A64 does support UHS-I, but, there is a big problem with SBCs, when even SoC vendors make their evaluating boards with new SoCs crippled down to the legacy only speeds (~20 MB/s). this is because somehow the need of providing 1.8V voltage regulator for the SD controller is often omitted and it runs 3.3V only. there is such a thing - PMIC, it supplies different voltages to the SoC power domains, if you wanna use UHS-I modes, your SD controller needs a separate voltage regulator inside a PMIC, then it queries voltage switch and goes from 3.3V to 1.8V - this is needed for UHS-I modes. so, it would be very probable to say, that pinephone is limited to the legacy modes just as Pine64 boards (and other a64 featuring products) are. Rock64, on the other hand, got UHS-I modes since the v3 of the board. kudos for Pine for taking that effort, even despite Rockchip's bad example of crippling down thier own SoCs (if I got it right, their own evaluation boards for rk3328, didn't get UHS-I, even the accompanying PMIC, intended for usage with rk3328 has been designed with suckily small amount of regulators).
theoretically, UHS-I could bring you an almost HDD speed of 104MB/s (read), but again, Rockchip sets the input frequency for eMMC/SD at 150 MHz (as opposed to 208 maximum, probably due to stability reasons), which with the fattest mode SDR104 will give us ~75 MB/s. I saw similar test results with Tinkerboard (rk3288) and here with RockPro64 (rk3399). for loading something meaty, a x3 speed increase is not bad. but it's not for a64. but maybe I'm wrong and Pine with pinephone has tried to overcome vendors stupidity just like they had with Rock64v3. the PMIC in question is AXP something and if it has enough regulators, it could be possible to allocate one for the a64's SD controller.
somehow, this very important aspect (putting the storage efficiency at its best possible level) gets overlooked by users, they are more concerned with such stuff like how "blobless"/open/schmopen/fsfapproved a device is or with those stupid kill switches, or keyboard running (buggy) firmware for powering off wifi. >_>
theoretically, UHS-I could bring you an almost HDD speed of 104MB/s (read), but again, Rockchip sets the input frequency for eMMC/SD at 150 MHz (as opposed to 208 maximum, probably due to stability reasons), which with the fattest mode SDR104 will give us ~75 MB/s. I saw similar test results with Tinkerboard (rk3288) and here with RockPro64 (rk3399). for loading something meaty, a x3 speed increase is not bad. but it's not for a64. but maybe I'm wrong and Pine with pinephone has tried to overcome vendors stupidity just like they had with Rock64v3. the PMIC in question is AXP something and if it has enough regulators, it could be possible to allocate one for the a64's SD controller.
somehow, this very important aspect (putting the storage efficiency at its best possible level) gets overlooked by users, they are more concerned with such stuff like how "blobless"/open/schmopen/fsfapproved a device is or with those stupid kill switches, or keyboard running (buggy) firmware for powering off wifi. >_>
ANT - my hobby OS for x86 and ARM.