<!--
{
  "availability" : [
    "iOS: 10.3.0 -",
    "iPadOS: 10.3.0 -",
    "macCatalyst: 13.0.0 -",
    "macOS: 10.15.0 -",
    "tvOS: 10.2.0 -",
    "visionOS: 1.0.0 -"
  ],
  "documentType" : "symbol",
  "framework" : "Metal",
  "identifier" : "/documentation/Metal/MTLCommandBuffer/kernelStartTime",
  "metadataVersion" : "0.1.0",
  "role" : "Instance Property",
  "symbol" : {
    "kind" : "Instance Property",
    "modules" : [
      "Metal"
    ],
    "preciseIdentifier" : "c:objc(pl)MTLCommandBuffer(py)kernelStartTime"
  },
  "title" : "kernelStartTime"
}
-->

# kernelStartTime

The host time, in seconds, when the CPU begins to schedule the command buffer.

```
var kernelStartTime: CFTimeInterval { get }
```

## Discussion

You can calculate how much time the kernel spends scheduling a command buffer by subtracting this value from [`kernelEndTime`](/documentation/Metal/MTLCommandBuffer/kernelEndTime).

The kernel start and end times remain `0.0` until the GPU driver (on the CPU) schedules the command buffer to run on the GPU. Apps typically use these values after the [`waitUntilScheduled()`](/documentation/Metal/MTLCommandBuffer/waitUntilScheduled()) method returns, or within a completion handler (see [`addScheduledHandler(_:)`](/documentation/Metal/MTLCommandBuffer/addScheduledHandler(_:)) and [`addCompletedHandler(_:)`](/documentation/Metal/MTLCommandBuffer/addCompletedHandler(_:))).

---

Copyright &copy; 2026 Apple Inc. All rights reserved. | [Terms of Use](https://www.apple.com/legal/internet-services/terms/site.html) | [Privacy Policy](https://www.apple.com/privacy/privacy-policy)