2.2. Performance Guide¶
2.2. Linux 09.00.00 Performance Guide¶
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/
2.2. System Benchmarks¶
2.2. 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 | am68_sk-fs: perf | am69_sk-fs: perf |
|---|---|---|
| af_unix_sock_stream_latency (microsec) | 15.11 | 14.06 |
| af_unix_socket_stream_bandwidth (MBs) | 1989.80 | 3409.24 |
| bw_file_rd-io-1mb (MB/s) | 2926.12 | 3621.00 |
| bw_file_rd-o2c-1mb (MB/s) | 1062.13 | 323.78 |
| bw_mem-bcopy-16mb (MB/s) | 3387.68 | 3129.89 |
| bw_mem-bcopy-1mb (MB/s) | 4347.83 | 9372.61 |
| bw_mem-bcopy-2mb (MB/s) | 3577.82 | 4456.00 |
| bw_mem-bcopy-4mb (MB/s) | 3381.23 | 3738.32 |
| bw_mem-bcopy-8mb (MB/s) | 3407.16 | 3273.77 |
| bw_mem-bzero-16mb (MB/s) | 10568.03 | 10862.19 |
| bw_mem-bzero-1mb (MB/s) | 8584.17 (min 4347.83, max 12820.51) | 11570.03 (min 9372.61, max 13767.44) |
| bw_mem-bzero-2mb (MB/s) | 7450.36 (min 3577.82, max 11322.89) | 9058.89 (min 4456.00, max 13661.77) |
| bw_mem-bzero-4mb (MB/s) | 7023.00 (min 3381.23, max 10664.77) | 8008.56 (min 3738.32, max 12278.80) |
| bw_mem-bzero-8mb (MB/s) | 6987.60 (min 3407.16, max 10568.03) | 7300.60 (min 3273.77, max 11327.43) |
| bw_mem-cp-16mb (MB/s) | 2119.77 | 2406.02 |
| bw_mem-cp-1mb (MB/s) | 7549.09 (min 2478.75, max 12619.43) | 8717.09 (min 3693.44, max 13740.73) |
| bw_mem-cp-2mb (MB/s) | 6702.99 (min 2089.50, max 11316.48) | 8218.82 (min 2743.95, max 13693.69) |
| bw_mem-cp-4mb (MB/s) | 6375.94 (min 2092.78, max 10659.09) | 7751.62 (min 2802.59, max 12700.65) |
| bw_mem-cp-8mb (MB/s) | 6336.49 (min 2090.96, max 10582.01) | 6910.97 (min 2450.23, max 11371.71) |
| bw_mem-fcp-16mb (MB/s) | 3367.36 | 3051.11 |
| bw_mem-fcp-1mb (MB/s) | 8456.01 (min 4091.50, max 12820.51) | 10197.42 (min 6627.39, max 13767.44) |
| bw_mem-fcp-2mb (MB/s) | 7403.95 (min 3485.00, max 11322.89) | 8668.11 (min 3674.44, max 13661.77) |
| bw_mem-fcp-4mb (MB/s) | 7049.72 (min 3434.66, max 10664.77) | 7877.93 (min 3477.05, max 12278.80) |
| bw_mem-fcp-8mb (MB/s) | 6970.83 (min 3373.63, max 10568.03) | 7264.14 (min 3200.85, max 11327.43) |
| bw_mem-frd-16mb (MB/s) | 4166.67 | 3565.86 |
| bw_mem-frd-1mb (MB/s) | 4687.06 (min 4091.50, max 5282.62) | 7284.10 (min 6627.39, max 7940.80) |
| bw_mem-frd-2mb (MB/s) | 4066.28 (min 3485.00, max 4647.56) | 4134.31 (min 3674.44, max 4594.18) |
| bw_mem-frd-4mb (MB/s) | 3827.04 (min 3434.66, max 4219.41) | 3698.35 (min 3477.05, max 3919.65) |
| bw_mem-frd-8mb (MB/s) | 3768.71 (min 3373.63, max 4163.78) | 3553.87 (min 3200.85, max 3906.89) |
| bw_mem-fwr-16mb (MB/s) | 10575.02 | 10832.77 |
| bw_mem-fwr-1mb (MB/s) | 8951.03 (min 5282.62, max 12619.43) | 10840.77 (min 7940.80, max 13740.73) |
| bw_mem-fwr-2mb (MB/s) | 7982.02 (min 4647.56, max 11316.48) | 9143.94 (min 4594.18, max 13693.69) |
| bw_mem-fwr-4mb (MB/s) | 7439.25 (min 4219.41, max 10659.09) | 8310.15 (min 3919.65, max 12700.65) |
| bw_mem-fwr-8mb (MB/s) | 7372.90 (min 4163.78, max 10582.01) | 7639.30 (min 3906.89, max 11371.71) |
| bw_mem-rd-16mb (MB/s) | 4883.26 | 4263.26 |
| bw_mem-rd-1mb (MB/s) | 5881.98 (min 4604.05, max 7159.90) | 15977.73 (min 14633.18, max 17322.27) |
| bw_mem-rd-2mb (MB/s) | 4323.14 (min 3001.88, max 5644.40) | 6988.59 (min 6321.33, max 7655.85) |
| bw_mem-rd-4mb (MB/s) | 3661.09 (min 2335.54, max 4986.64) | 4310.51 (min 3872.84, max 4748.18) |
| bw_mem-rd-8mb (MB/s) | 3567.05 (min 2235.89, max 4898.21) | 4265.04 (min 3775.96, max 4754.12) |
| bw_mem-rdwr-16mb (MB/s) | 2128.51 | 2645.50 |
| bw_mem-rdwr-1mb (MB/s) | 3318.17 (min 2478.75, max 4157.59) | 6653.54 (min 3693.44, max 9613.64) |
| bw_mem-rdwr-2mb (MB/s) | 2312.18 (min 2089.50, max 2534.85) | 3819.73 (min 2743.95, max 4895.50) |
| bw_mem-rdwr-4mb (MB/s) | 2176.81 (min 2092.78, max 2260.84) | 2969.19 (min 2802.59, max 3135.78) |
| bw_mem-rdwr-8mb (MB/s) | 2111.86 (min 2090.96, max 2132.76) | 2765.16 (min 2450.23, max 3080.08) |
| bw_mem-wr-16mb (MB/s) | 2206.29 | 3039.51 |
| bw_mem-wr-1mb (MB/s) | 4380.82 (min 4157.59, max 4604.05) | 13467.96 (min 9613.64, max 17322.27) |
| bw_mem-wr-2mb (MB/s) | 2768.37 (min 2534.85, max 3001.88) | 5608.42 (min 4895.50, max 6321.33) |
| bw_mem-wr-4mb (MB/s) | 2298.19 (min 2260.84, max 2335.54) | 3504.31 (min 3135.78, max 3872.84) |
| bw_mem-wr-8mb (MB/s) | 2184.33 (min 2132.76, max 2235.89) | 3428.02 (min 3080.08, max 3775.96) |
| bw_mmap_rd-mo-1mb (MB/s) | 8630.61 | 12919.43 |
| bw_mmap_rd-o2c-1mb (MB/s) | 979.43 | 1195.70 |
| bw_pipe (MB/s) | 848.26 | 886.10 |
| bw_unix (MB/s) | 1989.80 | 3409.24 |
| lat_connect (us) | 34.62 | 34.50 |
| lat_ctx-2-128k (us) | 3.57 | 3.42 |
| lat_ctx-2-256k (us) | 2.80 | 2.57 |
| lat_ctx-4-128k (us) | 3.76 | 3.17 |
| lat_ctx-4-256k (us) | 2.73 | 2.54 |
| lat_fs-0k (num_files) | 486.00 | 568.00 |
| lat_fs-10k (num_files) | 189.00 | 220.00 |
| lat_fs-1k (num_files) | 296.00 | 308.00 |
| lat_fs-4k (num_files) | 311.00 | 284.00 |
| lat_mem_rd-stride128-sz1000k (ns) | 13.34 | 5.72 |
| lat_mem_rd-stride128-sz125k (ns) | 5.57 | 5.65 |
| lat_mem_rd-stride128-sz250k (ns) | 5.57 | 5.65 |
| lat_mem_rd-stride128-sz31k (ns) | 3.35 | 3.40 |
| lat_mem_rd-stride128-sz50 (ns) | 2.00 | 2.00 |
| lat_mem_rd-stride128-sz500k (ns) | 5.90 | 5.65 |
| lat_mem_rd-stride128-sz62k (ns) | 5.57 | 5.65 |
| lat_mmap-1m (us) | 35.00 | 34.00 |
| lat_ops-double-add (ns) | 1.96 | 1.96 |
| lat_ops-double-div (ns) | 9.01 | 9.00 |
| lat_ops-double-mul (ns) | 2.00 | 2.00 |
| lat_ops-float-add (ns) | 1.96 | 1.96 |
| lat_ops-float-div (ns) | 5.50 | 5.50 |
| lat_ops-float-mul (ns) | 2.00 | 2.00 |
| lat_ops-int-add (ns) | 0.50 | 0.50 |
| lat_ops-int-bit (ns) | 0.33 | 0.33 |
| lat_ops-int-div (ns) | 4.00 | 4.00 |
| lat_ops-int-mod (ns) | 4.69 | 4.67 |
| lat_ops-int-mul (ns) | 1.52 | 1.52 |
| lat_ops-int64-add (ns) | 0.50 | 0.50 |
| lat_ops-int64-bit (ns) | 0.33 | 0.33 |
| lat_ops-int64-div (ns) | 3.00 | 3.00 |
| lat_ops-int64-mod (ns) | 5.67 | 5.67 |
| lat_ops-int64-mul (ns) | 2.52 | 2.52 |
| lat_pagefault (us) | 0.50 | 0.51 |
| lat_pipe (us) | 12.34 | 17.15 |
| lat_proc-exec (us) | 446.25 | 402.92 |
| lat_proc-fork (us) | 370.00 | 368.06 |
| lat_proc-proccall (us) | 0.00 | 0.00 |
| lat_select (us) | 11.10 | 11.15 |
| lat_sem (us) | 1.46 | 1.33 |
| lat_sig-catch (us) | 2.71 | 2.70 |
| lat_sig-install (us) | 0.51 | 0.52 |
| lat_sig-prot (us) | 0.45 | 0.42 |
| lat_syscall-fstat (us) | 1.29 | 1.26 |
| lat_syscall-null (us) | 0.38 | 0.38 |
| lat_syscall-open (us) | 237.26 | 2937.00 |
| lat_syscall-read (us) | 0.48 | 0.49 |
| lat_syscall-stat (us) | 1.68 | 1.63 |
| lat_syscall-write (us) | 0.42 | 0.44 |
| lat_tcp (us) | 0.82 | 0.76 |
| lat_unix (us) | 15.11 | 14.06 |
| latency_for_0.50_mb_block_size (nanosec) | 5.90 | 5.65 |
| latency_for_1.00_mb_block_size (nanosec) | 6.67 (min 0.00, max 13.34) | 2.86 (min 0.00, max 5.72) |
| pipe_bandwidth (MBs) | 848.26 | 886.10 |
| pipe_latency (microsec) | 12.34 | 17.15 |
| procedure_call (microsec) | 0.00 | 0.00 |
| select_on_200_tcp_fds (microsec) | 11.10 | 11.15 |
| semaphore_latency (microsec) | 1.46 | 1.33 |
| signal_handler_latency (microsec) | 0.51 | 0.52 |
| signal_handler_overhead (microsec) | 2.71 | 2.70 |
| tcp_ip_connection_cost_to_localhost (microsec) | 34.62 | 34.50 |
| tcp_latency_using_localhost (microsec) | 0.82 | 0.76 |
Table: LM Bench Metrics
2.2. 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 | am68_sk-fs: perf | am69_sk-fs: perf |
|---|---|---|
| cpu_clock (MHz) | 2000.00 | 2000.00 |
| dhrystone_per_mhz (DMIPS/MHz) | 4.70 | 5.70 |
| dhrystone_per_second (DhrystoneP) | 16666667.00 | 20000000.00 |
Table: Dhrystone Benchmark
2.2. Whetstone¶
Whetstone is a benchmark primarily measuring floating-point arithmetic performance.
Execute the benchmark with the following:
runWhetstone
| Benchmarks | am68_sk-fs: perf | am69_sk-fs: perf |
|---|---|---|
| whetstone (MIPS) | 10000.00 | 10000.00 |
Table: Whetstone Benchmark
2.2. Linpack¶
Linpack measures peak double precision (64 bit) floating point performance in solving a dense linear system.
| Benchmarks | am68_sk-fs: perf | am69_sk-fs: perf |
|---|---|---|
| linpack (Kflops) | 2508830.00 | 2614877.00 |
Table: Linpack Benchmark
2.2. 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 | am68_sk-fs: perf | am69_sk-fs: perf |
|---|---|---|
| add (MB/s) | 6213.90 | 6119.00 |
| copy (MB/s) | 6958.40 | 6635.20 |
| scale (MB/s) | 7075.30 | 6660.80 |
| triad (MB/s) | 6223.60 | 6113.20 |
Table: Stream
2.2. 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 | am68_sk-fs: perf | am69_sk-fs: perf |
|---|---|---|
| cjpeg-rose7-preset (workloads/) | 82.64 | 81.97 |
| core (workloads/) | 0.78 | 0.78 |
| coremark-pro () | 2472.41 | 2498.29 |
| linear_alg-mid-100x100-sp (workloads/) | 80.65 | 81.57 |
| loops-all-mid-10k-sp (workloads/) | 2.48 | 2.47 |
| nnet_test (workloads/) | 3.59 | 3.56 |
| parser-125k (workloads/) | 11.11 | 10.87 |
| radix2-big-64k (workloads/) | 251.51 | 271.37 |
| sha-test (workloads/) | 158.73 | 158.73 |
| zip-test (workloads/) | 47.62 | 50.00 |
Table: CoreMarkPro
2.2. 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 | am68_sk-fs: perf | am69_sk-fs: perf |
|---|---|---|
| 4m-check (workloads/) | 881.52 | 1039.93 |
| 4m-check-reassembly (workloads/) | 151.06 | 207.04 |
| 4m-check-reassembly-tcp (workloads/) | 98.04 | 119.05 |
| 4m-check-reassembly-tcp-cmykw2-rotatew2 (workloads/) | 41.64 | 38.68 |
| 4m-check-reassembly-tcp-x264w2 (workloads/) | 2.68 | 4.79 |
| 4m-cmykw2 (workloads/) | 312.01 | 600.60 |
| 4m-cmykw2-rotatew2 (workloads/) | 59.17 | 46.26 |
| 4m-reassembly (workloads/) | 131.06 | 155.28 |
| 4m-rotatew2 (workloads/) | 70.47 | 49.07 |
| 4m-tcp-mixed (workloads/) | 266.67 | 280.70 |
| 4m-x264w2 (workloads/) | 2.73 | 4.96 |
| idct-4m (workloads/) | 35.00 | 35.15 |
| idct-4mw1 (workloads/) | 35.01 | 35.11 |
| ippktcheck-4m (workloads/) | 881.52 | 1040.37 |
| ippktcheck-4mw1 (workloads/) | 869.26 | 1042.10 |
| ipres-4m (workloads/) | 165.38 | 208.91 |
| ipres-4mw1 (workloads/) | 164.65 | 208.91 |
| md5-4m (workloads/) | 45.98 | 47.48 |
| md5-4mw1 (workloads/) | 45.72 | 47.26 |
| rgbcmyk-4m (workloads/) | 163.13 | 164.20 |
| rgbcmyk-4mw1 (workloads/) | 162.87 | 163.93 |
| rotate-4ms1 (workloads/) | 52.03 | 55.01 |
| rotate-4ms1w1 (workloads/) | 51.02 | 55.01 |
| rotate-4ms64 (workloads/) | 52.69 | 55.43 |
| rotate-4ms64w1 (workloads/) | 52.91 | 55.62 |
| x264-4mq (workloads/) | 1.42 | 1.42 |
| x264-4mqw1 (workloads/) | 1.41 | 1.40 |
Table: Multibench
2.2. Boot-time Measurement¶
2.2. Boot media: MMCSD¶
| Boot Configuration | am68_sk-fs: boot time (sec) | am69_sk-fs: boot time (sec) |
|---|---|---|
| Kernel boot time test when bootloader, kernel and sdk-rootfs are in mmc-sd | 17.15 (min 16.84, max 17.34) | 15.54 (min 15.19, max 15.76) |
| Kernel boot time test when init is /bin/sh and bootloader, kernel and sdk-rootfs are in mmc-sd | 4.38 (min 4.38, max 4.39) | 4.98 (min 4.94, max 4.99) |
Table: Boot time MMC/SD
2.2. Graphics SGX/RGX Driver¶
2.2. Glmark2¶
Run Glmark2 and capture performance reported (Score). All display outputs (HDMI, Displayport and/or LCD) are connected when running these tests
| Benchmark | am68_sk-fs: Score | am69_sk-fs: Score |
|---|---|---|
| Glmark2-DRM | 55.00 | 55.00 |
| Glmark2-Wayland | 1297.00 | 1436.00 |
Table: Glmark2
2.2. 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
2.2. CPSW/CPSW2g/CPSW3g Ethernet Driver¶
- CPSW2g: AM65x, J7200, J721e, J721S2, J784S4
- CPSW3g: AM64x
TCP Bidirectional Throughput
| Command Used | am69_sk-fs: THROUGHPUT (Mbits/sec) | am69_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 | 181.25 | 3.03 |
Table: CPSW TCP Bidirectional Throughput
TCP Bidirectional Throughput Interrupt Pacing
| Command Used | am68_sk-fs: THROUGHPUT (Mbits/sec) | am68_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 | 179.92 | 9.38 |
Table: CPSW TCP Bidirectional Throughput Interrupt Pacing
2.2. CRYPTO Driver¶
2.2. OpenSSL Performance¶
| Algorithm | Buffer Size (in bytes) | am68_sk-fs: throughput (KBytes/Sec) | am69_sk-fs: throughput (KBytes/Sec) |
|---|---|---|---|
| aes-128-cbc | 1024 | 1419020.29 | 1401805.14 |
| aes-128-cbc | 16 | 312105.08 | 298655.59 |
| aes-128-cbc | 16384 | 1484745.39 | 1501779.29 |
| aes-128-cbc | 256 | 1271794.43 | 1206112.26 |
| aes-128-cbc | 64 | 863253.42 | 750428.14 |
| aes-128-cbc | 8192 | 1486826.15 | 1497565.87 |
| aes-192-cbc | 1024 | 1262488.58 | 1257734.83 |
| aes-192-cbc | 16 | 356885.48 | 358992.33 |
| aes-192-cbc | 16384 | 1324176.73 | 1311391.74 |
| aes-192-cbc | 256 | 1088377.09 | 1075945.22 |
| aes-192-cbc | 64 | 799344.66 | 799043.11 |
| aes-192-cbc | 8192 | 1322822.31 | 1322805.93 |
| aes-256-cbc | 1024 | 1088494.93 | 1076505.60 |
| aes-256-cbc | 16 | 347136.90 | 277460.52 |
| aes-256-cbc | 16384 | 1128819.37 | 1117454.34 |
| aes-256-cbc | 256 | 999211.69 | 956262.31 |
| aes-256-cbc | 64 | 726533.27 | 642237.78 |
| aes-256-cbc | 8192 | 1129141.59 | 1130558.81 |
| des3 | 1024 | 16290.47 | 16296.28 |
| des3 | 16 | 15234.17 | 15533.80 |
| des3 | 16384 | 16302.08 | 16307.54 |
| des3 | 256 | 16230.91 | 16246.44 |
| des3 | 64 | 16070.38 | 16110.72 |
| des3 | 8192 | 16302.08 | 16307.54 |
| md5 | 1024 | 268004.01 | 272543.06 |
| md5 | 16 | 23708.61 | 26673.35 |
| md5 | 16384 | 321809.07 | 322120.36 |
| md5 | 256 | 176228.44 | 182089.81 |
| md5 | 64 | 74984.49 | 79500.48 |
| md5 | 8192 | 317723.99 | 318521.34 |
| sha1 | 1024 | 739966.63 | 702369.11 |
| sha1 | 16 | 28852.33 | 28167.43 |
| sha1 | 16384 | 1091928.06 | 1087094.78 |
| sha1 | 256 | 328607.49 | 328636.50 |
| sha1 | 64 | 107892.71 | 107398.46 |
| sha1 | 8192 | 1058485.59 | 1051451.39 |
| sha224 | 1024 | 670352.04 | 666634.24 |
| sha224 | 16 | 32146.12 | 30489.96 |
| sha224 | 16384 | 967469.74 | 967546.20 |
| sha224 | 256 | 342687.23 | 339400.28 |
| sha224 | 64 | 115086.29 | 113553.15 |
| sha224 | 8192 | 936132.61 | 943158.61 |
| sha256 | 1024 | 650168.66 | 651953.83 |
| sha256 | 16 | 27388.27 | 29250.07 |
| sha256 | 16384 | 962494.46 | 960075.09 |
| sha256 | 256 | 322742.19 | 323623.77 |
| sha256 | 64 | 104748.25 | 106680.26 |
| sha256 | 8192 | 938912.43 | 939095.38 |
| sha384 | 1024 | 217740.29 | 216774.66 |
| sha384 | 16 | 15821.97 | 15635.32 |
| sha384 | 16384 | 271592.11 | 271766.87 |
| sha384 | 256 | 133338.11 | 131896.75 |
| sha384 | 64 | 63657.26 | 63300.20 |
| sha384 | 8192 | 268088.66 | 267785.56 |
| sha512 | 1024 | 220648.79 | 219712.85 |
| sha512 | 16 | 17191.00 | 16875.69 |
| sha512 | 16384 | 271187.97 | 272307.54 |
| sha512 | 256 | 137726.63 | 136462.25 |
| sha512 | 64 | 68889.02 | 67711.59 |
| sha512 | 8192 | 267815.59 | 268266.15 |
| Algorithm | am68_sk-fs: CPU Load | am69_sk-fs: CPU Load |
|---|---|---|
| aes-128-cbc | 99.00 | 99.00 |
| aes-192-cbc | 99.00 | 99.00 |
| aes-256-cbc | 99.00 | 99.00 |
| des-cbc | 60.00 | 100.00 |
| des3 | 99.00 | 99.00 |
| md5 | 99.00 | 99.00 |
| sha1 | 99.00 | 99.00 |
| sha224 | 99.00 | 99.00 |
| sha256 | 99.00 | 99.00 |
| sha384 | 99.00 | 99.00 |
| sha512 | 99.00 | 99.00 |
- Listed for each algorithm are the code snippets used to run each
- benchmark test.
- ::
- time -v openssl speed -elapsed -evp aes-128-cbc
2.2. DCAN Driver¶
Performance and Benchmarks not available in this release.