Set up cava benchmark

• Set up cava benchmark w/ criterion.
• GPU schedule for cava.
• Convert a bunch of benchmarks from using hercules refs to use hercules boxes.
• Define AlignedAlloc global allocator in hercules_rt that aligns everything to 32 bytes.
• Bump nom version.

juno_samples/matmul/src/main.rs

@@ -41,7 +41,9 @@ fn main() {
                 .await;
             let mut c_cpu: Box<[f32]> = vec![0.0; correct_c.len()].into_boxed_slice();
             c.to_cpu_ref(&mut c_cpu);
-            assert!(zip(c_cpu, correct_c).all(|(calc, correct)| (calc - correct).abs() < 0.00001));
+            for (calc, correct) in zip(c_cpu, correct_c) {
+                assert!((calc - correct).abs() < 0.00001, "{} != {}", calc, correct);
+            }
         }
     });
 }