redesign for faster cpu/gpu synch (#1869)

* redesign for faster cpu/gpu synch

* load + more async CPU

* use command encoder API and move more ops to use it

* make fence back-end generic + CPU only fence

* faster build

* fix async eval

* fixes + handle temporaries

* fix / improve cpu conv

* remove unused status, fix siblings

* fix extensions

* fix

* fix no cpu build

* format

* comments

* fix perf regression, remove unecessary abort

* fix events, task limit cpu

* fix waiting

* fix donation / temporaries in normalization

mlx/backend/cpu/gemm.h

@@ -7,14 +7,20 @@ namespace mlx::core {
 
 template <typename T>
 void matmul(
-    const array& a,
-    const array& b,
-    array& out,
+    const T* a,
+    const T* b,
+    T* out,
     bool a_transposed,
     bool b_transposed,
     size_t lda,
     size_t ldb,
+    size_t ldc,
     float alpha,
-    float beta);
+    float beta,
+    size_t batch_size,
+    const Shape& a_shape,
+    const Strides& a_strides,
+    const Shape& b_shape,
+    const Strides& b_strides);
 
 } // namespace mlx::core