Home / Adreno (TM) 830 / Compare

turnip_faux123_v1.2.0 vs Qualcomm Technologies Inc. Adreno Vulkan Driver

On Adreno (TM) 830 (AYANEO Pocket FIT Elite, HONOR AAK-AN00, HONOR Magic V5, nubia NX789J, OnePlus 13, OnePlus 13, OnePlus 13s, POCO F8 Pro, realme GT 7 Pro, Samsung Galaxy S25, Samsung Galaxy S25 Ultra, Samsung Galaxy S25 Ultra, Samsung Galaxy S25 Ultra, Samsung Galaxy S25+, Xiaomi 15, Xiaomi 15, Xiaomi 15 Ultra, ZTE NP05J). The first driver named is compared against the second, which is the baseline.

StockQualcomm Technologies Inc. Adreno Vulkan DriverDriver Build: f61dec9117, I68b0a64d7e, 1774510019

Customturnip_faux123_v1.2.0Mesa 26.3.0-devel (git-dd45fd0baa)

Graphics system version (Vulkan)1.3.284 to 1.4.359

Optional features (extensions)165 to 173+31 -23

Features that pass their check (validated)44 to 69turnip_faux123_v1.2.0: sent from a device this site could not confirm (unverified)25 pass only on turnip_faux123_v1.2.0 0 pass only on Qualcomm Technologies Inc. Adreno Vulkan Driver

Speed test scenes (benchmark)0 faster 4 slower37% slower overall (geometric mean)

Games each driver could run (an estimate from its feature list): green likely, yellow maybe, orange limited, red no

StockDX92002+DX102006+DX112009+DX122015+Switch

CustomDX92002+DX102006+DX112009+DX122015+Switch

hdr_2160p32% slower
oit_linked_lists43% slower
pipeline_barriers_2160p35% slower
subpasses_2160p39% slower

Custom driver against stock, per scene. Bars to the right are faster, to the left slower.

Small areas of the picture differ from the stock driver in 2 scenes (pipeline_barriers: 1432 of 2304 cells, pipeline_barriers_2160p: 1291 of 2304 cells). The picture matches overall, so this is a local rendering fault, not a different scene.

Not on equal terms: the GPU clock differed during 1 of these scenes (for example hdr_2160p: stock at 90% of the device's maximum, custom at 100%). Part of the frame-time difference comes from the clock the device chose, not from the driver.

Qualcomm Technologies Inc. Adreno Vulkan Driver details turnip_faux123_v1.2.0 details

Feature checks that differ (42)

Unverified turnip_faux123_v1.2.0: The results below were sent from a device this site could not confirm. A result the site counts replaces them.

Feature	Stock	Custom	Change
extendedDynamicState3ColorBlendEnable VK_EXT_extended_dynamic_state3	unsupported	pass	gained
extendedDynamicState3ColorBlendEquation VK_EXT_extended_dynamic_state3	unsupported	pass	gained
extendedDynamicState3ColorWriteMask VK_EXT_extended_dynamic_state3	unsupported	pass	gained
extendedDynamicState3DepthClampEnable VK_EXT_extended_dynamic_state3	unsupported	pass	gained
extendedDynamicState3LogicOpEnable VK_EXT_extended_dynamic_state3	unsupported	pass	gained
extendedDynamicState3PolygonMode VK_EXT_extended_dynamic_state3	unsupported	pass	gained
depthClamp VkPhysicalDeviceFeatures	untested	pass	gained
largePoints VkPhysicalDeviceFeatures	untested	pass	gained
compat.helper_loop_terminator core	untested	pass	gained
compat.nested_map core	untested	pass	gained
core.blend_advertised_formats core	untested	pass	gained
core.compute_big_stack core	untested	pass	gained
core.copy_wide_row_length core	fail	pass	gained
core.host_coherent_readback core	untested	pass	gained
core.mrt_partial_kill core	untested	pass	gained
core.ubo_update_buffer core	untested	pass	gained
core.ubo_written_after_record core	untested	pass	gained
core.varying_subset core	untested	pass	gained
mem.image_padding core	untested	pass	gained
perf.fb_local_barrier_large_draws core	untested	pass	gained
perf.fb_local_barrier_overdraw core	untested	pass	gained
stage.geometry_max_output core	untested	pass	gained
stage.geometry_point_size core	untested	pass	gained
stress.scratch_per_batch core	untested	pass	gained
sync.first_submit_clear_then_compute core	untested	pass	gained
extendedDynamicState3AlphaToCoverageEnable VK_EXT_extended_dynamic_state3	unsupported	untested	changed
extendedDynamicState3AlphaToOneEnable VK_EXT_extended_dynamic_state3	unsupported	untested	changed
extendedDynamicState3ConservativeRasterizationMode VK_EXT_extended_dynamic_state3	unsupported	untested	changed
extendedDynamicState3DepthClipEnable VK_EXT_extended_dynamic_state3	unsupported	untested	changed
extendedDynamicState3DepthClipNegativeOneToOne VK_EXT_extended_dynamic_state3	unsupported	untested	changed
extendedDynamicState3ExtraPrimitiveOverestimationSize VK_EXT_extended_dynamic_state3	unsupported	untested	changed
extendedDynamicState3LineRasterizationMode VK_EXT_extended_dynamic_state3	unsupported	untested	changed
extendedDynamicState3ProvokingVertexMode VK_EXT_extended_dynamic_state3	unsupported	untested	changed
extendedDynamicState3RasterizationSamples VK_EXT_extended_dynamic_state3	unsupported	untested	changed
extendedDynamicState3RasterizationStream VK_EXT_extended_dynamic_state3	unsupported	untested	changed
extendedDynamicState3SampleLocationsEnable VK_EXT_extended_dynamic_state3	unsupported	untested	changed
extendedDynamicState3SampleMask VK_EXT_extended_dynamic_state3	unsupported	untested	changed
extendedDynamicState3TessellationDomainOrigin VK_EXT_extended_dynamic_state3	unsupported	untested	changed
shaderResourceMinLod VkPhysicalDeviceFeatures	unsupported	untested	changed
shaderResourceResidency VkPhysicalDeviceFeatures	unsupported	untested	changed
sparseResidency16Samples VkPhysicalDeviceFeatures	untested	unsupported	changed
sparseResidencyAliased VkPhysicalDeviceFeatures	unsupported	untested	changed

Speed test scenes (benchmark)

unverified turnip_faux123_v1.2.0: The times below were sent from a device this site could not confirm. A result the site counts replaces them.

The two drivers were not measured on the same set of device models. Frame times from different devices do not compare drivers, so read these numbers with that in mind.

Scene	Stock ms	Custom ms	Difference	Starts	GPU clock, MHz	Picture vs stock
compute_nbody	8.51	n/a	only one side	3 of 3 vs 3 of 3	432 vs 525 not comparable	no large difference 0.0 no small difference worst 0.0 64x36
compute_nbody_2x	n/a	17.79	only one side	? vs 3 of 3	? vs 525 clock unknown	not checked
hdr_2160p	11.03 typical 10.28	16.11 typical 16.15	32% slower	3 of 3 vs 3 of 3	750 vs 525 not comparable	no large difference 0.0 no small difference worst 0.0 64x36
oit_linked_lists	21.66 typical 21.56	37.78 typical 37.67	43% slower	3 of 3 vs 3 of 3	832 vs 511	no large difference 0.0 no small difference worst 0.0 64x36
pipeline_barriers	2.38	n/a	too light, see pipeline_barriers_2160p	3 of 3 vs 3 of 3	832 vs 494 not comparable	no large difference 6.9 differs in 1432 small areas worst 44.0 64x36
pipeline_barriers_2160p	7.52 typical 7.32	11.52 typical 10.66	35% slower	3 of 3 vs 3 of 3	816 vs 511	no large difference 6.9 differs in 1291 small areas worst 70.0 64x36
subpasses	3.12	n/a	too light, see subpasses_2160p	3 of 3 vs 3 of 3	832 vs 525	not checked
subpasses_2160p	12.01 typical 11.94	19.84 typical 20.27	39% slower	3 of 3 vs 3 of 3	832 vs 525	not checked

Time to draw one frame, per scene: the average (mean) frame time of the scene's best round, which is how many frames the driver delivered; lower is faster. "Typical" beside it is the median frame, which can sit well under the average when a driver's frames come in two speeds. A pair of reports made before per-round averages were recorded is compared on the median. A scene whose rounds disagree by more than 15% is left out. A scene marked "paced" was held at that mean frame time by the driver or the display queue, together with at least one other scene of the same run; its time says how the frames were spaced, not how fast the driver rendered, so it is not compared. A scene marked "too light" took under 8 ms at the standard size (1920 by 1080) on both drivers, where the screen's frame queue spaces the frames more than the chip does; its heavier run is the one compared: the row ending in _2160p is the scene drawn at 3840 by 2160, and compute_nbody_2x is that scene with twice the particles. "Faster" and "slower" describe turnip_faux123_v1.2.0 against Qualcomm Technologies Inc. Adreno Vulkan Driver. "Picture vs stock" compares one fixed frame per scene as a coarse colour grid (0 is identical, above 40 is a different picture); it finds a scene that was not drawn. Where both reports carry the finer 64x36 grid, the second figure is how many of its 2,304 cells differ by more than 4 of 255, and "worst" is the largest single cell; this is what finds a small rendering error in a picture that otherwise matches. A row marked 16x9 comes from a report made before the finer grid existed. "Starts" is how many of a scene's process starts reached a measured round. A GPU clock marked "not comparable" differed by more than 5% of the device's maximum between the two timings.

Technical differences (Vulkan)

Extensions only in turnip_faux123_v1.2.0 (31)

VK_AMD_buffer_marker
VK_ARM_rasterization_order_attachment_access
VK_EXT_attachment_feedback_loop_dynamic_state
VK_EXT_attachment_feedback_loop_layout
VK_EXT_custom_resolve
VK_EXT_debug_marker
VK_EXT_depth_clamp_control
VK_EXT_dynamic_rendering_unused_attachments
VK_EXT_extended_dynamic_state3
VK_EXT_external_memory_dma_buf
VK_EXT_graphics_pipeline_library
VK_EXT_host_image_copy
VK_EXT_image_drm_format_modifier
VK_EXT_legacy_dithering
VK_EXT_legacy_vertex_attributes
VK_EXT_map_memory_placed
VK_EXT_memory_budget
VK_EXT_nested_command_buffer
VK_EXT_non_seamless_cube_map
VK_EXT_primitive_restart_index
VK_EXT_shader_replicated_composites
VK_EXT_shader_uniform_buffer_unsized_array
VK_EXT_zero_initialize_device_memory
VK_GOOGLE_decorate_string
VK_GOOGLE_hlsl_functionality1
VK_KHR_compute_shader_derivatives
VK_KHR_maintenance7
VK_KHR_pipeline_library
VK_KHR_shader_relaxed_extended_instruction
VK_NV_compute_shader_derivatives
VK_VALVE_mutable_descriptor_type

Extensions only in Qualcomm Technologies Inc. Adreno Vulkan Driver (23)

VK_ANDROID_external_format_resolve
VK_EXT_astc_decode_mode
VK_EXT_blend_operation_advanced
VK_EXT_depth_range_unrestricted
VK_EXT_device_fault
VK_EXT_external_memory_acquire_unmodified
VK_EXT_fragment_density_map
VK_EXT_fragment_density_map2
VK_EXT_image_compression_control_swapchain
VK_EXT_pipeline_protected_access
VK_KHR_ray_tracing_position_fetch
VK_KHR_shader_maximal_reconvergence
VK_KHR_shader_quad_control
VK_KHR_shared_presentable_image
VK_NV_optical_flow
VK_QCOM_filter_cubic_weights
VK_QCOM_fragment_density_map_offset
VK_QCOM_image_processing2
VK_QCOM_render_pass_store_ops
VK_QCOM_render_pass_transform
VK_QCOM_rotated_copy_commands
VK_QCOM_tile_properties
VK_QCOM_ycbcr_degamma

Core features that differ (20)

multiviewGeometryShader Vulkan 1.1	unsupported to pass
protectedMemory Vulkan 1.1	pass to unsupported
shaderInputAttachmentArrayDynamicIndexing Vulkan 1.2	pass to unsupported
shaderInputAttachmentArrayNonUniformIndexing Vulkan 1.2	pass to unsupported
bresenhamLines Vulkan 1.4	unsupported to pass
dynamicRenderingLocalRead Vulkan 1.4	unsupported to pass
globalPriorityQuery Vulkan 1.4	unsupported to pass
hostImageCopy Vulkan 1.4	unsupported to pass
indexTypeUint8 Vulkan 1.4	unsupported to pass
maintenance5 Vulkan 1.4	unsupported to pass
maintenance6 Vulkan 1.4	unsupported to pass
pipelineRobustness Vulkan 1.4	unsupported to pass
pushDescriptor Vulkan 1.4	unsupported to pass
rectangularLines Vulkan 1.4	unsupported to pass
shaderExpectAssume Vulkan 1.4	unsupported to pass
shaderFloatControls2 Vulkan 1.4	unsupported to pass
shaderSubgroupRotate Vulkan 1.4	unsupported to pass
shaderSubgroupRotateClustered Vulkan 1.4	unsupported to pass
vertexAttributeInstanceRateDivisor Vulkan 1.4	unsupported to pass
vertexAttributeInstanceRateZeroDivisor Vulkan 1.4	unsupported to pass

Limits that differ (20)

Limit	Stock	Custom
maxTexelBufferElements	134217728	536870911
bufferImageGranularity	1	64
sparseAddressSpaceSize	562949953421311	274877906944
maxPerStageResources	50331648	16777216
maxDescriptorSetUniformBuffersDynamic	32	16
maxDescriptorSetStorageBuffersDynamic	16	8
maxVertexInputAttributeOffset	4096	4095
maxGeometryInputComponents	128	64
maxGeometryOutputVertices	1024	256
maxGeometryTotalOutputComponents	131072	1024
maxFragmentInputComponents	112	128
minMemoryMapAlignment	64	4096
minUniformBufferOffsetAlignment	256	64
minStorageBufferOffsetAlignment	64	4
maxFramebufferLayers	2048	1024
discreteQueuePriorities	3	2
pointSizeRange	1, 4095	1, 4092
optimalBufferCopyOffsetAlignment	64	128
optimalBufferCopyRowPitchAlignment	64	128
nonCoherentAtomSize	1	64

Formats supported by only one driver (5 gained, 1 lost)

Only in turnip_faux123_v1.2.0:

B8G8R8A8_SSCALED
B8G8R8A8_USCALED
B8G8R8_SINT
B8G8R8_SNORM
B8G8R8_UINT

Only in Qualcomm Technologies Inc. Adreno Vulkan Driver:

G10X6_B10X6R10X6_2PLANE_420_UNORM_3PACK16