Etnaviv Open-Source Driver Adds GC7000 r6204 GPU Support For The NXP i.MX 8M Plus
One of Mesa's smaller drivers that continues advancing but not receiving as much attention as the big names is Etnaviv for providing open-source, reverse-engineered graphics support for Vivante graphics IP used across different SoCs.
Etnaviv continues advancing by its small development team for providing good open-source OpenGL support for Vivante graphics. Their latest accomplishment is now supporting the GC7000 r6204 IP, which is the GPU found within the NXP i.MX8MP.
The GC7000 r6204 / i.MX8MP GPU is similar to that of the already supported i.MX8MQ's GPU but lacks compression capabilities, only a single pixel pipe, and other minor differences that need accommodating by this Gallium3D driver.
This merge request that was open for one year but now merged into Mesa 22.2 is what gets this GPU support going should you be using an NXP i.MX8MP SoC.
Besides the Vivante GC7000UL graphics on the i.MX8MP (i.MX 8M Plus), this SoC features two or four Arm Cortex A53 cores, secondary core options with the Cortex-M7 and Cadence HiFi 4 DSP, and is intended for machine learning workloads with having a neural processing unit and dual image signal processors.
Etnaviv continues advancing by its small development team for providing good open-source OpenGL support for Vivante graphics. Their latest accomplishment is now supporting the GC7000 r6204 IP, which is the GPU found within the NXP i.MX8MP.
The GC7000 r6204 / i.MX8MP GPU is similar to that of the already supported i.MX8MQ's GPU but lacks compression capabilities, only a single pixel pipe, and other minor differences that need accommodating by this Gallium3D driver.
This merge request that was open for one year but now merged into Mesa 22.2 is what gets this GPU support going should you be using an NXP i.MX8MP SoC.
NXP i.MX 8M Plus Evaluation Board
Besides the Vivante GC7000UL graphics on the i.MX8MP (i.MX 8M Plus), this SoC features two or four Arm Cortex A53 cores, secondary core options with the Cortex-M7 and Cadence HiFi 4 DSP, and is intended for machine learning workloads with having a neural processing unit and dual image signal processors.
8 Comments