diff --git a/Ryujinx.Cpu/IVirtualMemoryManagerTracked.cs b/Ryujinx.Cpu/IVirtualMemoryManagerTracked.cs
index 23b4119f1..8004d39bc 100644
--- a/Ryujinx.Cpu/IVirtualMemoryManagerTracked.cs
+++ b/Ryujinx.Cpu/IVirtualMemoryManagerTracked.cs
@@ -8,6 +8,14 @@ namespace Ryujinx.Cpu
 {
     public interface IVirtualMemoryManagerTracked : IVirtualMemoryManager
     {
+        /// <summary>
+        /// Reads data from CPU mapped memory, with read tracking
+        /// </summary>
+        /// <typeparam name="T">Type of the data being read</typeparam>
+        /// <param name="va">Virtual address of the data in memory</param>
+        /// <returns>The data</returns>
+        T ReadTracked<T>(ulong va) where T : unmanaged;
+
         /// <summary>
         /// Writes data to CPU mapped memory, without write tracking.
         /// </summary>
diff --git a/Ryujinx.Cpu/MemoryManager.cs b/Ryujinx.Cpu/MemoryManager.cs
index 153bff629..2c11bab81 100644
--- a/Ryujinx.Cpu/MemoryManager.cs
+++ b/Ryujinx.Cpu/MemoryManager.cs
@@ -117,7 +117,7 @@ namespace Ryujinx.Cpu
         /// <inheritdoc/>
         public T Read<T>(ulong va) where T : unmanaged
         {
-            return MemoryMarshal.Cast<byte, T>(GetSpan(va, Unsafe.SizeOf<T>(), true))[0];
+            return MemoryMarshal.Cast<byte, T>(GetSpan(va, Unsafe.SizeOf<T>()))[0];
         }
 
         /// <inheritdoc/>
diff --git a/Ryujinx.Graphics.Gpu/Engine/MME/MacroHLE.cs b/Ryujinx.Graphics.Gpu/Engine/MME/MacroHLE.cs
index 77b44e814..05f3df0ee 100644
--- a/Ryujinx.Graphics.Gpu/Engine/MME/MacroHLE.cs
+++ b/Ryujinx.Graphics.Gpu/Engine/MME/MacroHLE.cs
@@ -2,8 +2,6 @@ using Ryujinx.Common.Logging;
 using Ryujinx.Graphics.Device;
 using Ryujinx.Graphics.GAL;
 using Ryujinx.Graphics.Gpu.Engine.GPFifo;
-using Ryujinx.Graphics.Gpu.Engine.Threed;
-using Ryujinx.Graphics.Gpu.Memory;
 using System;
 using System.Collections.Generic;
 
@@ -66,17 +64,42 @@ namespace Ryujinx.Graphics.Gpu.Engine.MME
             int arg2 = FetchParam().Word;
             int arg3 = FetchParam().Word;
 
-            int startOffset = arg0;
-            int endOffset = arg1;
+            int startDraw = arg0;
+            int endDraw = arg1;
             var topology = (PrimitiveTopology)arg2;
             int paddingWords = arg3;
-            int maxDrawCount = endOffset - startOffset;
             int stride = paddingWords * 4 + 0x14;
-            int indirectBufferSize = maxDrawCount * stride;
 
             ulong parameterBufferGpuVa = FetchParam().GpuVa;
-            ulong indirectBufferGpuVa = 0;
 
+            int maxDrawCount = endDraw - startDraw;
+
+            if (startDraw != 0)
+            {
+                int drawCount = _processor.MemoryManager.Read<int>(parameterBufferGpuVa, tracked: true);
+
+                // Calculate maximum draw count based on the previous draw count and current draw count.
+                if ((uint)drawCount <= (uint)startDraw)
+                {
+                    // The start draw is past our total draw count, so all draws were already performed.
+                    maxDrawCount = 0;
+                }
+                else
+                {
+                    // Perform just the missing number of draws.
+                    maxDrawCount = (int)Math.Min((uint)maxDrawCount, (uint)(drawCount - startDraw));
+                }
+            }
+
+            if (maxDrawCount == 0)
+            {
+                Fifo.Clear();
+                return;
+            }
+
+            int indirectBufferSize = maxDrawCount * stride;
+
+            ulong indirectBufferGpuVa = 0;
             int indexCount = 0;
 
             for (int i = 0; i < maxDrawCount; i++)
@@ -106,8 +129,10 @@ namespace Ryujinx.Graphics.Gpu.Engine.MME
             // It should be empty at this point, but clear it just to be safe.
             Fifo.Clear();
 
-            var parameterBuffer = _processor.MemoryManager.Physical.BufferCache.GetGpuBufferRange(_processor.MemoryManager, parameterBufferGpuVa, 4);
-            var indirectBuffer = _processor.MemoryManager.Physical.BufferCache.GetGpuBufferRange(_processor.MemoryManager, indirectBufferGpuVa, (ulong)indirectBufferSize);
+            var bufferCache = _processor.MemoryManager.Physical.BufferCache;
+
+            var parameterBuffer = bufferCache.GetGpuBufferRange(_processor.MemoryManager, parameterBufferGpuVa, 4);
+            var indirectBuffer = bufferCache.GetGpuBufferRange(_processor.MemoryManager, indirectBufferGpuVa, (ulong)indirectBufferSize);
 
             _processor.ThreedClass.MultiDrawIndirectCount(indexCount, topology, indirectBuffer, parameterBuffer, maxDrawCount, stride);
         }
diff --git a/Ryujinx.Graphics.Gpu/Memory/MemoryManager.cs b/Ryujinx.Graphics.Gpu/Memory/MemoryManager.cs
index b747b558f..2dc1edd24 100644
--- a/Ryujinx.Graphics.Gpu/Memory/MemoryManager.cs
+++ b/Ryujinx.Graphics.Gpu/Memory/MemoryManager.cs
@@ -63,10 +63,33 @@ namespace Ryujinx.Graphics.Gpu.Memory
         /// </summary>
         /// <typeparam name="T">Type of the data</typeparam>
         /// <param name="va">GPU virtual address where the data is located</param>
+        /// <param name="tracked">True if read tracking is triggered on the memory region</param>
         /// <returns>The data at the specified memory location</returns>
-        public T Read<T>(ulong va) where T : unmanaged
+        public T Read<T>(ulong va, bool tracked = false) where T : unmanaged
         {
-            return MemoryMarshal.Cast<byte, T>(GetSpan(va, Unsafe.SizeOf<T>()))[0];
+            int size = Unsafe.SizeOf<T>();
+
+            if (IsContiguous(va, size))
+            {
+                ulong address = Translate(va);
+
+                if (tracked)
+                {
+                    return Physical.ReadTracked<T>(address);
+                }
+                else
+                {
+                    return Physical.Read<T>(address);
+                }
+            }
+            else
+            {
+                Span<byte> data = new byte[size];
+
+                ReadImpl(va, data, tracked);
+
+                return MemoryMarshal.Cast<byte, T>(data)[0];
+            }
         }
 
         /// <summary>
diff --git a/Ryujinx.Graphics.Gpu/Memory/PhysicalMemory.cs b/Ryujinx.Graphics.Gpu/Memory/PhysicalMemory.cs
index 8df3c8fb3..fd2a74766 100644
--- a/Ryujinx.Graphics.Gpu/Memory/PhysicalMemory.cs
+++ b/Ryujinx.Graphics.Gpu/Memory/PhysicalMemory.cs
@@ -139,11 +139,22 @@ namespace Ryujinx.Graphics.Gpu.Memory
         /// Reads data from the application process.
         /// </summary>
         /// <typeparam name="T">Type of the structure</typeparam>
-        /// <param name="gpuVa">Address to read from</param>
+        /// <param name="address">Address to read from</param>
         /// <returns>The data at the specified memory location</returns>
         public T Read<T>(ulong address) where T : unmanaged
         {
-            return MemoryMarshal.Cast<byte, T>(GetSpan(address, Unsafe.SizeOf<T>()))[0];
+            return _cpuMemory.Read<T>(address);
+        }
+
+        /// <summary>
+        /// Reads data from the application process, with write tracking.
+        /// </summary>
+        /// <typeparam name="T">Type of the structure</typeparam>
+        /// <param name="address">Address to read from</param>
+        /// <returns>The data at the specified memory location</returns>
+        public T ReadTracked<T>(ulong address) where T : unmanaged
+        {
+            return _cpuMemory.ReadTracked<T>(address);
         }
 
         /// <summary>