Benchmarks
Y como no, también hemos utilizado el famoso SBC de Thomas Kaiser que se descarga a instala así …
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 |
root@orangepi5ultra:/home/orangepi# wget https://raw.githubusercontent.com/ThomasKaiser/sbc-bench/master/sbc-bench.sh --2025-05-28 17:02:19-- https://raw.githubusercontent.com/ThomasKaiser/sbc-bench/master/sbc-bench.sh Resolviendo raw.githubusercontent.com (raw.githubusercontent.com)... 185.199.109.133, 185.199.111.133, 185.199.108.133, ... Conectando con raw.githubusercontent.com (raw.githubusercontent.com)[185.199.109.133]:443... conectado. Petición HTTP enviada, esperando respuesta... 200 OK Longitud: 451249 (441K) [text/plain] Guardando como: ‘sbc-bench.sh’ sbc-bench.sh 100%[=================================================================>] 440,67K --.-KB/s en 0,02s 2025-05-28 17:02:19 (18,8 MB/s) - ‘sbc-bench.sh’ guardado [451249/451249] |
y cuyos resultados os dejamos aquí.
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 |
root@orangepi5ultra:/home/orangepi# sudo /bin/bash ./sbc-bench.sh -r Starting to examine hardware/software for review purposes... Average load and/or CPU utilization too high (too much background activity). Waiting... Too busy for benchmarking: 17:27:17 up 1:11, 6 users, load average: 1.42, 3.19, 2.74, cpu: 0% Too busy for benchmarking: 17:27:22 up 1:11, 6 users, load average: 1.39, 3.15, 2.73, cpu: 0% Too busy for benchmarking: 17:27:27 up 1:11, 6 users, load average: 1.36, 3.12, 2.72, cpu: 0% Too busy for benchmarking: 17:27:32 up 1:12, 6 users, load average: 1.33, 3.08, 2.71, cpu: 0% Too busy for benchmarking: 17:27:37 up 1:12, 6 users, load average: 1.30, 3.05, 2.71, cpu: 0% Too busy for benchmarking: 17:27:42 up 1:12, 6 users, load average: 1.28, 3.01, 2.70, cpu: 0% sbc-bench v0.9.71 Installing needed tools: apt-get -f -qq -y install dmidecode lshw links stress-ng. Updating (can't build cpuminer) Done. Checking cpufreq OPP. Done. Executing tinymembench. Done. Executing RAM latency tester. Done. Executing OpenSSL benchmark. Done. Executing 7-zip benchmark. Done. Throttling test: heating up the device, 5 more minutes to wait. Done. Checking cpufreq OPP again. Done (15 minutes elapsed). Results validation: * Advertised vs. measured max CPU clockspeed: -3.7% before, -4.3% after -> https://tinyurl.com/32w9rr94 * No swapping * Background activity (%system) OK * No throttling Full results uploaded to https://0x0.st/8xFU.bin # RK3588 OPi 5 Ultra Tested with sbc-bench v0.9.71 on Wed, 28 May 2025 17:43:07 +0200. Full info: [https://0x0.st/8xFU.bin](http://0x0.st/8xFU.bin) ### General information: Information courtesy of cpufetch: SoC: Rockchip RK3588 Technology: 8nm CPU 1: Microarchitecture: Cortex-A55 Max Frequency: 1.800 GHz Cores: 4 cores Features: NEON,SHA1,SHA2,AES,CRC32 CPU 2: Microarchitecture: Cortex-A76 Max Frequency: 2.352 GHz Cores: 4 cores Features: NEON,SHA1,SHA2,AES,CRC32 The CPU features 3 clusters consisting of 2 different core types: Rockchip RK3588 (35881000 / 35 88 12 fe 21 4e 5a 58 59 38 00 00 00 00), Kernel: aarch64, Userland: arm64 CPU sysfs topology (clusters, cpufreq members, clockspeeds) cpufreq min max CPU cluster policy speed speed core type 0 0 0 408 1800 Cortex-A55 / r2p0 1 0 0 408 1800 Cortex-A55 / r2p0 2 0 0 408 1800 Cortex-A55 / r2p0 3 0 0 408 1800 Cortex-A55 / r2p0 4 0 4 408 2352 Cortex-A76 / r4p0 5 0 4 408 2352 Cortex-A76 / r4p0 6 0 6 408 2352 Cortex-A76 / r4p0 7 0 6 408 2352 Cortex-A76 / r4p0 7934 KB available RAM ### Governors/policies (performance vs. idle consumption): Original governor settings: cpufreq-policy0: ondemand / 1008 MHz (interactive conservative ondemand userspace powersave performance schedutil / 408 600 816 1008 1200 1416 1608 1800) cpufreq-policy4: ondemand / 1200 MHz (interactive conservative ondemand userspace powersave performance schedutil / 408 600 816 1008 1200 1416 1608 1800 2016 2208 2352) cpufreq-policy6: ondemand / 600 MHz (interactive conservative ondemand userspace powersave performance schedutil / 408 600 816 1008 1200 1416 1608 1800 2016 2208 2352) dmc: dmc_ondemand / 2400 MHz (rknpu_ondemand dmc_ondemand userspace powersave performance simple_ondemand / 534 1320 1968 2400) fb000000.gpu: simple_ondemand / 300 MHz (rknpu_ondemand dmc_ondemand userspace powersave performance simple_ondemand / 300 400 500 600 700 800 900 1000) fdab0000.npu: rknpu_ondemand / 300 MHz (rknpu_ondemand dmc_ondemand userspace powersave performance simple_ondemand / 300 400 500 600 700 800 900 1000) Tuned governor settings: cpufreq-policy0: performance / 1800 MHz cpufreq-policy4: performance / 2352 MHz cpufreq-policy6: performance / 2352 MHz dmc: performance / 2400 MHz fb000000.gpu: performance / 1000 MHz fdab0000.npu: performance / 1000 MHz Status of performance related policies found below /sys: /sys/devices/platform/fb000000.gpu/power_policy: [coarse_demand] always_on /sys/module/pcie_aspm/parameters/policy: default [performance] powersave powersupersave ### Clockspeeds (idle vs. heated up): Before at 56.4°C: cpu0-cpu3 (Cortex-A55): OPP: 1800, Measured: 1781 (-1.1%) cpu4-cpu5 (Cortex-A76): OPP: 2352, Measured: 2269 (-3.5%) cpu6-cpu7 (Cortex-A76): OPP: 2352, Measured: 2265 (-3.7%) After at 74.8°C: cpu0-cpu3 (Cortex-A55): OPP: 1800, Measured: 1770 (-1.7%) cpu4-cpu5 (Cortex-A76): OPP: 2352, Measured: 2254 (-4.2%) cpu6-cpu7 (Cortex-A76): OPP: 2352, Measured: 2251 (-4.3%) ### Performance baseline * cpu0 (Cortex-A55): memcpy: 6539.3 MB/s, memchr: 2738.9 MB/s, memset: 21632.3 MB/s * cpu4 (Cortex-A76): memcpy: 12178.7 MB/s, memchr: 15863.9 MB/s, memset: 27540.4 MB/s * cpu6 (Cortex-A76): memcpy: 12151.2 MB/s, memchr: 15837.2 MB/s, memset: 27432.3 MB/s * cpu0 (Cortex-A55) 16M latency: 140.1 140.7 137.8 140.3 137.4 141.4 218.5 383.6 * cpu4 (Cortex-A76) 16M latency: 141.3 130.3 139.6 130.3 139.4 133.7 128.6 127.5 * cpu6 (Cortex-A76) 16M latency: 139.7 129.6 140.4 129.9 139.1 129.0 126.1 127.2 * cpu0 (Cortex-A55) 128M latency: 162.0 163.2 161.7 163.0 161.4 163.1 235.1 407.3 * cpu4 (Cortex-A76) 128M latency: 153.8 152.3 153.3 152.1 153.5 152.2 152.9 159.4 * cpu6 (Cortex-A76) 128M latency: 155.6 155.0 155.4 155.0 155.6 154.6 154.7 162.4 * 7-zip MIPS (3 consecutive runs): 15590, 15688, 15678 (15650 avg), single-threaded: 2955 * `aes-256-cbc 145614.48k 382933.21k 645486.08k 778974.89k 829057.71k 833049.94k (Cortex-A55)` * `aes-256-cbc 574930.81k 996768.64k 1204175.79k 1265561.94k 1289377.11k 1291709.10k (Cortex-A76)` * `aes-256-cbc 574853.17k 993951.47k 1201168.04k 1263089.32k 1286845.78k 1289136.81k (Cortex-A76)` ### PCIe and storage devices: * Realtek RTL8125 2.5GbE: Speed 5GT/s (ok), Width x1 (ok), driver in use: r8169, ASPM Disabled * 238.5GB "BR 256GB" SSD as /dev/nvme0: Speed 8GT/s (ok), Width x4 (ok), 0% worn out, drive temp: 40°C, ASPM Disabled * 58.3GB "ISOCOM" HS400 Enhanced strobe eMMC 5.1 card as /dev/mmcblk0: date 11/2024, manfid/oemid: 0x000004/0x0101, hw/fw rev: 0x0/0x1000000000000000 * 14.5GB "SanDisk SD16G" UHS SDR50 SD card as /dev/mmcblk1: date 05/2014, manfid/oemid: 0x000003/0x5344, hw/fw rev: 0x3/0x0 * Xmc XM25QU128C 16MB SPI NOR flash, drivers in use: spi-nor/rockchip-sfc ### Swap configuration: * /dev/zram0: 3.9G (0K used, lzo-rle, 8 streams, 4K data, 74B compressed, 12K total) ### Software versions: * Ubuntu 22.04.5 LTS (jammy) * Build scripts: https://github.com/orangepi-xunlong/orangepi-build, 1.0.0, Orange Pi 5 Ultra, rockchip-rk3588, rockchip-rk3588 * Compiler: /usr/bin/gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0 / aarch64-linux-gnu * OpenSSL 3.0.2, built on 15 Mar 2022 (Library: OpenSSL 3.0.2 15 Mar 2022) ### Kernel info: * `/proc/cmdline: root=UUID=69cb8be7-f4f7-4515-a53b-ca446b668190 rootwait rootfstype=ext4 splash plymouth.ignore-serial-consoles console=ttyS2,1500000 console=tty1 consoleblank=0 loglevel=1 ubootpart=6b53f235-db5e-c54f-b2d2-9edf86801bd8 usb-storage.quirks= cma=128M cgroup_enable=cpuset cgroup_memory=1 cgroup_enable=memory swapaccount=1` * Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled via prctl * Vulnerability Spectre v1: Mitigation; __user pointer sanitization * Vulnerability Spectre v2: Vulnerable: Unprivileged eBPF enabled * Kernel 6.1.43-rockchip-rk3588 / CONFIG_HZ=300 All known settings adjusted for performance. Device now ready for benchmarking. Once finished stop with [ctrl]-[c] to get info about throttling, frequency cap and too high background activity all potentially invalidating benchmark scores. All changes with storage and PCIe devices as well as suspicious dmesg contents will be reported too. Time cpu0/cpu4/cpu6 load %cpu %sys %usr %nice %io %irq Temp 17:43:09: 1800/2352/2352MHz 6.95 11% 0% 9% 0% 1% 0% 66.5°C 17:44:09: 1800/2352/2352MHz 3.18 0% 0% 0% 0% 0% 0% 60.1°C 17:45:09: 1800/2352/2352MHz 1.80 0% 0% 0% 0% 0% 0% 58.2°C 17:46:09: 1800/2352/2352MHz 1.36 0% 0% 0% 0% 0% 0% 58.2°C 17:47:09: 1800/2352/2352MHz 1.13 0% 0% 0% 0% 0% 0% 58.2°C 17:48:09: 1800/2352/2352MHz 1.05 0% 0% 0% 0% 0% 0% 57.3°C 17:49:09: 1800/2352/2352MHz 1.02 0% 0% 0% 0% 0% 0% 57.3°C 17:50:09: 1800/2352/2352MHz 1.11 0% 0% 0% 0% 0% 0% 57.3°C 17:51:09: 1800/2352/2352MHz 1.04 0% 0% 0% 0% 0% 0% 57.3°C 17:52:09: 1800/2352/2352MHz 1.01 0% 0% 0% 0% 0% 0% 56.4°C 17:53:09: 1800/2352/2352MHz 1.07 0% 0% 0% 0% 0% 0% 57.3°C ^C Cleaning up. Done. Checking cpufreq OPP again. Done. Clockspeeds now at 58.2°C: cpu0-cpu3 (Cortex-A55): OPP: 1800, Measured: 1781 (-1.1%) cpu4-cpu5 (Cortex-A76): OPP: 2352, Measured: 2268 (-3.6%) cpu6-cpu7 (Cortex-A76): OPP: 2352, Measured: 2264 (-3.7%) ATTENTION: some noise in kernel ring buffer since start of monitoring: [ 5279.478653] [dhd] [wlan0] wl_run_escan : LEGACY_SCAN sync ID: 51, bssidx: 0 [ 5400.476177] [dhd] [wlan0] wl_run_escan : LEGACY_SCAN sync ID: 52, bssidx: 0 [ 5521.479663] [dhd] [wlan0] wl_run_escan : LEGACY_SCAN sync ID: 53, bssidx: 0 [ 5642.481037] [dhd] [wlan0] wl_run_escan : LEGACY_SCAN sync ID: 54, bssidx: 0 [ 5763.488804] [dhd] [wlan0] wl_run_escan : LEGACY_SCAN sync ID: 55, bssidx: 0 [ 5884.495641] [dhd] [wlan0] wl_run_escan : LEGACY_SCAN sync ID: 56, bssidx: 0 Results validation: * Advertised vs. measured max CPU clockspeed: -3.7% before, -4.3% after -> https://tinyurl.com/32w9rr94 * No swapping * Background activity (%system) OK * No throttling * ondemand cpufreq governor used by distro but io_is_busy not set to 1 on all cores -> http://tinyurl.com/44pbmw79 |
Nos quedamos con que la velocidad máxima en ningún momento alcanzó los 75ºC (y el ruido del ventilador al máximo sin molestar lo más mínimo) y sin necesidad de thermal throttling para mantener funcionando sin problemas el procesador.
4739
Resulta cuando menos curioso que haya sido publicar nuestra revisión para que ahora el RPi5 Ultra … se niegue a pararse.
Lo peor del caso es que parece no somos los únicos a los que no funciona el HALT en Debían 12 desde la terminal.
En cuanto lo tengamos resuelto os lo contamos en la revisión.
RESUELTO!
En lugar de utilizar halt … con shutdown funciona perfectamente. Una cosa más que hemos aprendido
Una cosa «negativa» que se nos ha olvidado incluir en la revisión es la FALTA DE PATITAS en la caja metálica para evitar «roces» contra el mueble en el que lo colocas y que no se te enfade la parienta
Si todo va de acuerdo de lo previsto en unos días empezaremos a revisar el nuevo Orange Pi 6 Plus. Además de nuestras típicas pruebas intentaremos probar el desarrollo de aplicaciones de IA aprovechando la potencia del nuevo SBC.