc(HANDLE, DWORD, LPVOID ptr, SIZE_T size) {
     PRINT("Called: HeapReAlloc with size: %ull and ptr: %p\n", size, ptr);
     if (!ptr) return Malloc(size);
     LPVOID newPtr = Malloc(size);
     if (!newPtr) return nullptr;
     // No original size info; can't fully emulate, user must handle
-    CopyMem(newPtr, ptr, size);
+    memcpy(newPtr, ptr, size);
     Free(ptr);
     return newPtr;
 }
@@ -223,6 +237,58 @@ DWORD WaitForSingleObject(HANDLE, DWORD) {
     return 0;
 }
 
+DWORD ResumeThread(HANDLE hThread) {
+    PRINT("Called: ResumeThread\n");
+    return 1;
+}
+
+BOOL SetThreadPriority(HANDLE hThread, int nPriority) {
+    PRINT("Called: SetThreadPriority\n");
+    return TRUE;
+}
+
+void InitializeCriticalSection(LPCRITICAL_SECTION lpCriticalSection) {
+    PRINT("Called: InitializeCriticalSection\n");
+}
+
+void DeleteCriticalSection(LPCRITICAL_SECTION lpCriticalSection) {
+    PRINT("Called: DeleteCriticalSection\n");
+}
+
+void EnterCriticalSection(LPCRITICAL_SECTION lpCriticalSection) {
+    PRINT("Called: EnterCriticalSection\n");
+}
+
+void LeaveCriticalSection(LPCRITICAL_SECTION lpCriticalSection) {
+    PRINT("Called: LeaveCriticalSection\n");
+}
+
+DWORD GetCurrentThreadId() {
+    PRINT("Called: GetCurrentThreadId\n");
+    return 1;
+}
+
+HANDLE CreateMutexA(LPSECURITY_ATTRIBUTES lpMutexAttributes, BOOL bInitialOwner, LPCTSTR lpName) {
+    PRINT("Called: CreateMutexA with name: lpName: %s\n", lpName);
+    return reinterpret_cast<HANDLE>(1);
+}
+
+BOOL ReleaseMutex(HANDLE hMutex) {
+    PRINT("Called: ReleaseMutex\n");
+    return TRUE;
+}
+
+HANDLE CreateSemaphoreA(LPSECURITY_ATTRIBUTES lpSemaphoreAttributes, LONG lInitialCount, LONG lMaximumCount, LPCTSTR lpName) {
+    PRINT("Called: CreateSemaphoreA with name: lpName: %s\n", lpName);
+    return reinterpret_cast<HANDLE>(1);
+}
+
+BOOL ReleaseSemaphore(HANDLE hSemaphore, LONG lReleaseCount, LPLONG lpPreviousCount) {
+    PRINT("Called: ReleaseSemaphore\n");
+    if (lpPreviousCount) *lpPreviousCount = 0;
+    return TRUE;
+}
+
 // --- Timing ---
 static inline unsigned long long read_tsc() {
     _mm_lfence();
@@ -282,7 +348,28 @@ BOOL QueryPerformanceCounter(LARGE_INTEGER* lpPerformanceCount)
 DWORD GetTickCount() {
     LARGE_INTEGER counter;
     QueryPerformanceCounter(&counter);
-    return (DWORD)(counter.QuadPart / 1000ULL); // ms
+    return (DWORD)counter.QuadPart;
+}
+
+void Sleep(DWORD dwMilliseconds) {
+    if (dwMilliseconds == 0) return;
+
+    // Ensure calibration has happened
+    if (!g_QpcInitialized) {
+        LARGE_INTEGER freq;
+        QueryPerformanceFrequency(&freq);
+    }
+
+    unsigned long long start = read_tsc();
+    
+    // Calculate how many total TSC ticks we need to wait
+    unsigned long long ticksToWait = (unsigned long long)dwMilliseconds * g_TicksPerMs;
+    
+    // Busy-wait until the TSC difference reaches the target
+    // We use _mm_pause() to be friendly to the CPU pipeline during the loop
+    while ((read_tsc() - start) < ticksToWait) {
+        _mm_pause(); 
+    }
 }
 
 // --- Module / File ---
@@ -448,12 +535,6 @@ DWORD GetModuleFileNameA(HMODULE hModule, LPSTR lpFilename, DWORD nSize) {
     return 0;
 }
 
-void Sleep(DWORD dwMilliseconds) {
-    PRINT("Called: Sleep\n");
-    UINTN microseconds = (UINTN)dwMilliseconds * 1000;
-    BS->Stall(microseconds);
-}
-
 DWORD SetErrorMode(DWORD uMode) {
     PRINT("Called: SetErrorMode\n");
     return 0;
@@ -527,60 +608,8 @@ DWORD_PTR SetThreadAffinityMask(HANDLE hThread, DWORD_PTR dwThreadAffinityMask)
     return 1;
 }
 
-DWORD ResumeThread(HANDLE hThread) {
-    PRINT("Called: ResumeThread\n");
-    return 1;
-}
-
-BOOL SetThreadPriority(HANDLE hThread, int nPriority) {
-    PRINT("Called: SetThreadPriority\n");
-    return TRUE;
-}
-
-void InitializeCriticalSection(LPCRITICAL_SECTION lpCriticalSection) {
-    PRINT("Called: InitializeCriticalSection\n");
-}
-
-void DeleteCriticalSection(LPCRITICAL_SECTION lpCriticalSection) {
-    PRINT("Called: DeleteCriticalSection\n");
-}
-
-void EnterCriticalSection(LPCRITICAL_SECTION lpCriticalSection) {
-    PRINT("Called: EnterCriticalSection\n");
-}
-
-void LeaveCriticalSection(LPCRITICAL_SECTION lpCriticalSection) {
-    PRINT("Called: LeaveCriticalSection\n");
-}
-
-DWORD GetCurrentThreadId() {
-    PRINT("Called: GetCurrentThreadId\n");
-    return 1;
-}
-
-HANDLE CreateMutexA(LPSECURITY_ATTRIBUTES lpMutexAttributes, BOOL bInitialOwner, LPCTSTR lpName) {
-    PRINT("Called: CreateMutexA with name: lpName: %s\n", lpName);
-    return reinterpret_cast<HANDLE>(1);
-}
-
-BOOL ReleaseMutex(HANDLE hMutex) {
-    PRINT("Called: ReleaseMutex\n");
-    return TRUE;
-}
-
-HANDLE CreateSemaphoreA(LPSECURITY_ATTRIBUTES lpSemaphoreAttributes, LONG lInitialCount, LONG lMaximumCount, LPCTSTR lpName) {
-    PRINT("Called: CreateSemaphoreA with name: lpName: %s\n", lpName);
-    return reinterpret_cast<HANDLE>(1);
-}
-
-BOOL ReleaseSemaphore(HANDLE hSemaphore, LONG lReleaseCount, LPLONG lpPreviousCount) {
-    PRINT("Called: ReleaseSemaphore\n");
-    if (lpPreviousCount) *lpPreviousCount = 0;
-    return TRUE;
-}
-
 BOOL TerminateProcess(HANDLE hProcess, UINT uExitCode) {
-    PRINT("Called: TerminateProcess\n");
+    PRINT("!! Called: TerminateProcess\n");
     return TRUE;
 }
 
diff --git a/main.cpp b/main.cpp
index 99fe54b..0c96603 100644
--- a/main.cpp
+++ b/main.cpp
@@ -467,16 +467,15 @@ EFI_STATUS efi_main(EFI_HANDLE ImageHandle, EFI_SYSTEM_TABLE* SystemTable) {
         }
         
         // Wait for the next frame
-        if (pBinkWait(bink)) {
+        while (pBinkWait(bink)) {
             _mm_pause(); // don't burn CPU cycles
-            continue; // wait for the next frame
         }
 
         // Check if we should skip the frame
-        if (pBinkShouldSkip(bink)) {
+        while (pBinkShouldSkip(bink)) {
             pBinkNextFrame(bink);
+            pBinkDoFrame(bink);
             skipped++;
-            continue; // skip this frame
         }
 
         int code = pBinkDoFrame(bink);
@@ -484,6 +483,7 @@ EFI_STATUS efi_main(EFI_HANDLE ImageHandle, EFI_SYSTEM_TABLE* SystemTable) {
             printf("BinkDoFrame failed!\n");
             pBinkClose(bink);
             Free(video_file);
+            Free(framebuffer);
             gBS->Stall(5 * 1000 * 1000);
             gST->RuntimeServices->ResetSystem(EfiResetShutdown, EFI_SUCCESS, 0, NULL);
             return 1;
@@ -503,6 +503,7 @@ EFI_STATUS efi_main(EFI_HANDLE ImageHandle, EFI_SYSTEM_TABLE* SystemTable) {
             printf("BinkCopyToBuffer failed!\n");
             pBinkClose(bink);
             Free(video_file);
+            Free(framebuffer);
             gBS->Stall(5 * 1000 * 1000);
             gST->RuntimeServices->ResetSystem(EfiResetShutdown, EFI_SUCCESS, 0, NULL);
             return 1;
diff --git a/video.bk2 b/video.bk2
new file mode 100644
index 0000000..2503128
Binary files /dev/null and b/video.bk2 differ