| Comment by: Rob (82.15.19.75) | At: 15 Dec 2007, 13:23 | File version: 1.3 |
7457 @ 1 Ghz 2x512MB Radeon 9200 256Mb
Raw transfer speed
------------------
Screen | Register to | FAST RAM to
depth | video RAM | video RAM
-------+-------------+-------------
15 | 169.0 MB/s | 111.3 MB/s
16 | 169.3 MB/s | 125.8 MB/s
24 | 169.2 MB/s | 116.9 MB/s
32 | 169.3 MB/s | 119.1 MB/s
WritePixelArray() 320x240
-------------------------
Screen | Source: LUT8 | Source: ARGB
depth | secs | fps | MB/s | secs | fps | MB/s
-------+-------+-------+------+-------+-------+------
8 | 0.11 | 1852.4 | 135.7 | ----- | ----- | ----
15 | 0.40 | 498.6 | 146.1 | 0.31 | 324.3 | 95.0
16 | 0.20 | 983.2 | 144.0 | 0.18 | 545.7 | 79.9
24 | 0.39 | 507.7 | 148.8 | 0.31 | 320.1 | 93.8
32 | 0.41 | 483.2 | 141.6 | 0.32 | 317.3 | 93.0
ScalePixelArray() 320x240 -> 640x480
------------------------------------
Screen | Source: LUT8 | Source: ARGB
depth | secs | fps | MB/s | secs | fps | MB/s
-------+-------+-------+------+-------+-------+------
8 | ----- | ----- | ---- | ----- | ----- | ----
15 | ----- | ----- | ---- | 0.66 | 151.0 | 44.2
16 | ----- | ----- | ---- | 1.10 | 91.2 | 13.4
24 | ----- | ----- | ---- | 0.67 | 150.2 | 44.0
|
| |
| |
| Comment by: jahc (118.90.71.210) | At: 13 Dec 2007, 23:09 | File version: 1.3 |
Results from AmigaOneXE G3@800MHz, 512mb RAM, Radeon 9250SE:
Raw transfer speed
------------------
Screen | Register to | FAST RAM to
depth | video RAM | video RAM
-------+-------------+-------------
15 | 167.7 MB/s | 72.5 MB/s
16 | 167.5 MB/s | 97.1 MB/s
24 | 167.6 MB/s | 77.9 MB/s
32 | 167.5 MB/s | 77.9 MB/s
WritePixelArray() 320x240
-------------------------
Screen | Source: LUT8 | Source: ARGB
depth | secs | fps | MB/s | secs | fps | MB/s
-------+-------+-------+------+-------+-------+------
8 | 0.10 | 2022.6 | 148.1 | ----- | ----- | ----
15 | 0.61 | 327.2 | 95.8 | 0.53 | 187.1 | 54.8
16 | 0.31 | 647.1 | 94.8 | 0.25 | 403.2 | 59.1
24 | 0.63 | 317.1 | 92.9 | 0.54 | 186.7 | 54.7
32 | 0.61 | 327.4 | 95.9 | 0.54 | 186.8 | 54.7
ScalePixelArray() 320x240 -> 640x480
------------------------------------
Screen | Source: LUT8 | Source: ARGB
depth | secs | fps | MB/s | secs | fps | MB/s
-------+-------+-------+------+-------+-------+------
8 | ----- | ----- | ---- | ----- | ----- | ----
15 | ----- | ----- | ---- | 0.85 | 117.1 | 34.3
16 | ----- | ----- | ---- | 1.12 | 89.2 | 13.1
24 | ----- | ----- | ---- | 0.85 | 117.3 | 34.4
|
| |
| |
| Comment by: Petrol (90.1.21.200) | At: 13 Dec 2007, 20:11 | File version: 1.3 |
here's mine (A1XE G4-800 512MB - Radeon 9200 128MB)
Raw transfer speed
------------------
Screen | Register to | FAST RAM to
depth | video RAM | video RAM
-------+-------------+-------------
15 | 169.2 MB/s | 110.2 MB/s
16 | 169.0 MB/s | 110.9 MB/s
24 | 169.1 MB/s | 109.9 MB/s
32 | 169.1 MB/s | 109.4 MB/s
WritePixelArray() 320x240
-------------------------
Screen | Source: LUT8 | Source: ARGB
depth | secs | fps | MB/s | secs | fps | MB/s
-------+-------+-------+------+-------+-------+------
8 | 0.12 | 1674.6 | 122.6 | ----- | ----- | ----
15 | 0.45 | 443.3 | 129.9 | 0.31 | 322.6 | 94.5
16 | 0.24 | 818.2 | 119.9 | 0.28 | 357.3 | 52.3
24 | 0.46 | 435.5 | 127.6 | 0.31 | 319.9 | 93.7
32 | 0.43 | 463.6 | 135.8 | 0.33 | 302.3 | 88.6
ScalePixelArray() 320x240 -> 640x480
------------------------------------
Screen | Source: LUT8 | Source: ARGB
depth | secs | fps | MB/s | secs | fps | MB/s
-------+-------+-------+------+-------+-------+------
8 | ----- | ----- | ---- | ----- | ----- | ----
15 | ----- | ----- | ---- | 0.64 | 155.3 | 45.5
16 | ----- | ----- | ---- | 1.11 | 89.8 | 13.2
24 | ----- | ----- | ---- | 0.62 | 162.4 | 47.6
|
| |
| |
| Comment by: Jack (132.68.249.117) | At: 13 Dec 2007, 18:06 | File version: 1.3 |
@Lazi
You meant SPA() in 16 bit is slower, me curious too,
same here (radeon9k pro, A1G4XE):
WritePixelArray() 320x240
-------------------------
Screen | Source: LUT8 | Source: ARGB
depth | secs | fps | MB/s | secs | fps | MB/s
-------+-------+-------+------+-------+-------+------
15 | 0.40 | 506.1 | 148.3 | 0.35 | 288.3 | 84.5
16 | 0.23 | 862.0 | 126.3 | 0.20 | 492.6 | 72.2
24 | 0.39 | 507.3 | 148.6 | 0.35 | 287.8 | 84.3
32 | 0.40 | 504.8 | 147.9 | 0.35 | 286.8 | 84.0
ScalePixelArray() 320x240 -> 640x480
------------------------------------
Screen | Source: LUT8 | Source: ARGB
depth | secs | fps | MB/s | secs | fps | MB/s
-------+-------+-------+------+-------+-------+------
15 | ----- | ----- | ---- | 0.72 | 138.2 | 40.5
16 | ----- | ----- | ---- | 1.19 | 84.3 | 12.3
24 | ----- | ----- | ---- | 0.73 | 136.8 | 40.1
|
| |
| |
| Comment by: Lazi (84.1.59.181) | At: 13 Dec 2007, 14:57 | File version: 1.3 |
I got these results on an uA1:
What is the reason behind that WPA() is faster on 16bit and SPA() is faster on
16 bit?
WritePixelArray() 320x240
-------------------------
Screen | Source: LUT8 | Source: ARGB
depth | secs | fps | MB/s | secs | fps | MB/s
-------+-------+-------+------+-------+-------+------
15 | 0.71 | 281.1 | 82.3 | 0.83 | 121.1 | 35.5
16 | 0.36 | 549.2 | 80.4 | 0.17 | 571.6 | 83.7
ScalePixelArray() 320x240 -> 640x480
15 | ----- | ----- | ---- | 0.85 | 117.6 | 34.5
16 | ----- | ----- | ---- | 1.73 | 57.7 | 8.5
|
| |
| |
Audio (347)