diff --git a/core/klibc.c b/core/klibc.c
index 1a928e2..e5c9537 100644
--- a/core/klibc.c
+++ b/core/klibc.c
@@ -16,14 +16,66 @@ int memcmp(const void *aptr, const void *bptr, size_t size)
     return 0;
 }
 
-void *memcpy(void *dst, const void *src, size_t n)
+// Inspirated by https://interrupt.memfault.com/blog/memcpy-newlib-nano
+/* Nonzero if either X or Y is not aligned on a "long" boundary.  */
+#define UNALIGNED(X, Y) \
+  (((long)X & (sizeof (long) - 1)) | ((long)Y & (sizeof (long) - 1)))
+
+/* How many bytes are copied each iteration of the 4X unrolled loop.  */
+#define BIGBLOCKSIZE    (sizeof (long) << 2)
+
+/* How many bytes are copied each iteration of the word copy loop.  */
+#define LITTLEBLOCKSIZE (sizeof (long))
+
+/* Threshhold for punting to the byte copier.  */
+#define TOO_SMALL(LEN)  ((LEN) < BIGBLOCKSIZE)
+
+void *memcpy(void *dst0, const void *src0, size_t len0)
 {
-    char *dstChar       = dst;
-    const char *srcChar = src;
-    for (size_t i = 0; i < n; i++) {
+#if 0
+    char *dstChar       = dst0;
+    const char *srcChar = src0;
+    for (size_t i = 0; i < len0; i++) {
         *(dstChar++) = *(srcChar++);
     }
-    return dst;
+    return dst0;
+#else
+    char *dst       = dst0;
+    const char *src = src0;
+    long *aligned_dst;
+    const long *aligned_src;
+
+    /* If the size is small, or either SRC or DST is unaligned,
+       then punt into the byte copy loop.  This should be rare.  */
+    if (!TOO_SMALL(len0) && !UNALIGNED(src, dst)) {
+        aligned_dst = (long *)dst;
+        aligned_src = (long *)src;
+
+        /* Copy 4X long words at a time if possible.  */
+        while (len0 >= BIGBLOCKSIZE) {
+            *aligned_dst++ = *aligned_src++;
+            *aligned_dst++ = *aligned_src++;
+            *aligned_dst++ = *aligned_src++;
+            *aligned_dst++ = *aligned_src++;
+            len0 -= BIGBLOCKSIZE;
+        }
+
+        /* Copy one long word at a time if possible.  */
+        while (len0 >= LITTLEBLOCKSIZE) {
+            *aligned_dst++ = *aligned_src++;
+            len0 -= LITTLEBLOCKSIZE;
+        }
+
+        /* Pick up any residual with a byte copier.  */
+        dst = (char *)aligned_dst;
+        src = (char *)aligned_src;
+    }
+
+    while (len0--)
+        *dst++ = *src++;
+
+    return dst0;
+#endif
 }
 
 void *memset(void *src, int c, size_t n)
diff --git a/core/time.c b/core/time.c
index 4bd7f42..b08b77d 100644
--- a/core/time.c
+++ b/core/time.c
@@ -27,3 +27,10 @@ unsigned long usecs_to_jiffies(const unsigned int u)
     // This could overflow
     return (u * HZ) / 1000000L;
 }
+
+#include <x86intrin.h>
+inline uint64_t read_cycle_counter()
+{
+        uint64_t tsc = __rdtsc();
+        return tsc;
+}
diff --git a/core/time.h b/core/time.h
index 699b33d..a5a4bdd 100644
--- a/core/time.h
+++ b/core/time.h
@@ -1,4 +1,5 @@
 #pragma once
+#include <stdint.h>
 
 #define HZ 100
 /*
@@ -50,3 +51,5 @@ unsigned int jiffies_to_msecs(const unsigned long j);
 unsigned int jiffies_to_usecs(const unsigned long j);
 unsigned long msecs_to_jiffies(const unsigned int m);
 unsigned long usecs_to_jiffies(const unsigned int u);
+
+uint64_t read_cycle_counter();
diff --git a/tests/test.c b/tests/test.c
index 2a87674..b1a11c1 100644
--- a/tests/test.c
+++ b/tests/test.c
@@ -11,6 +11,29 @@
 #include "synchro.h"
 #include "time.h"
 
+void testMemcpyPerf()
+{
+    struct test_struct {
+        char data[4096];
+    };
+    // instantiate 2 structs. for our purposes, we don't care what data is in
+    // there. set them to `volatile` so the compiler won't optimize away what we
+    // do with them
+    volatile struct test_struct dest, source;
+
+    printf("Test Memcpy perf\n");
+    // run through powers-of-two memcpy's, printing stats for each test
+    for (size_t len = 1; len <= sizeof(dest); len <<= 1) {
+        uint32_t start = read_cycle_counter(); // << Start count
+        memcpy((void *)&dest, (void *)&source, len);
+        uint32_t stop = read_cycle_counter(); // << Stop count
+
+        // print out the cycles consumed
+        printf("len = %d, %d %d cyccnt = %d, cycles/byte = %d\n", (uint32_t)len, stop, start,
+               stop - start, (stop - start) / len);
+    }
+}
+
 void testPhymem(void)
 {
     printf("Testing memory PHY\n");
@@ -47,7 +70,8 @@ void testPhymem(void)
     assert(freePageStatFree == freePageStatBegin);
     assert(usedPageStatFree == usedPageStatBegin);
 
-    assertmsg((page = (struct phyMemDesc *)allocPhyPage(1)) != NULL, "Cannot allocate memory\n");
+    assertmsg((page = (struct phyMemDesc *)allocPhyPage(1)) != NULL,
+              "Cannot allocate memory\n");
     unrefPhyPage((ulong)page);
 }
 
@@ -137,7 +161,8 @@ static void testPaging(void)
     }
     printf("%d pages freed\n", freeCount);
 
-    assertmsg((page = (struct phyMemDesc *)allocPhyPage(1)) != NULL, "Cannot allocate memory\n");
+    assertmsg((page = (struct phyMemDesc *)allocPhyPage(1)) != NULL,
+              "Cannot allocate memory\n");
     unrefPhyPage((ulong)page);
 }
 
@@ -315,6 +340,7 @@ void testKthread()
 
 void run_test(void)
 {
+    testMemcpyPerf();
     {
         int test             = 1000;
         long long int test64 = 0x100000000;