diff --git a/nx/include/switch/gfx/nvioctl.h b/nx/include/switch/gfx/nvioctl.h
index a109f9d6..735b622c 100644
--- a/nx/include/switch/gfx/nvioctl.h
+++ b/nx/include/switch/gfx/nvioctl.h
@@ -101,7 +101,7 @@ Result nvioctlNvmap_GetId(u32 fd, u32 nvmap_handle, u32 *id);
 Result nvioctlChannel_SetNvmapFd(u32 fd, u32 nvmap_fd);
 Result nvioctlChannel_SubmitGpfifo(u32 fd, nvioctl_gpfifo_entry *entries, u32 num_entries, u32 flags, nvioctl_fence *fence_out);
 Result nvioctlChannel_AllocObjCtx(u32 fd, u32 class_num, u32 flags);
-Result nvioctlChannel_ZCullBind(u32 fd, u32 in[4]);
+Result nvioctlChannel_ZCullBind(u32 fd, u64 gpu_va, u32 mode);
 Result nvioctlChannel_SetErrorNotifier(u32 fd, u64 offset, u64 size, u32 nvmap_handle);
 Result nvioctlChannel_SetPriority(u32 fd, u32 priority);
 Result nvioctlChannel_AllocGpfifoEx2(u32 fd, u32 num_entries, u32 flags, u32 unk0, u32 unk1, u32 unk2, u32 unk3, nvioctl_fence *fence_out);
diff --git a/nx/source/gfx/ioctl/channel.c b/nx/source/gfx/ioctl/channel.c
index 6bf88f28..e3fd4c4d 100644
--- a/nx/source/gfx/ioctl/channel.c
+++ b/nx/source/gfx/ioctl/channel.c
@@ -58,14 +58,16 @@ Result nvioctlChannel_AllocObjCtx(u32 fd, u32 class_num, u32 flags) {
     return nvIoctl(fd, _IOWR(0x48, 0x09, data), &data);
 }
 
-Result nvioctlChannel_ZCullBind(u32 fd, u32 in[4]) {
-    // Fixme: Needs work
+Result nvioctlChannel_ZCullBind(u32 fd, u64 gpu_va, u32 mode) {
     struct {
-        u32 in[4];
+        __in u64 gpu_va;
+        __in u32 mode;
+        __in u32 padding;
     } data;
 
     memset(&data, 0, sizeof(data));
-    memcpy(data.in, in, sizeof(data.in));
+    data.gpu_va = gpu_va;
+    data.mode = mode;
 
     return nvIoctl(fd, _IOWR(0x48, 0x0B, data), &data);
 }
diff --git a/nx/source/gfx/nvgfx.c b/nx/source/gfx/nvgfx.c
index 5e53a3ed..24367c62 100644
--- a/nx/source/gfx/nvgfx.c
+++ b/nx/source/gfx/nvgfx.c
@@ -38,6 +38,7 @@ size_t g_nvgfx_singleframebuf_size = /*0x3c0000*/ 1280*768*4;
 
 static nvmapobj nvmap_objs[18];
 
+static u64 nvmap_obj4_mapbuffer_x0_offset;
 static u64 nvmap_obj6_mapbuffer_xdb_offset;
 
 //Some of this struct is based on tegra_dc_ext_flip_windowattr.
@@ -129,7 +130,6 @@ Result nvgfxInitialize(void) {
     u64 *ptr64 = (u64*)g_gfxprod_BufferInitData;
     if(g_nvgfxInitialized)return 0;
 
-    u32 zcullbind_data[4] = {0x58000, 0x5, 0x2, 0x0};
     u32 framebuf_nvmap_handle = 0;//Special handle ID for framebuf/windowbuf.
 
     //nvioctl_gpfifo_entry gpfifo_entries[2] = {{0x00030000, 0x00177a05}, {0x00031778, 0x80002e05}};
@@ -155,6 +155,7 @@ Result nvgfxInitialize(void) {
     memset(&g_nvgfx_nvhostgpu_gpfifo_fence, 0, sizeof(g_nvgfx_nvhostgpu_gpfifo_fence));
     g_nvgfx_nvhostasgpu_allocspace_offset = 0;
     g_nvgfx_zcullctxsize = 0;
+    nvmap_obj4_mapbuffer_x0_offset = 0;
     nvmap_obj6_mapbuffer_xdb_offset = 0;
     g_nvgfx_nvhostctrl_eventres = 0;
 
@@ -246,10 +247,10 @@ Result nvgfxInitialize(void) {
 
     if (R_SUCCEEDED(rc)) rc = nvmapobjSetup(&nvmap_objs[4], 0, 0x1, 0x20000, 0);
 
-    if (R_SUCCEEDED(rc)) rc = nvioctlNvhostAsGpu_MapBufferEx(g_nvgfx_fd_nvhostasgpu, 4, 0, nvmap_objs[4].handle, 0x10000, 0, 0, 0, NULL);
+    if (R_SUCCEEDED(rc)) rc = nvioctlNvhostAsGpu_MapBufferEx(g_nvgfx_fd_nvhostasgpu, 4, 0, nvmap_objs[4].handle, 0x10000, 0, 0, 0, &nvmap_obj4_mapbuffer_x0_offset);
     if (R_SUCCEEDED(rc)) rc = nvioctlNvhostAsGpu_MapBufferEx(g_nvgfx_fd_nvhostasgpu, 4, 0xfe, nvmap_objs[4].handle, 0x10000, 0, 0, 0, NULL);
 
-    if (R_SUCCEEDED(rc)) rc = nvioctlChannel_ZCullBind(g_nvgfx_fd_nvhostgpu, zcullbind_data);
+    if (R_SUCCEEDED(rc)) rc = nvioctlChannel_ZCullBind(g_nvgfx_fd_nvhostgpu, nvmap_obj4_mapbuffer_x0_offset+0x8000, 0x2);
 
     //Officially, ipcQueryPointerBufferSize and NVGPU_IOCTL_CHANNEL_SUBMIT_GPFIFO(nvioctlChannel_SubmitGPFIFO) are used here with the duplicate service session setup during nv serv init.
     //TODO: This is probably used for GPU rendering? Is this really needed when not doing actual GPU rendering?