radek-pesina commented on issue #9080: URL: https://github.com/apache/nuttx/issues/9080#issuecomment-1569643272
Hi, Below are some measured speeds with the round-robin interval reduced to 10ms (was initially 200ms, however the improvement was marginal from reducing it to 10ms). NOTE: This is running on a slightly faster core than the previous measurements, hence why the initial baseline is ~100KiB/s instead of 75KiB/s. I believe you are more interested in the read/write speeds for eMMC without littlefs on it - I'm trying to DD from /dev/zero to /dev/mmcsd0 however I receive error 6 (no such device or address). Is there a method you suggest could be used for testing the transfer speed for eMMC? ``` // From block device (RAM) to block device (RAM) nsh> dd if=/dev/zero of=/dev/null bs=512 count=2000 1024000 bytes copied, 300000 usec, 3333 KB/s nsh> dd if=/dev/zero of=/dev/null bs=1024 count=2000 2048000 bytes copied, 340000 usec, 5882 KB/s nsh> dd if=/dev/zero of=/dev/null bs=4095 count=2000 8190000 bytes copied, 570000 usec, 14031 KB/s nsh> dd if=/dev/zero of=/dev/null bs=16384 count=2000 32768000 bytes copied, 1520000 usec, 21052 KB/s // From block device (RAM) to littleFS on (/sys - eMMC) nsh> dd if=/dev/zero of=/sys/t512 bs=512 count=2000 1024000 bytes copied, 9650000 usec, 103 KB/s nsh> dd if=/dev/zero of=/sys/t1024 bs=1024 count=100 102400 bytes copied, 680000 usec, 147 KB/s nsh> dd if=/dev/zero of=/sys/t4096 bs=4096 count=100 409600 bytes copied, 2970000 usec, 134 KB/s nsh> dd if=/dev/zero of=/sys/t16384 bs=16384 count=50 819200 bytes copied, 5940000 usec, 134 KB/s // from littleFS (on /sys - eMMC) to block device (RAM) nsh> dd if=/sys/t512 of=/dev/null bs=512 1024000 bytes copied, 9160000 usec, 109 KB/s nsh> dd if=/sys/t1024 of=/dev/null bs=1024 102400 bytes copied, 510000 usec, 196 KB/s nsh> dd if=/sys/t4096 of=/dev/null bs=4096 409600 bytes copied, 2800000 usec, 142 KB/s ``` -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
