-
Notifications
You must be signed in to change notification settings - Fork 96
Allwinner A33
Icenowy Zheng edited this page Jul 25, 2017
·
3 revisions
A33 Q8 tablet 2c789849709d837b4bd114c11ed2d9bdc65afbc6 CFLAGS='-O3 -mcpu=cortex-a7' Note: When disassembling the tablet, I found that only one 8-bit DDR3 chip is soldered.
tinymembench v0.4.9 (simple benchmark for memory throughput and latency)
==========================================================================
== Memory bandwidth tests ==
== ==
== Note 1: 1MB = 1000000 bytes ==
== Note 2: Results for 'copy' tests show how many bytes can be ==
== copied per second (adding together read and writen ==
== bytes would have provided twice higher numbers) ==
== Note 3: 2-pass copy means that we are using a small temporary buffer ==
== to first fetch data into it, and only then write it to the ==
== destination (source -> L1 cache, L1 cache -> destination) ==
== Note 4: If sample standard deviation exceeds 0.1%, it is shown in ==
== brackets ==
==========================================================================
C copy backwards : 201.7 MB/s (0.9%)
C copy backwards (32 byte blocks) : 601.4 MB/s (5.8%)
C copy backwards (64 byte blocks) : 627.4 MB/s (0.1%)
C copy : 626.7 MB/s
C copy prefetched (32 bytes step) : 629.6 MB/s
C copy prefetched (64 bytes step) : 655.2 MB/s
C 2-pass copy : 592.2 MB/s
C 2-pass copy prefetched (32 bytes step) : 592.0 MB/s
C 2-pass copy prefetched (64 bytes step) : 605.1 MB/s
C fill : 1762.0 MB/s
C fill (shuffle within 16 byte blocks) : 1762.3 MB/s
C fill (shuffle within 32 byte blocks) : 272.7 MB/s
C fill (shuffle within 64 byte blocks) : 276.6 MB/s
---
standard memcpy : 646.3 MB/s
standard memset : 1762.4 MB/s
---
NEON read : 1023.3 MB/s
NEON read prefetched (32 bytes step) : 1144.7 MB/s
NEON read prefetched (64 bytes step) : 1164.9 MB/s
NEON read 2 data streams : 304.5 MB/s
NEON read 2 data streams prefetched (32 bytes step) : 572.2 MB/s
NEON read 2 data streams prefetched (64 bytes step) : 597.8 MB/s
NEON copy : 628.0 MB/s
NEON copy prefetched (32 bytes step) : 630.3 MB/s
NEON copy prefetched (64 bytes step) : 679.2 MB/s
NEON unrolled copy : 620.7 MB/s
NEON unrolled copy prefetched (32 bytes step) : 641.0 MB/s (0.1%)
NEON unrolled copy prefetched (64 bytes step) : 668.6 MB/s
NEON copy backwards : 601.0 MB/s
NEON copy backwards prefetched (32 bytes step) : 622.5 MB/s
NEON copy backwards prefetched (64 bytes step) : 670.4 MB/s
NEON 2-pass copy : 597.5 MB/s
NEON 2-pass copy prefetched (32 bytes step) : 643.7 MB/s
NEON 2-pass copy prefetched (64 bytes step) : 658.8 MB/s
NEON unrolled 2-pass copy : 584.7 MB/s
NEON unrolled 2-pass copy prefetched (32 bytes step) : 578.1 MB/s
NEON unrolled 2-pass copy prefetched (64 bytes step) : 598.7 MB/s
NEON fill : 1762.7 MB/s
NEON fill backwards : 1761.9 MB/s
VFP copy : 622.5 MB/s (0.8%)
VFP 2-pass copy : 586.6 MB/s
ARM fill (STRD) : 1762.5 MB/s
ARM fill (STM with 8 registers) : 1762.4 MB/s
ARM fill (STM with 4 registers) : 1762.2 MB/s
ARM copy prefetched (incr pld) : 670.5 MB/s
ARM copy prefetched (wrap pld) : 619.5 MB/s
ARM 2-pass copy prefetched (incr pld) : 634.3 MB/s
ARM 2-pass copy prefetched (wrap pld) : 605.6 MB/s
==========================================================================
== Framebuffer read tests. ==
== ==
== Many ARM devices use a part of the system memory as the framebuffer, ==
== typically mapped as uncached but with write-combining enabled. ==
== Writes to such framebuffers are quite fast, but reads are much ==
== slower and very sensitive to the alignment and the selection of ==
== CPU instructions which are used for accessing memory. ==
== ==
== Many x86 systems allocate the framebuffer in the GPU memory, ==
== accessible for the CPU via a relatively slow PCI-E bus. Moreover, ==
== PCI-E is asymmetric and handles reads a lot worse than writes. ==
== ==
== If uncached framebuffer reads are reasonably fast (at least 100 MB/s ==
== or preferably >300 MB/s), then using the shadow framebuffer layer ==
== is not necessary in Xorg DDX drivers, resulting in a nice overall ==
== performance improvement. For example, the xf86-video-fbturbo DDX ==
== uses this trick. ==
==========================================================================
NEON read (from framebuffer) : 44.2 MB/s
NEON copy (from framebuffer) : 42.3 MB/s (0.1%)
NEON 2-pass copy (from framebuffer) : 42.8 MB/s
NEON unrolled copy (from framebuffer) : 42.9 MB/s
NEON 2-pass unrolled copy (from framebuffer) : 42.4 MB/s
VFP copy (from framebuffer) : 247.7 MB/s
VFP 2-pass copy (from framebuffer) : 233.5 MB/s
ARM copy (from framebuffer) : 151.0 MB/s
ARM 2-pass copy (from framebuffer) : 144.1 MB/s
==========================================================================
== Memory latency test ==
== ==
== Average time is measured for random memory accesses in the buffers ==
== of different sizes. The larger is the buffer, the more significant ==
== are relative contributions of TLB, L1/L2 cache misses and SDRAM ==
== accesses. For extremely large buffer sizes we are expecting to see ==
== page table walk with several requests to SDRAM for almost every ==
== memory access (though 64MiB is not nearly large enough to experience ==
== this effect to its fullest). ==
== ==
== Note 1: All the numbers are representing extra time, which needs to ==
== be added to L1 cache latency. The cycle timings for L1 cache ==
== latency can be usually found in the processor documentation. ==
== Note 2: Dual random read means that we are simultaneously performing ==
== two independent memory accesses at a time. In the case if ==
== the memory subsystem can't handle multiple outstanding ==
== requests, dual random read has the same timings as two ==
== single reads performed one after another. ==
==========================================================================
block size : single random read / dual random read
1024 : 0.0 ns / 0.0 ns
2048 : 0.0 ns / 0.0 ns
4096 : 0.0 ns / 0.0 ns
8192 : 0.0 ns / 0.0 ns
16384 : 0.0 ns / 0.0 ns
32768 : 0.0 ns / 0.0 ns
65536 : 6.2 ns / 10.8 ns
131072 : 9.6 ns / 15.1 ns
262144 : 11.4 ns / 16.8 ns
524288 : 14.3 ns / 20.3 ns
1048576 : 123.0 ns / 193.8 ns
2097152 : 186.6 ns / 258.9 ns
4194304 : 219.4 ns / 283.5 ns
8388608 : 239.2 ns / 297.6 ns
16777216 : 255.7 ns / 316.0 ns
33554432 : 274.1 ns / 345.3 ns
67108864 : 302.8 ns / 400.2 ns
Kernel 4.9.140-tegra #1 SMP PREEMPT Wed Mar 13 00:32:22 PDT 2019 aarch64 GNU/Linux Under xorg, no compositor active, no browser or other cpu hogs.
tinymembench v0.4.9 (simple benchmark for memory thr
==========================================================================
== Memory bandwidth tests ==
== ==
== Note 1: 1MB = 1000000 bytes ==
== Note 2: Results for 'copy' tests show how many bytes can be ==
== copied per second (adding together read and writen ==
== bytes would have provided twice higher numbers) ==
== Note 3: 2-pass copy means that we are using a small temporary buffer ==
== to first fetch data into it, and only then write it to the ==
== destination (source -> L1 cache, L1 cache -> destination) ==
== Note 4: If sample standard deviation exceeds 0.1%, it is shown in ==
== brackets ==
==========================================================================
C copy backwards : 2949.7 MB/s (3.8%)
C copy backwards (32 byte blocks) : 3011.8 MB/s
C copy backwards (64 byte blocks) : 3029.2 MB/s
C copy : 3642.2 MB/s (4.1%)
C copy prefetched (32 bytes step) : 3824.4 MB/s (0.3%)
C copy prefetched (64 bytes step) : 3825.3 MB/s (0.4%)
C 2-pass copy : 2726.2 MB/s
C 2-pass copy prefetched (32 bytes step) : 2902.6 MB/s (2.5%)
C 2-pass copy prefetched (64 bytes step) : 2928.3 MB/s (0.3%)
C fill : 8541.0 MB/s (0.2%)
C fill (shuffle within 16 byte blocks) : 8518.5 MB/s (2.1%)
C fill (shuffle within 32 byte blocks) : 8537.1 MB/s (0.1%)
C fill (shuffle within 64 byte blocks) : 8528.7 MB/s (0.2%)
---
standard memcpy : 3558.8 MB/s
standard memset : 8520.2 MB/s
---
NEON LDP/STP copy : 3633.9 MB/s (4.2%)
NEON LDP/STP copy pldl2strm (32 bytes step) : 1451.0 MB/s (0.3%)
NEON LDP/STP copy pldl2strm (64 bytes step) : 1450.9 MB/s (0.5%)
NEON LDP/STP copy pldl1keep (32 bytes step) : 3882.5 MB/s (3.9%)
NEON LDP/STP copy pldl1keep (64 bytes step) : 3884.0 MB/s (0.4%)
NEON LD1/ST1 copy : 3630.8 MB/s (0.3%)
NEON STP fill : 8537.8 MB/s
NEON STNP fill : 8544.9 MB/s (1.7%)
ARM LDP/STP copy : 3635.8 MB/s (0.3%)
ARM STP fill : 8544.8 MB/s (0.1%)
ARM STNP fill : 8549.2 MB/s (1.0%)
==========================================================================
== Framebuffer read tests. ==
== ==
== Many ARM devices use a part of the system memory as the framebuffer, ==
== typically mapped as uncached but with write-combining enabled. ==
== Writes to such framebuffers are quite fast, but reads are much ==
== slower and very sensitive to the alignment and the selection of ==
== CPU instructions which are used for accessing memory. ==
== ==
== Many x86 systems allocate the framebuffer in the GPU memory, ==
== accessible for the CPU via a relatively slow PCI-E bus. Moreover, ==
== PCI-E is asymmetric and handles reads a lot worse than writes. ==
== ==
== If uncached framebuffer reads are reasonably fast (at least 100 MB/s ==
== or preferably >300 MB/s), then using the shadow framebuffer layer ==
== is not necessary in Xorg DDX drivers, resulting in a nice overall ==
== performance improvement. For example, the xf86-video-fbturbo DDX ==
== uses this trick. ==
==========================================================================
NEON LDP/STP copy (from framebuffer) : 766.0 MB/s
NEON LDP/STP 2-pass copy (from framebuffer) : 688.8 MB/s
NEON LD1/ST1 copy (from framebuffer) : 770.6 MB/s (0.1%)
NEON LD1/ST1 2-pass copy (from framebuffer) : 681.3 MB/s (0.3%)
ARM LDP/STP copy (from framebuffer) : 766.1 MB/s
ARM LDP/STP 2-pass copy (from framebuffer) : 689.1 MB/s
==========================================================================
== Memory latency test ==
== ==
== Average time is measured for random memory accesses in the buffers ==
== of different sizes. The larger is the buffer, the more significant ==
== are relative contributions of TLB, L1/L2 cache misses and SDRAM ==
== accesses. For extremely large buffer sizes we are expecting to see ==
== page table walk with several requests to SDRAM for almost every ==
== memory access (though 64MiB is not nearly large enough to experience ==
== this effect to its fullest). ==
== ==
== Note 1: All the numbers are representing extra time, which needs to ==
== be added to L1 cache latency. The cycle timings for L1 cache ==
== latency can be usually found in the processor documentation. ==
== Note 2: Dual random read means that we are simultaneously performing ==
== two independent memory accesses at a time. In the case if ==
== the memory subsystem can't handle multiple outstanding ==
== requests, dual random read has the same timings as two ==
== single reads performed one after another. ==
==========================================================================
block size : single random read / dual random read, [MADV_NOHUGEPAGE]
1024 : 0.0 ns / 0.1 ns
2048 : 0.0 ns / 0.1 ns
4096 : 0.0 ns / 0.1 ns
8192 : 0.0 ns / 0.1 ns
16384 : 0.1 ns / 0.1 ns
32768 : 1.7 ns / 2.9 ns
65536 : 6.4 ns / 9.5 ns
131072 : 9.6 ns / 12.3 ns
262144 : 13.7 ns / 17.0 ns
524288 : 15.8 ns / 19.7 ns
1048576 : 17.3 ns / 22.1 ns
2097152 : 42.1 ns / 64.2 ns
4194304 : 98.5 ns / 138.1 ns
8388608 : 143.9 ns / 186.3 ns
16777216 : 167.2 ns / 211.2 ns
33554432 : 180.1 ns / 227.1 ns
67108864 : 200.0 ns / 260.2 ns
block size : single random read / dual random read, [MADV_HUGEPAGE]
1024 : 0.0 ns / 0.0 ns
2048 : 0.0 ns / 0.0 ns
4096 : 0.0 ns / 0.0 ns
8192 : 0.0 ns / 0.0 ns
16384 : 0.0 ns / 0.0 ns
32768 : 0.0 ns / 0.0 ns
65536 : 6.4 ns / 9.4 ns
131072 : 9.5 ns / 12.2 ns
262144 : 11.2 ns / 13.1 ns
524288 : 12.1 ns / 13.5 ns
1048576 : 12.8 ns / 13.6 ns
2097152 : 27.0 ns / 33.0 ns
4194304 : 90.6 ns / 127.8 ns
8388608 : 123.9 ns / 153.8 ns
16777216 : 139.5 ns / 161.2 ns
33554432 : 147.2 ns / 163.6 ns
67108864 : 154.0 ns / 167.6 ns