Skip to main content
Apple silicon has a unified memory architecture. The CPU and GPU have direct access to the same memory pool. MLX is designed to take advantage of that. Concretely, when you make an array in MLX you don’t have to specify its location: 
a = mx.random.normal((100,))
b = mx.random.normal((100,))
Both a and b live in unified memory.

​
Operations on Any Device

In MLX, rather than moving arrays to devices, you specify the device when you run the operation. Any device can perform any operation on a and b without needing to move them from one memory location to another. For example: 
mx.add(a, b, stream=mx.cpu)
mx.add(a, b, stream=mx.gpu)
In the above, both the CPU and the GPU will perform the same add operation. The operations can (and likely will) be run in parallel since there are no dependencies between them.
See the Streams documentation for more information on the semantics of streams in MLX.

​
Automatic Dependency Management

In the above add example, there are no dependencies between operations, so there is no possibility for race conditions. If there are dependencies, the MLX scheduler will automatically manage them. For example: 
c = mx.add(a, b, stream=mx.cpu)
d = mx.add(a, c, stream=mx.gpu)
In the above case, the second add runs on the GPU but it depends on the output of the first add which is running on the CPU. MLX will automatically insert a dependency between the two streams so that the second add only starts executing after the first is complete and c is available.

​
A Simple Example

Here is a more interesting (albeit slightly contrived example) of how unified memory can be helpful. Suppose we have the following computation: 
def fun(a, b, d1, d2):
  x = mx.matmul(a, b, stream=d1)
  for _ in range(500):
      b = mx.exp(b, stream=d2)
  return x, b
which we want to run with the following arguments: 
a = mx.random.uniform(shape=(4096, 512))
b = mx.random.uniform(shape=(512, 4))
The first matmul operation is a good fit for the GPU since it’s more compute dense. The second sequence of operations are a better fit for the CPU, since they are very small and would probably be overhead bound on the GPU.
If we time the computation fully on the GPU, we get 2.8 milliseconds. But if we run the computation with d1=mx.gpu and d2=mx.cpu, then the time is only about 1.4 milliseconds, about twice as fast. These times were measured on an M1 Max.
By leveraging unified memory, MLX allows you to optimize performance by running different parts of your computation on the most appropriate device, all without explicit memory transfers.

Build docs developers (and LLMs) love

Get started for freeTalk to us