Values are not correct when building NormalZMap texture - directx-11

Question

now I am trying to learn about DX11.
I am trying to practicing the SSAO effect.
When i try to build a NormalZmap, it looks quite weird.
Where I store the normals value and Z value in RGBA channel, but the pixel value inside the texture are not quite the same.
the output value of pixel shader
pixel value of texture
Looks like the Z value is passed well, but the normals value are messed up.
Here is my texture setting
D3D11_TEXTURE2D_DESC texDesc;
texDesc.Width = SCREEN_WIDTH;
texDesc.Height = SCREEN_HEIGHT;
texDesc.MipLevels = 1;
texDesc.ArraySize = 1;
texDesc.Format = DXGI_FORMAT_R32G32B32A32_FLOAT;
texDesc.SampleDesc.Count = 1;
texDesc.SampleDesc.Quality = 0;
texDesc.Usage = D3D11_USAGE_DEFAULT;
texDesc.BindFlags = D3D11_BIND_RENDER_TARGET | D3D11_BIND_SHADER_RESOURCE;
texDesc.CPUAccessFlags = 0;
texDesc.MiscFlags = 0;
GetDevice()->CreateTexture2D(&texDesc, NULL, &g_NormalZMap);
GetDevice()->CreateRenderTargetView(g_NormalZMap, NULL, &g_NormalZMapRTV);
GetDevice()->CreateShaderResourceView(g_NormalZMap, NULL, &g_NormalZMapSRV);
texDesc.Format = DXGI_FORMAT_D24_UNORM_S8_UINT;
texDesc.BindFlags = D3D11_BIND_DEPTH_STENCIL;
GetDevice()->CreateTexture2D(&texDesc, NULL, &g_NormalZMapDS);
GetDevice()->CreateDepthStencilView(g_NormalZMapDS, NULL, &g_NormalZMapDSV);
Here is my inputlayout setting
D3D11_INPUT_ELEMENT_DESC normalZMapLayout[] ={
{ "POSITION", 0, DXGI_FORMAT_R32G32B32_FLOAT, 0, D3D11_APPEND_ALIGNED_ELEMENT, D3D11_INPUT_PER_VERTEX_DATA, 0 },
{ "NORMAL", 0, DXGI_FORMAT_R32G32B32_FLOAT, 0, D3D11_APPEND_ALIGNED_ELEMENT, D3D11_INPUT_PER_VERTEX_DATA, 0 },
{ "COLOR", 0, DXGI_FORMAT_R32G32B32A32_FLOAT, 0, D3D11_APPEND_ALIGNED_ELEMENT, D3D11_INPUT_PER_VERTEX_DATA, 0 },
{ "TEXCOORD", 0, DXGI_FORMAT_R32G32_FLOAT, 0, D3D11_APPEND_ALIGNED_ELEMENT, D3D11_INPUT_PER_VERTEX_DATA, 0 },
};
UINT normalZMapNumElements = ARRAYSIZE(normalZMapLayout);
GetDevice()->CreateInputLayout(normalZMapLayout,
normalZMapNumElements,
pVSBlob->GetBufferPointer(),
pVSBlob->GetBufferSize(),
&g_InputLayoutSSAONormalZMap);
pVSBlob->Release();
pVSBlob = NULL;
Here is my HLSL code
cbuffer WorldBuffer : register(b0)
{
matrix World;
}
cbuffer ViewBuffer : register(b1)
{
matrix View;
}
cbuffer ProjectionBuffer : register(b2)
{
matrix Projection;
}
struct VSINPUT
{
float4 Position : POSITION0;
float4 Normal : NORMAL0;
float4 Diffuse : COLOR0;
float2 TexCoord : TEXCOORD0;
};
struct PSINPUT
{
float4 Position : SV_POSITION;
float4 Normal : NORMAL0;
float2 TexCoord : TEXCOORD0;
float4 ViewPos : POSITION0;
};
struct PSOUTPUT
{
float4 Diffuse : SV_Target;
};
PSINPUT NormalZMapVS(VSINPUT input)
{
PSINPUT output;
float4x4 wv = mul(World, View);
float4x4 wvp = mul(wv, Projection);
float4x4 invTransWv = transpose(inverse(wv));
output.Position = mul(input.Position, wvp);
output.Normal.xyz = normalize(mul(input.Normal.xyz, (float3x3)invTransWv));
output.TexCoord = input.TexCoord;
output.ViewPos = mul(input.Position, wv);
return output;
}
PSOUTPUT NormalZMapPS(PSINPUT input)
{
PSOUTPUT output;
float4 texColor = g_Texture.Sample(g_SamplerState, input.TexCoord);
clip(texColor.a - 0.1f);
float4 Out;
Out.xyz = input.Normal.xyz;
Out.w = input.ViewPos.z;
output.Diffuse = Out;
return output;
}