| This Site has been put up for Testing Purposes — Bryce Render Tests (Updated 22-JAN-2009) | |
 |
The Test In order to get an idea about the time the embeded fanless system needs to render a picture with depth of field (DOF), the picture at left was used and rendered to a size of 800 x 600 pixels. If you would like to repeat the test, get Bryce 5.5 file by rightclicking on the link (45 KB). This is nothing you will have any use for, except for testing. The text was done with Wings3D, the object imported to Bryce, scaled, put on a material and rendered. Nothing really fancy. |
The .br5 file opens in Bryce 5.0, 5.5, 6.0 and 6.1. Gamma Correction engaged for all tests. | |
Computer |
Bryce 5.0 |
Bryce 5.5 |
Bryce 6.0 |
Bryce 6.1 |
Processor |
Clock |
FSB |
Instruction |
|---|---|---|---|---|---|---|---|---|
P2L97S |
- |
1:13:46 |
- |
- |
Pentium II |
233 |
67 |
MMX |
EPIAM1 |
- |
0:53:48 |
1:00:25 |
1:01:33 |
VIA C3 |
600 |
133 |
MMX, 3DNOW |
AB40V12 |
0:18:12 |
0:16:50 |
0:16:52 |
0:16:23 |
Pentium IV |
1'600 |
100 |
MMX, SSE, SSE2 |
Laptorn |
0:17:27 |
0:15:44 |
0:16:00 |
0:16:30 |
Celeron |
2'600 |
100 |
MMX, SSE, SSE2 |
Foxconn *a) |
0:08:52 |
0:08:13 |
- |
0:08:05 |
Pentium D 950 |
3'400 |
800 |
MMX, SSE, SSE2, SSE3, EM64T |
Foxconn *b) |
0:08:52 + | + 0:08:24 |
0:04:33 |
0:04:28 |
Pentium D 950 |
3'400 |
800 |
MMX, SSE, SSE2, SSE3, EM64T | |
*a): only one core in use for Bryce 5.0, 5.5 and 6.1. |
*b): Bryce 5.0 on core 1, Bryce 5.5 on core 2 together; Bryce 6.0 uses both cores, 6.1 High Priority. | |||||||
Foxconn *c) |
- |
- |
0:03:42 |
Pentium D 950 |
3'400 |
800 |
MMX, SSE, SSE2, SSE3, EM64T | |
*c): all of the above, running Win2K as a virtual machine on Foxconn so both processors are used. Without P2L97S and Epiam1 render time is 3:30. | ||||||||
LEKMISNB0016 |
- |
- |
- |
0:03:11 |
Intel Core Duo T2500 |
2'000 |
677 |
? |
Using two similar machines connected over a virtual network (1.2 Mbps) using Lightning 2.0b (Bryce 6.0) renders in 2'23" (4 cores). | ||||||||
| ||||||||
Eugene Babiak |
- |
0:11:40 |
- |
- |
AMD Duron |
1'300 |
200 |
? |
Eugene Babiak |
- |
- |
0:03:06 |
0:03:04 |
AMD Athelon64 |
2'216 |
800 |
MMX, 3DNow!, SSE, SSE2, SSE3 |
David Brinnen |
- |
0:05:14 |
0:05:52 |
- |
Pentium 4 Prescott |
3'400 |
800 |
MMX, SSE, SSE2, SSE3 |
Christian Weiten |
- |
0:03:35 |
- |
0:01:03 |
QX6800 |
2'930 |
? |
5.5: 1/4 core used, |
Robert Glaisen |
- |
- |
- |
0:12:31 |
PowerPC G4 |
1'200 |
? |
MAC 10.4.11 |
Robert Glaisen |
- |
- |
- |
0:00:41 |
Intel Core i5 |
2'666 |
? |
MAC 10.6.2 |
James Dykes |
- |
- |
- |
0:10:00 |
Pentiun IV |
3'000 |
? |
- |
Rashad Carter |
- |
- |
- |
0:00:46 |
2 x Intel Xeon Quadcore Processors |
2'500 |
1'333 |
(8 cores) |
lucindrea |
- |
- |
- |
0:02:28 |
AMD 64 Opteron |
2'750 |
305 |
(2 cores) |
Spektyr |
- |
- |
- |
0:11:30 |
Dual P3 |
700 |
100 |
(2 cores) |
Spektyr |
- |
- |
- |
0:08:41 |
Dual P3 |
933 |
133 |
(2 cores) |
Spektyr |
- |
- |
- |
0:07:58 |
Dual P3 |
1'000 |
100 |
(2 cores) |
Steve L |
- |
0:16:22 |
- |
0:08:22 |
Pentium III |
930 |
133 |
|
Steve L |
- |
0:07:56 |
- |
0:04:50 |
Pentium D925 |
3'000 |
800 |
(2 cores) |
Note on Network Rendering: Bryce 5.0, 5.5, 6.0 and 6.1 can be used as network render server/host and have separate network render clients. It is possible to use the host as client as well. Note that the slowest machine is usually also the last to finish and hence pushes the total time up.
Bryce 6.1 uses only 1 core on dual core machines. Network rendering the test file using P2L97S, EPIAM1, AB40V12, Laptorn (all Win2k) and Foxconn (XP) results in 4:34, 6 seconds more than just using Foxconn (b). If the render job is started on AB40V12 instead of Foxconn, the file needs 6:00. Windows 2K needs over 1 minute to start the render on all clients, XP (Foxconn) starts after less than 5 seconds.
In order to use both cores in a dual core machine, install a virtual PC and run Lightning on the main machine and on the virtual machine — it gets another IP an can thus take the role of a second client. This work around has been suggested by Hans Schneider (schneider01). There are some caveats using a virtual PC. I performed some Tests with all Lightning clients. Lightning 2.0b that came with Bryce 6.0.18.6 uses all cores as mentioned by Buaya in a DAZ forum. This network render application is compatible with Bryce 6.0 and Bryce 6.1.
Foxconn:
This is currently the main machine. It is performing at
• 1'685 Million Double Precision Instructions per Second (MIPS); or 1'685 Whetstones
per processor and
• 2'577 Million Integer Instructions per Second (MIPS); or 2'577 Drystones per
processor.
AB40V12:
This is the old main machine. It is performing at
• 803 Million Double Precision Instructions per Second (MIPS); or 803 Whetstones and
• 1'523 Million Integer Instructions per Second (MIPS); or 1'523 Drystones.
Whet- & Drystones measured by BOINC (Berkeley Open Infrastructure for Network Computing).
< Zurück zur Hauptseite.