diff --git a/rpcs3/Emu/RSX/VK/VKGSRender.cpp b/rpcs3/Emu/RSX/VK/VKGSRender.cpp
index e928493f4b..f382061c58 100644
--- a/rpcs3/Emu/RSX/VK/VKGSRender.cpp
+++ b/rpcs3/Emu/RSX/VK/VKGSRender.cpp
@@ -509,7 +509,7 @@ VKGSRender::VKGSRender(utils::serial* ar) noexcept : GSRender(ar)
 	// This first set is bound persistently, so grow notifications are enabled.
 	m_attrib_ring_info.create(VK_BUFFER_USAGE_UNIFORM_TEXEL_BUFFER_BIT, VK_ATTRIB_RING_BUFFER_SIZE_M * 0x100000, vk::heap_pool_default, "attrib buffer", 0x400000, VK_TRUE);
 	m_fragment_env_ring_info.create(VK_BUFFER_USAGE_UNIFORM_BUFFER_BIT, VK_UBO_RING_BUFFER_SIZE_M * 0x100000, vk::heap_pool_low_latency, "fragment env buffer", 0x10000, VK_TRUE);
-	m_vertex_env_ring_info.create(VK_BUFFER_USAGE_STORAGE_BUFFER_BIT, VK_UBO_RING_BUFFER_SIZE_M * 0x100000, vk::heap_pool_low_latency, "vertex env buffer", 0x10000, VK_TRUE);
+	m_vertex_env_ring_info.create(VK_BUFFER_USAGE_UNIFORM_BUFFER_BIT, VK_UBO_RING_BUFFER_SIZE_M * 0x100000, vk::heap_pool_default, "vertex env buffer", 0x10000, VK_TRUE);
 	m_fragment_texture_params_ring_info.create(VK_BUFFER_USAGE_UNIFORM_BUFFER_BIT, VK_UBO_RING_BUFFER_SIZE_M * 0x100000, vk::heap_pool_low_latency, "fragment texture params buffer", 0x10000, VK_TRUE);
 	m_vertex_layout_ring_info.create(VK_BUFFER_USAGE_STORAGE_BUFFER_BIT, VK_UBO_RING_BUFFER_SIZE_M * 0x100000, vk::heap_pool_low_latency, "vertex layout buffer", 0x10000, VK_TRUE);
 	m_fragment_constants_ring_info.create(VK_BUFFER_USAGE_UNIFORM_BUFFER_BIT, VK_UBO_RING_BUFFER_SIZE_M * 0x100000, vk::heap_pool_low_latency, "fragment constants buffer", 0x10000, VK_TRUE);
@@ -563,6 +563,11 @@ VKGSRender::VKGSRender(utils::serial* ar) noexcept : GSRender(ar)
 	const auto& limits = m_device->gpu().get_limits();
 	m_texbuffer_view_size = std::min(limits.maxTexelBufferElements, VK_ATTRIB_RING_BUFFER_SIZE_M * 0x100000u);
 
+	// Initialize bulk allocators
+	m_vertex_env_allocator = std::make_unique<rsx::data_heap::bulk_allocator<256, 96>>(
+		m_vertex_env_ring_info,
+		std::min<u32>(limits.maxUniformBufferRange / 96u, 1024u));
+
 	if (m_texbuffer_view_size < 0x800000)
 	{
 		// Warn, only possibly expected on macOS
@@ -1951,7 +1956,9 @@ void VKGSRender::load_program_env()
 	if (update_vertex_env)
 	{
 		// Vertex state. Note, we're now on std430 alignment here, not hardware alignment.
-		const auto [mem, buf] = m_vertex_env_ring_info.alloc_and_map<16, char>(96);
+		// Use the bulk allocator here
+		const auto mem = m_vertex_env_allocator->alloc();
+		auto buf = m_vertex_env_ring_info.map<char>(mem, 96);
 
 		m_draw_processor.fill_scale_offset_data(buf, false);
 		m_draw_processor.fill_user_clip_data(buf + 64);
@@ -1962,6 +1969,9 @@ void VKGSRender::load_program_env()
 
 		m_vertex_env_ring_info.unmap();
 		m_vertex_env_dynamic_offset = mem;
+
+		m_vertex_env_buffer_info = m_vertex_env_ring_info.window<256>(m_vertex_env_dynamic_offset, 96, gpu_limits.maxUniformBufferRange);
+		m_vertex_env_dynamic_offset -= m_vertex_env_buffer_info.offset;
 	}
 
 	if (update_instancing_data)
@@ -2288,6 +2298,11 @@ void VKGSRender::patch_transform_constants(rsx::context* /*ctx*/, u32 index, u32
 
 	rsx::io_buffer iobuf(allocate_mem);
 	upload_transform_constants(iobuf);
+
+	if (!iobuf.empty())
+	{
+		m_transform_constants_ring_info.unmap();
+	}
 }
 
 void VKGSRender::init_buffers(rsx::framebuffer_creation_context context, bool)
diff --git a/rpcs3/Emu/RSX/VK/VKGSRender.h b/rpcs3/Emu/RSX/VK/VKGSRender.h
index da1491afcc..a8a6bb6745 100644
--- a/rpcs3/Emu/RSX/VK/VKGSRender.h
+++ b/rpcs3/Emu/RSX/VK/VKGSRender.h
@@ -161,6 +161,8 @@ private:
 	u64 m_texture_parameters_dynamic_offset = 0;
 	u64 m_stipple_array_dynamic_offset = 0;
 
+	std::unique_ptr<rsx::data_heap::bulk_allocator<256, 96>> m_vertex_env_allocator;
+
 	std::vector<vk::frame_context_t> m_frame_context_storage;
 	u32 m_max_async_frames = 0u;
 	// Temp frame context to use if the real frame queue is overburdened. Only used for storage
diff --git a/rpcs3/Emu/RSX/VK/VKVertexProgram.cpp b/rpcs3/Emu/RSX/VK/VKVertexProgram.cpp
index 375d35e942..0186045dd1 100644
--- a/rpcs3/Emu/RSX/VK/VKVertexProgram.cpp
+++ b/rpcs3/Emu/RSX/VK/VKVertexProgram.cpp
@@ -75,6 +75,8 @@ void VKVertexDecompilerThread::insertHeader(std::stringstream &OS)
 
 	OS <<
 		"#version 450\n\n"
+		"#extension GL_EXT_scalar_block_layout : require\n"
+		"#extension GL_EXT_uniform_buffer_unsized_array : require\n"
 		"#extension GL_ARB_separate_shader_objects : enable\n\n";
 
 	glsl::insert_subheader_block(OS);
@@ -88,7 +90,7 @@ void VKVertexDecompilerThread::insertHeader(std::stringstream &OS)
 		"#define get_user_clip_config() get_vertex_context().user_clip_configuration_bits\n\n";
 
 	OS <<
-		"layout(std430, set=0, binding=" << vk_prog->binding_table.context_buffer_location << ") readonly restrict buffer VertexContextBuffer\n"
+		"layout(std430, set=0, binding=" << vk_prog->binding_table.context_buffer_location << ") uniform VertexContextBuffer\n"
 		"{\n"
 		"	vertex_context_t vertex_contexts[];\n"
 		"};\n\n";
@@ -96,7 +98,7 @@ void VKVertexDecompilerThread::insertHeader(std::stringstream &OS)
 	const vk::glsl::program_input context_input
 	{
 		.domain = glsl::glsl_vertex_program,
-		.type = vk::glsl::input_type_storage_buffer,
+		.type = vk::glsl::input_type_uniform_buffer,
 		.set = vk::glsl::binding_set_index_vertex,
 		.location = vk_prog->binding_table.context_buffer_location,
 		.name = "VertexContextBuffer"