Metal Unary: Add benchmarks and process kernels in a tile based fashion (#2056)

* add basic unary bench for sqrt

* process unary commands in tiles of 4

* re-enable all benchmarks

* rename helper to unary

* modify approach to split up tiled and non-tiled operations

* undo bench ignore for other tests

* update tile size to 2

* only perform the optimization on the contiguous even numbered element case

candle-core/benches/benchmarks/mod.rs

@@ -3,6 +3,7 @@ pub(crate) mod conv_transpose2d;
 pub(crate) mod matmul;
 pub(crate) mod qmatmul;
 pub(crate) mod random;
+pub(crate) mod unary;
 pub(crate) mod where_cond;
 
 use candle_core::{Device, Result};

candle-core/benches/benchmarks/unary.rs

@@ -0,0 +1,49 @@
+use crate::benchmarks::{BenchDevice, BenchDeviceHandler};
+use candle_core::{DType, Device, Tensor};
+use criterion::{black_box, criterion_group, Criterion, Throughput};
+use std::time::Instant;
+
+fn run(a: &Tensor) {
+    a.sqrt().unwrap();
+}
+
+fn run_unary_benchmark(c: &mut Criterion, device: &Device, dtype: DType, name: &str) {
+    let b = 1;
+    let m = 1024;
+    let k = 1024;
+
+    let tensor = Tensor::arange(0.0f32, (b * m * k) as f32, &device)
+        .unwrap()
+        .to_dtype(dtype)
+        .unwrap()
+        .reshape((b, m, k))
+        .unwrap();
+
+    let flops = b * m * k * dtype.size_in_bytes();
+
+    let mut group = c.benchmark_group(device.bench_name(name));
+    group.throughput(Throughput::Bytes(flops as u64));
+    group.bench_function("iter", move |b| {
+        b.iter_custom(|iters| {
+            let start = Instant::now();
+            for _i in 0..iters {
+                run(black_box(&tensor));
+            }
+            device.sync().unwrap();
+            start.elapsed()
+        })
+    });
+    group.finish();
+}
+
+fn criterion_benchmark(c: &mut Criterion) {
+    let handler = BenchDeviceHandler::new().unwrap();
+    for device in handler.devices {
+        for dtype in [DType::F32, DType::BF16, DType::F16] {
+            let name = format!("sqrt_{:?}", dtype);
+            run_unary_benchmark(c, &device, dtype, &name);
+        }
+    }
+}
+
+criterion_group!(benches, criterion_benchmark);