Part Number: CC2340R5-Q1
Hello TI Team,
I am using SDK version 9.14.02.16.
Using the code below, I measured the processing time required to write to and erase one sector.
NVS_Params lp_nvs_params;
/* NVS has been erased. */
/* Init NVS */
NVS_init();
NVS_Params_init(&lp_nvs_params);
gh_nvs_handle = NVS_open(CONFIG_NVS_0, &lp_nvs_params);
/* gu32_d_test … 2048byte All:0x00 or 0xFF */
/* Measure NVS_write [1 Sector] time */
GPIO_write(CONFIG_GPIO_0, GPIO_CFG_OUT_HIGH);
(void)NVS_write(gh_nvs_handle, (size_t)0, gu32_d_test, (size_t)2048, 0);
GPIO_write(CONFIG_GPIO_0, GPIO_CFG_OUT_LOW);
/* Measure NVS_erase [1 Sector] time */
GPIO_write(CONFIG_GPIO_0, GPIO_CFG_OUT_HIGH);
(void)NVS_erase(gh_nvs_handle, (size_t)0, (size_t)2048);
GPIO_write(CONFIG_GPIO_0, GPIO_CFG_OUT_LOW);
Table: Measurement Results
| API | Condition | Time | Unit | Notes |
| NVS_write[1Ssector] | Write "0x00"to all locations | 16.5 | msec | Write flags = None(0) |
| ↑ | Write "0xFF" to all locations | 9.36 | msec | Write flags = None(0) |
| NVS_erase[1Ssector] | Erase all locations containing "0x00" | 6.3 | msec | |
| ↑ | Erase all locations containing "0xFF" | 0.77 | msec |
Question 1:
There is a difference between my measurements and the values shown in the datasheet. Under what conditions—such as the number of write/erase cycles—were the typical values specified in the datasheet measured?
There is a difference between my measurements and the values shown in the datasheet. Under what conditions—such as the number of write/erase cycles—were the typical values specified in the datasheet measured?
Question 2:
Are the processing times measured through the API different from the values specified in the datasheet? If so, approximately how much overhead does the API add?
Are the processing times measured through the API different from the values specified in the datasheet? If so, approximately how much overhead does the API add?
Question 3:
The processing time varies significantly depending on the value being written and the data stored before the erase operation. What causes this variation?
The processing time varies significantly depending on the value being written and the data stored before the erase operation. What causes this variation?
Additional information:
The measurements were taken on a device with fewer than 1,000 total write/erase cycles.
The measurements were taken on a device with fewer than 1,000 total write/erase cycles.
Best regards,
