diff --git a/tools/blkdump.py b/tools/blkdump.py
new file mode 100755
index 0000000..793602f
--- /dev/null
+++ b/tools/blkdump.py
@@ -0,0 +1,59 @@
+#!/usr/bin/python
+
+# Read specified number of bytes in specified blkdev ID and spit it to stdout.
+# The proper blkdev has to be selected and placed already.
+
+import argparse
+import os
+import sys
+import time
+
+# Some place where it's safe to write 0xff bytes.
+MEMPTR = '9000'
+
+def sendcmd(fd, cmd):
+    # The serial link echoes back all typed characters and expects us to read
+    # them. We have to send each char one at a time.
+    print("Executing {}".format(cmd.decode()), file=sys.stderr)
+    for c in cmd:
+        os.write(fd, bytes([c]))
+        os.read(fd, 1)
+    os.write(fd, b'\n')
+    os.read(fd, 2)  # sends back \r\n
+
+
+def main():
+    parser = argparse.ArgumentParser()
+    parser.add_argument('device')
+    parser.add_argument('bytecount')
+    args = parser.parse_args()
+
+    try:
+        bytecount = int(args.bytecount, 16)
+    except ValueError:
+        print("bytecount has to be hexadecimal without prefix.")
+        return 1
+    fd = os.open(args.device, os.O_RDWR)
+    sendcmd(fd, 'mptr {}'.format(MEMPTR).encode())
+    os.read(fd, 9)
+    while bytecount > 0:
+        toread = min(bytecount, 0x100)
+        sendcmd(fd, 'load {:x}'.format(toread & 0xff).encode())
+        os.read(fd, 5)
+        sendcmd(fd, 'peek {:x}'.format(toread & 0xff).encode())
+        peek = b''
+        while len(peek) < toread * 2:
+            peek += os.read(fd, 1)
+            time.sleep(0.0001)
+        os.read(fd, 5)
+        while peek:
+            c = peek[:2]
+            sys.stdout.buffer.write(bytes([int(c, 16)]))
+            peek = peek[2:]
+        bytecount -= toread
+    os.close(fd)
+    return 0
+
+if __name__ == '__main__':
+    sys.exit(main())
+
diff --git a/tools/download.py b/tools/memdump.py
similarity index 100%
rename from tools/download.py
rename to tools/memdump.py