AMD Zacate E-350 Performance
|
Tests
To remind you, this is just a preview of our new test method, so maybe (just maybe) today's results will be incomparable with those obtained with the final version. For this reason there won't be a traditional Excel table with complete results this time. Instead, we'll provide those right by the diagrams below.
Traditionally (for your convenience) we chose a reference processor performance of which is considered 100 points. This time it was Intel Atom 330 with NVIDIA Ion.
3D modeling software
|
Atom 330 NVIDIA Ion |
Atom D510 GMA 3150 |
Zacate E-350 HD 6310 |
Athlon II X2 250u HD 4290 |
Result |
Score |
Result |
Score |
Result |
Score |
Result |
Score |
Maya |
0.29 |
100 |
- |
- |
0.53 |
183 |
0.84 |
290 |
SolidWorks |
177.95 |
100 |
735.42 |
24 |
87.37 |
204 |
78.49 |
227 |
Creo Elements |
5355 |
100 |
- |
- |
3447 |
155 |
2038 |
263 |
Overall |
|
100 |
|
8 |
|
181 |
|
260 |
First, let's explain the 8 points of Atom D510. According to a rule introduced in this new test method, a testbed gets zero points for a benchmark it fails to complete. And the average score is naturally the sum of all points divided by the number of benchmarks (an arithmetic mean). So, of course, every failed benchmark takes a large chunk out of average score. After all, failing a benchmark is not good and should be 'rewarded' adequately.
As to this particular group of benchmarks, the reason for failure was the GMA 3150 graphics core. The corresponding testbed froze every time in SPECapc for Maya, and couldn't even visualize the application window of Creo Elements (former Pro/ENGINEER). As for SolidWorks, the completion was as formal as "the benchmark started, produced some images and finished." But just so you imagine the actual performance, a car (one of 3D models used in the test) had square wheels instead of round ones, thanks to GMA 3150.
By the way, let's have a look at specifications of graphics cores featured today. Who knows what further tests may reveal.
|
GMA 3150 (Intel Atom) |
Ion (NVIDIA Ion) |
HD 4290 (AMD 890GX) |
HD 6310 (AMD Zacate) |
Core/shader clock rate |
400/400 MHz |
450/1100 MHz |
700/700 MHz |
500/500 MHz |
RAMDAC clock rate |
400 MHz |
300 MHz |
400 MHz |
400 MHz |
Pixel pipelines |
2 |
8 |
4 |
8 |
Hardware T&L |
- |
+ |
+ |
+ |
Vertex shaders |
2.0/3.0 (software) |
4.0 |
4.1 |
5.0 |
Pixel shaders |
2.0 |
4.0 |
4.1 |
5.0 |
DirectX |
9.0c |
10 |
10.1 |
11 |
OpenGL |
1.5 |
3.3 |
3.3 |
4.1 |
OpenCL |
- |
- |
1.0 |
1.1 |
Hardware decoding |
MPEG-2 |
MPEG-2, H.264, VC-1 |
MPEG-2/4, H.264, VC-1 |
MPEG-2/4, H.264, VC-1 |
Video outputs |
D-Sub |
DP, HDMI, DVI, D-Sub |
DP, HDMI, DVI, D-Sub |
DP, HDMI, DVI, D-Sub |
Even at first glance you can see that GMA 3150 is very obsolete. Well, at least it can handle operating system's GUI and regular software. No surprise it failed two graphics benchmarks out of three.
Final rendering
|
Atom 330 NVIDIA Ion |
Atom D510 GMA 3150 |
Zacate E-350 HD 6310 |
Athlon II X2 250u HD 4290 |
Result |
Score |
Result |
Score |
Result |
Score |
Result |
Score |
Maya |
2:44:47 |
100 |
2:36:17 |
105 |
2:08:13 |
129 |
1:27:10 |
189 |
Lightwave |
1:13:14 |
100 |
1:08:19 |
107 |
1:04:27 |
114 |
0:45:15 |
162 |
3ds max |
1:50:07 |
100 |
1:34:38 |
116 |
1:29:28 |
123 |
0:53:54 |
204 |
Overall |
|
100 |
|
109 |
|
122 |
|
185 |
A perfect explanation why some people call nettop and netbook processors "socket stubs." Considering the results of Athlon II X2 250u, the difference between Atoms and Zacate seems uninteresting. But that's not the only group of benchmarks where processor had the strongest effect on results. Just so you continue reading, we can assure you that it's not that bad in further tests.
Write a comment below. No registration needed!
|
|
|
|
|