2.4.2. Linux 09.02.00 Performance Guide¶
Read This First
All performance numbers provided in this document are gathered using following Evaluation Modules unless otherwise specified.
Name |
Description |
|---|---|
AM62Px SK |
AM62Px Starter Kit rev E1 with ARM running at 1.4GHz, DDR data rate 3200 MT/S |
Table: Evaluation Modules
About This Manual
This document provides performance data for each of the device drivers which are part of the Processor SDK Linux package. This document should be used in conjunction with release notes and user guides provided with the Processor SDK Linux package for information on specific issues present with drivers included in a particular release.
If You Need Assistance
For further information or to report any problems, contact http://e2e.ti.com/ or http://support.ti.com/
2.4.2.1. System Benchmarks¶
2.4.2.1.1. LMBench¶
LMBench is a collection of microbenchmarks of which the memory bandwidth and latency related ones are typically used to estimate processor memory system performance. More information about lmbench at http://lmbench.sourceforge.net/whatis_lmbench.html and http://lmbench.sourceforge.net/man/lmbench.8.html
Latency: lat_mem_rd-stride128-szN, where N is equal to or smaller than the cache size at given level measures the cache miss penalty. N that is at least double the size of last level cache is the latency to external memory.
Bandwidth: bw_mem_bcopy-N, where N is equal to or smaller than the cache size at a given level measures the achievable memory bandwidth from software doing a memcpy() type operation. Typical use is for external memory bandwidth calculation. The bandwidth is calculated as byte read and written counts as 1 which should be roughly half of STREAM copy result.
Execute the LMBench with the following:
cd /opt/ltp
./runltp -P j721e-idk-gw -f ddt/lmbench -s LMBENCH_L_PERF_0001
Benchmarks |
am62pxx_sk-fs: perf |
|---|---|
af_unix_sock_stream_latency (microsec) |
24.11 |
af_unix_socket_stream_bandwidth (MBs) |
1157.09 |
bw_file_rd-io-1mb (MB/s) |
1259.67 |
bw_file_rd-o2c-1mb (MB/s) |
700.16 |
bw_mem-bcopy-16mb (MB/s) |
1840.35 |
bw_mem-bcopy-1mb (MB/s) |
1998.40 |
bw_mem-bcopy-2mb (MB/s) |
1765.85 |
bw_mem-bcopy-4mb (MB/s) |
1814.33 |
bw_mem-bcopy-8mb (MB/s) |
1812.83 |
bw_mem-bzero-16mb (MB/s) |
7811.23 |
bw_mem-bzero-1mb (MB/s) |
4892.32 (min 1998.40, max 7786.23) |
bw_mem-bzero-2mb (MB/s) |
4781.56 (min 1765.85, max 7797.27) |
bw_mem-bzero-4mb (MB/s) |
4807.19 (min 1814.33, max 7800.04) |
bw_mem-bzero-8mb (MB/s) |
4808.86 (min 1812.83, max 7804.88) |
bw_mem-cp-16mb (MB/s) |
916.17 |
bw_mem-cp-1mb (MB/s) |
1241.66 (min 890.00, max 1593.31) |
bw_mem-cp-2mb (MB/s) |
1132.75 (min 886.66, max 1378.83) |
bw_mem-cp-4mb (MB/s) |
1168.97 (min 932.94, max 1404.99) |
bw_mem-cp-8mb (MB/s) |
1247.39 (min 954.54, max 1540.24) |
bw_mem-fcp-16mb (MB/s) |
1729.73 |
bw_mem-fcp-1mb (MB/s) |
4739.30 (min 1692.37, max 7786.23) |
bw_mem-fcp-2mb (MB/s) |
4699.41 (min 1601.54, max 7797.27) |
bw_mem-fcp-4mb (MB/s) |
4736.38 (min 1672.71, max 7800.04) |
bw_mem-fcp-8mb (MB/s) |
4712.00 (min 1619.11, max 7804.88) |
bw_mem-frd-16mb (MB/s) |
1871.13 |
bw_mem-frd-1mb (MB/s) |
1826.58 (min 1692.37, max 1960.78) |
bw_mem-frd-2mb (MB/s) |
1673.53 (min 1601.54, max 1745.51) |
bw_mem-frd-4mb (MB/s) |
1752.67 (min 1672.71, max 1832.62) |
bw_mem-frd-8mb (MB/s) |
1714.64 (min 1619.11, max 1810.16) |
bw_mem-fwr-16mb (MB/s) |
1548.59 |
bw_mem-fwr-1mb (MB/s) |
1777.05 (min 1593.31, max 1960.78) |
bw_mem-fwr-2mb (MB/s) |
1562.17 (min 1378.83, max 1745.51) |
bw_mem-fwr-4mb (MB/s) |
1618.81 (min 1404.99, max 1832.62) |
bw_mem-fwr-8mb (MB/s) |
1675.20 (min 1540.24, max 1810.16) |
bw_mem-rd-16mb (MB/s) |
16.00 |
bw_mem-rd-1mb (MB/s) |
2002.94 (min 1840.60, max 2165.28) |
bw_mem-rd-2mb (MB/s) |
1791.11 (min 1654.81, max 1927.40) |
bw_mem-rd-4mb (MB/s) |
1849.46 (min 1780.15, max 1918.77) |
bw_mem-rd-8mb (MB/s) |
1869.24 (min 1822.53, max 1915.94) |
bw_mem-rdwr-16mb (MB/s) |
1855.29 |
bw_mem-rdwr-1mb (MB/s) |
1285.97 (min 890.00, max 1681.93) |
bw_mem-rdwr-2mb (MB/s) |
1262.00 (min 886.66, max 1637.33) |
bw_mem-rdwr-4mb (MB/s) |
1297.38 (min 932.94, max 1661.82) |
bw_mem-rdwr-8mb (MB/s) |
1347.41 (min 954.54, max 1740.27) |
bw_mem-wr-16mb (MB/s) |
1831.08 |
bw_mem-wr-1mb (MB/s) |
1761.27 (min 1681.93, max 1840.60) |
bw_mem-wr-2mb (MB/s) |
1646.07 (min 1637.33, max 1654.81) |
bw_mem-wr-4mb (MB/s) |
1720.99 (min 1661.82, max 1780.15) |
bw_mem-wr-8mb (MB/s) |
1781.40 (min 1740.27, max 1822.53) |
bw_mmap_rd-mo-1mb (MB/s) |
2165.67 |
bw_mmap_rd-o2c-1mb (MB/s) |
646.73 |
bw_pipe (MB/s) |
717.53 |
bw_unix (MB/s) |
1157.09 |
lat_connect (us) |
52.27 |
lat_ctx-2-128k (us) |
4.79 |
lat_ctx-2-256k (us) |
3.86 |
lat_ctx-4-128k (us) |
4.51 |
lat_ctx-4-256k (us) |
3.49 |
lat_fs-0k (num_files) |
283.00 |
lat_fs-10k (num_files) |
124.00 |
lat_fs-1k (num_files) |
183.00 |
lat_fs-4k (num_files) |
165.00 |
lat_mem_rd-stride128-sz1000k (ns) |
31.26 |
lat_mem_rd-stride128-sz125k (ns) |
5.54 |
lat_mem_rd-stride128-sz250k (ns) |
5.83 |
lat_mem_rd-stride128-sz31k (ns) |
3.65 |
lat_mem_rd-stride128-sz50 (ns) |
2.15 |
lat_mem_rd-stride128-sz500k (ns) |
10.41 |
lat_mem_rd-stride128-sz62k (ns) |
5.19 |
lat_mmap-1m (us) |
56.00 |
lat_ops-double-add (ns) |
2.86 |
lat_ops-double-div (ns) |
15.73 |
lat_ops-double-mul (ns) |
2.86 |
lat_ops-float-add (ns) |
2.86 |
lat_ops-float-div (ns) |
9.30 |
lat_ops-float-mul (ns) |
2.86 |
lat_ops-int-add (ns) |
0.72 |
lat_ops-int-bit (ns) |
0.48 |
lat_ops-int-div (ns) |
4.29 |
lat_ops-int-mod (ns) |
4.53 |
lat_ops-int-mul (ns) |
3.07 |
lat_ops-int64-add (ns) |
0.72 |
lat_ops-int64-bit (ns) |
0.48 |
lat_ops-int64-div (ns) |
6.79 |
lat_ops-int64-mod (ns) |
5.25 |
lat_ops-int64-mul (ns) |
3.55 |
lat_pagefault (us) |
1.39 |
lat_pipe (us) |
19.48 |
lat_proc-exec (us) |
782.00 |
lat_proc-fork (us) |
606.11 |
lat_proc-proccall (us) |
0.01 |
lat_select (us) |
33.39 |
lat_sem (us) |
1.79 |
lat_sig-catch (us) |
5.28 |
lat_sig-install (us) |
0.66 |
lat_sig-prot (us) |
0.55 |
lat_syscall-fstat (us) |
2.73 |
lat_syscall-null (us) |
0.46 |
lat_syscall-open (us) |
149.03 |
lat_syscall-read (us) |
0.77 |
lat_syscall-stat (us) |
3.94 |
lat_syscall-write (us) |
0.64 |
lat_tcp (us) |
0.93 |
lat_unix (us) |
24.11 |
latency_for_0.50_mb_block_size (nanosec) |
10.41 |
latency_for_1.00_mb_block_size (nanosec) |
6.25 (min 0.00, max 31.26) |
pipe_bandwidth (MBs) |
717.53 |
pipe_latency (microsec) |
19.48 |
procedure_call (microsec) |
0.01 |
select_on_200_tcp_fds (microsec) |
33.39 |
semaphore_latency (microsec) |
1.79 |
signal_handler_latency (microsec) |
0.66 |
signal_handler_overhead (microsec) |
5.28 |
tcp_ip_connection_cost_to_localhost (microsec) |
52.27 |
tcp_latency_using_localhost (microsec) |
0.93 |
2.4.2.1.2. Dhrystone¶
Dhrystone is a core only benchmark that runs from warm L1 caches in all modern processors. It scales linearly with clock speed. For standard ARM cores the DMIPS/MHz score will be identical with the same compiler and flags.
Execute the benchmark with the following:
runDhrystone
Benchmarks |
am62pxx_sk-fs: perf |
|---|---|
cpu_clock (MHz) |
1400.00 |
dhrystone_per_mhz (DMIPS/MHz) |
2.90 |
dhrystone_per_second (DhrystoneP) |
7142857.00 |
2.4.2.1.3. Whetstone¶
Whetstone is a benchmark primarily measuring floating-point arithmetic performance.
Execute the benchmark with the following:
runWhetstone
Benchmarks |
am62pxx_sk-fs: perf |
|---|---|
whetstone (MIPS) |
5000.00 |
2.4.2.1.4. Linpack¶
Linpack measures peak double precision (64 bit) floating point performance in solving a dense linear system.
Benchmarks |
am62pxx_sk-fs: perf |
|---|---|
linpack (Kflops) |
580494.00 |
2.4.2.1.5. NBench¶
NBench which stands for Native Benchmark is used to measure macro benchmarks for commonly used operations such as sorting and analysis algorithms. More information about NBench at https://en.wikipedia.org/wiki/NBench and https://nbench.io/articles/index.html
Benchmarks |
am62pxx_sk-fs: perf |
|---|---|
assignment (Iterations) |
13.94 |
fourier (Iterations) |
22639.00 |
fp_emulation (Iterations) |
92.34 |
huffman (Iterations) |
1169.40 |
idea (Iterations) |
3444.90 |
lu_decomposition (Iterations) |
533.43 |
neural_net (Iterations) |
8.82 |
numeric_sort (Iterations) |
598.41 |
string_sort (Iterations) |
164.95 |
2.4.2.1.6. Stream¶
STREAM is a microbenchmark for measuring data memory system performance without any data reuse. It is designed to miss on caches and exercise data prefetcher and speculative accesses. It uses double precision floating point (64bit) but in most modern processors the memory access will be the bottleneck. The four individual scores are copy, scale as in multiply by constant, add two numbers, and triad for multiply accumulate. For bandwidth, a byte read counts as one and a byte written counts as one, resulting in a score that is double the bandwidth LMBench will show.
Execute the benchmark with the following:
stream_c
Benchmarks |
am62pxx_sk-fs: perf |
|---|---|
add (MB/s) |
2702.00 |
copy (MB/s) |
3765.90 |
scale (MB/s) |
3488.00 |
triad (MB/s) |
2402.50 |
2.4.2.1.7. CoreMarkPro¶
CoreMark®-Pro is a comprehensive, advanced processor benchmark that works with and enhances the market-proven industry-standard EEMBC CoreMark® benchmark. While CoreMark stresses the CPU pipeline, CoreMark-Pro tests the entire processor, adding comprehensive support for multicore technology, a combination of integer and floating-point workloads, and data sets for utilizing larger memory subsystems.
Benchmarks |
am62pxx_sk-fs: perf |
|---|---|
cjpeg-rose7-preset (workloads/) |
41.84 |
core (workloads/) |
0.30 |
coremark-pro () |
944.26 |
linear_alg-mid-100x100-sp (workloads/) |
14.70 |
loops-all-mid-10k-sp (workloads/) |
0.71 |
nnet_test (workloads/) |
1.09 |
parser-125k (workloads/) |
8.77 |
radix2-big-64k (workloads/) |
75.73 |
sha-test (workloads/) |
81.97 |
zip-test (workloads/) |
21.74 |
2.4.2.1.8. MultiBench¶
MultiBench™ is a suite of benchmarks that allows processor and system designers to analyze, test, and improve multicore processors. It uses three forms of concurrency: Data decomposition: multiple threads cooperating on achieving a unified goal and demonstrating a processor’s support for fine grain parallelism. Processing multiple data streams: uses common code running over multiple threads and demonstrating how well a processor scales over scalable data inputs. Multiple workload processing: shows the scalability of general-purpose processing, demonstrating concurrency over both code and data. MultiBench combines a wide variety of application-specific workloads with the EEMBC Multi-Instance-Test Harness (MITH), compatible and portable with most any multicore processors and operating systems. MITH uses a thread-based API (POSIX-compliant) to establish a common programming model that communicates with the benchmark through an abstraction layer and provides a flexible interface to allow a wide variety of thread-enabled workloads to be tested.
Benchmarks |
am62pxx_sk-fs: perf |
|---|---|
4m-check (workloads/) |
412.47 |
4m-check-reassembly (workloads/) |
128.37 |
4m-check-reassembly-tcp (workloads/) |
60.68 |
4m-check-reassembly-tcp-cmykw2-rotatew2 (workloads/) |
33.43 |
4m-check-reassembly-tcp-x264w2 (workloads/) |
1.95 |
4m-cmykw2 (workloads/) |
251.26 |
4m-cmykw2-rotatew2 (workloads/) |
50.34 |
4m-reassembly (workloads/) |
88.42 |
4m-rotatew2 (workloads/) |
53.45 |
4m-tcp-mixed (workloads/) |
120.30 |
4m-x264w2 (workloads/) |
2.01 |
idct-4m (workloads/) |
19.22 |
idct-4mw1 (workloads/) |
19.23 |
ippktcheck-4m (workloads/) |
413.29 |
ippktcheck-4mw1 (workloads/) |
412.95 |
ipres-4m (workloads/) |
112.28 |
ipres-4mw1 (workloads/) |
111.69 |
md5-4m (workloads/) |
28.96 |
md5-4mw1 (workloads/) |
29.38 |
rgbcmyk-4m (workloads/) |
65.94 |
rgbcmyk-4mw1 (workloads/) |
65.94 |
rotate-4ms1 (workloads/) |
23.58 |
rotate-4ms1w1 (workloads/) |
23.54 |
rotate-4ms64 (workloads/) |
23.84 |
rotate-4ms64w1 (workloads/) |
23.79 |
x264-4mq (workloads/) |
0.58 |
x264-4mqw1 (workloads/) |
0.58 |
2.4.2.2. Boot-time Measurement¶
2.4.2.2.1. Boot media: MMCSD¶
Boot Configuration |
am62pxx_sk-fs: boot time (sec) |
|---|---|
Kernel boot time test when bootloader, kernel and sdk-rootfs are in mmc-sd |
105.28 (min 15.02, max 128.07) |
Kernel boot time test when init is /bin/sh and bootloader, kernel and sdk-rootfs are in mmc-sd |
3.70 (min 3.67, max 3.77) |
2.4.2.3. Graphics SGX/RGX Driver¶
2.4.2.3.1. Glmark2¶
Run Glmark2 and capture performance reported (Score). All display outputs (HDMI, Displayport and/or LCD) are connected when running these tests
Benchmark |
am62pxx_sk-fs: Score |
|---|---|
Glmark2-DRM |
57.00 |
Glmark2-Wayland |
913.00 |
2.4.2.4. Ethernet¶
Ethernet performance benchmarks were measured using Netperf 2.7.1 https://hewlettpackard.github.io/netperf/doc/netperf.html Test procedures were modeled after those defined in RFC-2544: https://tools.ietf.org/html/rfc2544, where the DUT is the TI device and the “tester” used was a Linux PC. To produce consistent results, it is recommended to carry out performance tests in a private network and to avoid running NFS on the same interface used in the test. In these results, CPU utilization was captured as the total percentage used across all cores on the device, while running the performance test over one external interface.
UDP Throughput (0% loss) was measured by the procedure defined in RFC-2544 section 26.1: Throughput. In this scenario, netperf options burst_size (-b) and wait_time (-w) are used to limit bandwidth during different trials of the test, with the goal of finding the highest rate at which no loss is seen. For example, to limit bandwidth to 500Mbits/sec with 1472B datagram:
burst_size = <bandwidth (bits/sec)> / 8 (bits -> bytes) / <UDP datagram size> / 100 (seconds -> 10 ms)
burst_size = 500000000 / 8 / 1472 / 100 = 425
wait_time = 10 milliseconds (minimum supported by Linux PC used for testing)
UDP Throughput (possible loss) was measured by capturing throughput and packet loss statistics when running the netperf test with no bandwidth limit (remove -b/-w options).
In order to start a netperf client on one device, the other device must have netserver running. To start netserver:
netserver [-p <port_number>] [-4 (IPv4 addressing)] [-6 (IPv6 addressing)]
Running the following shell script from the DUT will trigger netperf clients to measure bidirectional TCP performance for 60 seconds and report CPU utilization. Parameter -k is used in client commands to summarize selected statistics on their own line and -j is used to gain additional timing measurements during the test.
#!/bin/bash
for i in 1
do
netperf -H <tester ip> -j -c -l 60 -t TCP_STREAM --
-k DIRECTION,THROUGHPUT,MEAN_LATENCY,LOCAL_CPU_UTIL,REMOTE_CPU_UTIL,LOCAL_BYTES_SENT,REMOTE_BYTES_RECVD,LOCAL_SEND_SIZE &
netperf -H <tester ip> -j -c -l 60 -t TCP_MAERTS --
-k DIRECTION,THROUGHPUT,MEAN_LATENCY,LOCAL_CPU_UTIL,REMOTE_CPU_UTIL,LOCAL_BYTES_SENT,REMOTE_BYTES_RECVD,LOCAL_SEND_SIZE &
done
Running the following commands will trigger netperf clients to measure UDP burst performance for 60 seconds at various burst/datagram sizes and report CPU utilization.
For UDP egress tests, run netperf client from DUT and start netserver on tester.
netperf -H <tester ip> -j -c -l 60 -t UDP_STREAM -b <burst_size> -w <wait_time> -- -m <UDP datagram size>
-k DIRECTION,THROUGHPUT,MEAN_LATENCY,LOCAL_CPU_UTIL,REMOTE_CPU_UTIL,LOCAL_BYTES_SENT,REMOTE_BYTES_RECVD,LOCAL_SEND_SIZE
For UDP ingress tests, run netperf client from tester and start netserver on DUT.
netperf -H <DUT ip> -j -C -l 60 -t UDP_STREAM -b <burst_size> -w <wait_time> -- -m <UDP datagram size>
-k DIRECTION,THROUGHPUT,MEAN_LATENCY,LOCAL_CPU_UTIL,REMOTE_CPU_UTIL,LOCAL_BYTES_SENT,REMOTE_BYTES_RECVD,LOCAL_SEND_SIZE
2.4.2.4.1. CPSW/CPSW2g/CPSW3g Ethernet Driver¶
CPSW3g: AM62px
TCP Bidirectional Throughput
Command Used |
am62pxx_sk-fs: THROUGHPUT (Mbits/sec) |
am62pxx_sk-fs: CPU Load % (LOCAL_CPU_UTIL) |
|---|---|---|
netperf -H 192.168.0.1 -j -c -C -l 60 -t TCP_STREAM; netperf -H 192.168.0.1 -j -c -C -l 60 -t TCP_MAERTS |
1800.0 |
59.22 |
Interrupt pacing value is kept as 250 microseconds for tx and rx interrupts.
TCP Bidirectional Throughput Interrupt Pacing
Command Used |
am62pxx_sk-fs: THROUGHPUT (Mbits/sec) |
am62pxx_sk-fs: CPU Load % (LOCAL_CPU_UTIL) |
|---|---|---|
netperf -H 192.168.0.1 -j -c -C -l 60 -t TCP_STREAM; netperf -H 192.168.0.1 -j -c -C -l 60 -t TCP_MAERTS |
1344.61 |
37.04 |
CPSW UDP Egress Throughput
Frame Size(bytes) |
am62pxx_sk-fs: UDP Datagram Size(bytes) (LOCAL_SEND_SIZE) |
am62pxx_sk-fs: THROUGHPUT (Mbits/sec) |
am62pxx_sk-fs: CPU Load %(LOCAL_CPU_UTIL) |
|---|---|---|---|
64 |
18.00 |
19.89 |
41.4 |
128 |
82.00 |
90.14 |
41.57 |
256 |
210.00 |
225.72 |
41.14 |
512 |
466.00 |
498.49 |
41.15 |
1024 |
978.00 |
936.62 |
42.7 |
1280 |
1234.00 |
947.78 |
33.1 |
1518 |
1472.00 |
955.52 |
33.21 |
CPSW UDP Ingress Throughput
Frame Size(bytes) |
am62pxx_sk-fs: UDP Datagram Size(bytes) (LOCAL_SEND_SIZE) |
am62pxx_sk-fs: THROUGHPUT (Mbits/sec) |
am62pxx_sk-fs: CPU Load %(LOCAL_CPU_UTIL) |
am62pxx_sk-fs: Packet loss % |
|---|---|---|---|---|
64 |
18.00 |
7.3 |
31.08 |
63.54 |
128 |
82.00 |
34.9 |
30.41 |
58.7 |
256 |
210.00 |
94.08 |
31.17 |
50.21 |
512 |
466.00 |
201.31 |
32.04 |
13.06 |
1024 |
978.00 |
916.22 |
40.04 |
0.02 |
1280 |
1234.00 |
930.71 |
39.42 |
0.0 |
1518 |
1472.00 |
916.74 |
38.17 |
0.0 |
2.4.2.5. EMMC Driver¶
Warning
IMPORTANT: The performance numbers can be severely affected if the media is mounted in sync mode. Hot plug scripts in the filesystem mount removable media in sync mode to ensure data integrity. For performance sensitive applications, umount the auto-mounted filesystem and re-mount in async mode.
2.4.2.5.1. AM62PXX-SK¶
Buffer size (bytes) |
am62pxx_sk-fs: Write EXT4 Throughput (Mbytes/sec) |
am62pxx_sk-fs: Write EXT4 CPU Load (%) |
am62pxx_sk-fs: Read EXT4 Throughput (Mbytes/sec) |
am62pxx_sk-fs: Read EXT4 CPU Load (%) |
|---|---|---|---|---|
1m |
89.80 |
1.21 |
285.00 |
2.24 |
4m |
95.70 |
0.94 |
286.00 |
1.76 |
4k |
79.20 |
26.99 |
92.80 |
25.24 |
256k |
90.30 |
1.71 |
284.00 |
3.43 |
2.4.2.6. UBoot EMMC Driver¶
2.4.2.6.1. AM62PXX-SK¶
Table: UBOOT EMMC RAW
File size (bytes in hex) |
am62pxx_sk-fs: Write Throughput (Kbytes/sec) |
am62pxx_sk-fs: Read Throughput (Kbytes/sec) |
|---|---|---|
2000000 |
95533.53 |
268590.16 |
4000000 |
98254.87 |
232397.16 |
2.4.2.7. MMC/SD Driver¶
Warning
IMPORTANT: The performance numbers can be severely affected if the media is mounted in sync mode. Hot plug scripts in the filesystem mount removable media in sync mode to ensure data integrity. For performance sensitive applications, umount the auto-mounted filesystem and re-mount in async mode.
2.4.2.7.1. AM62PXX-SK¶
Buffer size (bytes) |
am62pxx_sk-fs: Write EXT4 Throughput (Mbytes/sec) |
am62pxx_sk-fs: Write EXT4 CPU Load (%) |
am62pxx_sk-fs: Read EXT4 Throughput (Mbytes/sec) |
am62pxx_sk-fs: Read EXT4 CPU Load (%) |
|---|---|---|---|---|
1m |
42.70 |
0.81 |
87.50 |
1.05 |
4m |
43.40 |
0.60 |
87.30 |
0.81 |
4k |
3.60 |
1.64 |
16.40 |
5.05 |
256k |
39.60 |
1.00 |
85.50 |
1.24 |
The performance numbers were captured using the following:
SanDisk 8GB MicroSDHC Class 10 Memory Card
Partition was mounted with async option
2.4.2.8. UBoot MMC/SD Driver¶
2.4.2.8.1. AM62PXX-SK¶
Table: UBOOT MMCSD FAT
File size (bytes in hex) |
am62pxx_sk-fs: Write Throughput (Kbytes/sec) |
am62pxx_sk-fs: Read Throughput (Kbytes/sec) |
|---|---|---|
400000 |
37236.36 |
60235.29 |
800000 |
40960.00 |
72495.58 |
1000000 |
41583.76 |
81512.44 |
2.4.2.9. USB Driver¶
2.4.2.9.1. USB Device Controller¶
Table: USBDEVICE HIGHSPEED SLAVE_READ_THROUGHPUT
Number of Blocks |
am62pxx_sk-fs: Throughput (MB/sec) |
|---|---|
150 |
44.10 |
Table: USBDEVICE HIGHSPEED SLAVE_WRITE_THROUGHPUT
Number of Blocks |
am62pxx_sk-fs: Throughput (MB/sec) |
|---|---|
150 |
39.90 |
2.4.2.10. CRYPTO Driver¶
2.4.2.10.1. OpenSSL Performance¶
Table: OpenSSL Performance
Algorithm |
Buffer Size (in bytes) |
am62pxx_sk-fs: throughput (KBytes/Sec) |
|---|---|---|
aes-128-cbc |
1024 |
27031.21 |
aes-128-cbc |
16 |
552.76 |
aes-128-cbc |
16384 |
88069.46 |
aes-128-cbc |
256 |
8216.49 |
aes-128-cbc |
64 |
2187.29 |
aes-128-cbc |
8192 |
76024.49 |
aes-128-ecb |
1024 |
27399.85 |
aes-128-ecb |
16 |
573.77 |
aes-128-ecb |
16384 |
91215.19 |
aes-128-ecb |
256 |
8339.88 |
aes-128-ecb |
64 |
2251.84 |
aes-128-ecb |
8192 |
78157.14 |
aes-192-cbc |
1024 |
26395.65 |
aes-192-cbc |
16 |
553.54 |
aes-192-cbc |
16384 |
79527.94 |
aes-192-cbc |
256 |
8087.89 |
aes-192-cbc |
64 |
2193.92 |
aes-192-cbc |
8192 |
69823.15 |
aes-192-ecb |
1024 |
27090.60 |
aes-192-ecb |
16 |
570.65 |
aes-192-ecb |
16384 |
82482.52 |
aes-192-ecb |
256 |
8261.63 |
aes-192-ecb |
64 |
2266.73 |
aes-192-ecb |
8192 |
71778.30 |
aes-256-cbc |
1024 |
25508.18 |
aes-256-cbc |
16 |
555.50 |
aes-256-cbc |
16384 |
72783.19 |
aes-256-cbc |
256 |
7644.93 |
aes-256-cbc |
64 |
2168.23 |
aes-256-cbc |
8192 |
64143.36 |
aes-256-ecb |
1024 |
26324.31 |
aes-256-ecb |
16 |
568.96 |
aes-256-ecb |
16384 |
74880.34 |
aes-256-ecb |
256 |
8201.64 |
aes-256-ecb |
64 |
2240.98 |
aes-256-ecb |
8192 |
66183.17 |
sha256 |
1024 |
36728.83 |
sha256 |
16 |
613.09 |
sha256 |
16384 |
283989.33 |
sha256 |
256 |
9599.83 |
sha256 |
64 |
2425.77 |
sha256 |
8192 |
195665.92 |
sha512 |
1024 |
25829.72 |
sha512 |
16 |
600.75 |
sha512 |
16384 |
69058.56 |
sha512 |
256 |
8569.77 |
sha512 |
64 |
2403.75 |
sha512 |
8192 |
61964.29 |
Table: OpenSSL CPU Load
Algorithm |
am62pxx_sk-fs: CPU Load |
|---|---|
aes-128-cbc |
32.00 |
aes-128-ecb |
34.00 |
aes-192-cbc |
32.00 |
aes-192-ecb |
33.00 |
aes-256-cbc |
31.00 |
aes-256-ecb |
33.00 |
sha256 |
98.00 |
sha512 |
98.00 |
Listed for each algorithm are the code snippets used to run each benchmark test.
time -v openssl speed -elapsed -evp aes-128-cbc
2.4.2.11. Low Power Performance¶
Table: Deep sleep
Rail name |
Rail voltage(V) |
am62pxx_sk-fs |
|---|---|---|
vdd_core |
0.85 |
14.08 |
vddr_core |
0.85 |
1.24 |
soc_dvdd_3v3 |
3.30 |
5.36 |
soc_dvdd_1v8 |
1.80 |
2.87 |
vdda_1v8 |
1.80 |
9.26 |
vdd_lpddr4/vdd_ddr4 |
1.10 |
5.20 |
Total |
38.02 |
Table: MCU only
Rail name |
Rail voltage(V) |
am62pxx_sk-fs |
|---|---|---|
vdd_core |
0.85 |
202.22 |
vddr_core |
0.85 |
2.67 |
soc_dvdd_3v3 |
3.30 |
5.28 |
soc_dvdd_1v8 |
1.80 |
2.84 |
vdda_1v8 |
1.80 |
17.62 |
vdd_lpddr4/vdd_ddr4 |
1.10 |
3.36 |
Total |
233.99 |
Partial I/O Data - All voltage rails were measured to be near 0V
Further optimizations are possible for these low power modes. Please refer to the AM62x Power Consumption App Note (https://www.ti.com/lit/pdf/spradg1)