FauxVkDB
简体中文

报告 #6 跑分

参考设备 有效 至尊支持者
GPU
Adreno (TM) 830
驱动
turnip_faux123_v1.1.9-rc13 Mesa 26.3.0-devel (git-dd45fd0baa)
设备
AYN Odin 3
芯片
Snapdragon 8 Elite (CQ8725S)
CPU
6 x 3.53 GHz + 2 x 4.32 GHz, arm64-v8a
内存
10.9 GB
上报为
ayn / Odin3
收到时间
Android
15
库文件 SHA-256
49de278a13c27a6e1ead2293e4e4b2c4475b5289ee67b28ee81c18781023baf3

原始 JSON

测试条件

seconds_per_scene
6
warmup_frames
30
render_width
1920
render_height
1080
heavy_width
3840
heavy_height
2160
heavy_below_ms
8
display_width
1080
display_height
1920
display_refresh_hz
120
rounds
3
scene_set
3
vsync
OFF

Every scene draws at 1920 by 1080, whatever the screen. A scene under 8.0 ms there is too light for the chip and draws again at 3840 by 2160, as the row named with _2160p. Compare drivers on one device; across devices, compare only rows that kept the GPU busy.

测试期间显卡芯片频率低于峰值的 90%(GPU 频率):oit_linked_lists, subpasses, compute_nbody, pipeline_barriers, subpasses_2160p, compute_nbody_2160p, pipeline_barriers_2160p。可能是设备压住了显卡芯片(发热或省电模式),也可能是场景太轻,用不着更高频率。这些数据只适合和频率相近的测试比较。

场景

场景中位数 ms平均 msp95 ms轮次开测时已发热GPU 频率
oit_linked_lists
19.31
19.3219.4730967 / 1100 MHz
subpasses
2.62
2.873.4930967 / 1100 MHz
compute_nbody
3.56
3.774.4330967 / 1100 MHz
pipeline_barriers
1.58
1.762.5030967 / 1100 MHz
subpasses_2160p
9.80
10.3411.1230967 / 1100 MHz
compute_nbody_2160p
5.88
5.927.5730967 / 1100 MHz
pipeline_barriers_2160p
5.62
6.127.0530967 / 1100 MHz
单次运行(21)
场景轮状态中位数 ms帧数
oit_linked_lists0ok19.31277
subpasses0ok2.622057
compute_nbody0ok3.571544
pipeline_barriers0ok1.583344
oit_linked_lists1ok19.31277
subpasses1ok2.622060
compute_nbody1ok3.581542
pipeline_barriers1ok1.583340
oit_linked_lists2ok19.31277
subpasses2ok2.622056
compute_nbody2ok3.561547
pipeline_barriers2ok1.583340
subpasses_2160p0ok9.80546
compute_nbody_2160p0ok5.92963
pipeline_barriers_2160p0ok5.63928
subpasses_2160p1ok10.83549
compute_nbody_2160p1ok5.88964
pipeline_barriers_2160p1ok5.62928
subpasses_2160p2ok10.84549
compute_nbody_2160p2ok5.91962
pipeline_barriers_2160p2ok5.63925