Ridgeland
Member
- Dec 14, 2010
- 285
- 9
ROMs compared:
HCV9 - B6 from
http://www.androidtablets.net/forum...-tc21-hcv925-b7-kyros-7015-all-4-02-11-a.html
Watery-Kyros v1.2b from
http://www.androidtablets.net/forum...tery-kyros-v1-2b-custom-firmware-mid7015.html
And the Coby ROM update.zip
From the Coby ROM I tested three variations.
1. Base - almost nothing except a fresh install of Coby's update.zip with the addition of gscript, AndExplorer, SuperUser and AndTerm.
2. Market - Android Market plus Amazon Appstore
3. Performance - for lack of a better name, This is my everyday system, It does not have Market or Amazon Appstore. It has all the apps I have chosen as keepers.
Repeating that makes five system benchmarks:
HCV9-B6
WK1.2b
Coby {Base, Market, Performance}
For each system I used four benchmark tools:
CaffeineMark 3.0 - cm3.apk pulled from HCV9-B6 /system/app
System Benchmark (ABenchMark.apk) from Market
NBench from Market
Quadrant Standard Edition from Market
And the results:
[FONT=Courier New, monospace]CaffeineMark 3.0[/FONT]
[FONT=Courier New, monospace]- BenchMark - ...... HCV9 ..... WK ..... C-Base .. C-Mkt .. C-Perf[/FONT]
[FONT=Courier New, monospace]Sieve score = ....... 798 .... 2016 ..... 758 ..... 855 ..... 830[/FONT]
[FONT=Courier New, monospace]Loop score = ........ 946 .... 3272 ..... 934 .... 1022 .... 1021[/FONT]
[FONT=Courier New, monospace]Logic score = ....... 677 .... 3478 ..... 671 ..... 736 ..... 739[/FONT]
[FONT=Courier New, monospace]String score = ...... 846 .... 1164 ..... 784 ..... 807 ..... 828[/FONT]
[FONT=Courier New, monospace]Float scroe = ....... 676 .... 1656 ..... 610 ..... 733 ..... 463[/FONT]
[FONT=Courier New, monospace]Method score = ...... 708 .... 1811 ..... 655 ..... 763 ..... 764[/FONT]
[FONT=Courier New, monospace]Overall score = ..... 769 .... 2076 ..... 727 ..... 813 ..... 753[/FONT]
[FONT=Courier New, monospace]System Benchmark (aka ABenchMark.apk )[/FONT]
[FONT=Courier New, monospace]- BenchMark - ...... HCV9 ..... WK ..... C-Base .. C-Mkt .. C-Perf[/FONT]
[FONT=Courier New, monospace]Memory: ............. 101 ..... 108 ..... 106 ..... 103 ..... 105[/FONT]
[FONT=Courier New, monospace]CPU Integer: ........ 221 ..... 239 ..... 236 ..... 233 ..... 236[/FONT]
[FONT=Courier New, monospace]CPU Float: .......... 148 ..... 162 ..... 159 ..... 157 ..... 161[/FONT]
[FONT=Courier New, monospace]2D Graphics: ........ 200 ..... 207 ..... 197 ..... 202 ..... 213[/FONT]
[FONT=Courier New, monospace]3D Graphics: ......... 87 ...... 93 ...... 91 ...... 91 ...... 92[/FONT]
[FONT=Courier New, monospace]Database IO: ........ 100 ..... 100 ...... 85 ...... 85 ...... 75[/FONT]
[FONT=Courier New, monospace]SD Card Write: ....... 32 ...... 33 ...... 35 ...... 34 ...... 34[/FONT]
[FONT=Courier New, monospace]SD Card Read: ........ 74 ...... 73 ...... 74 ...... 70 ...... 73[/FONT]
[FONT=Courier New, monospace]Total scores: ....... 963 .... 1015 ..... 983 ..... 975 ..... 989[/FONT]
[FONT=Courier New, monospace]NBench - [/FONT]
[FONT=Courier New, monospace]- BenchMark - ...... HCV9 ..... WK ..... C-Base .. C-Mkt .. C-Perf[/FONT]
[FONT=Courier New, monospace]Memory: ........... 1.054 ... 1.133 ... 1.129 ... 1.111 ... 1.123[/FONT]
[FONT=Courier New, monospace]Integer: .......... 1.919 ... 2.084 ... 2.08 .... 2.045 ... 2.079[/FONT]
[FONT=Courier New, monospace]Float: ............ 0.546 ... 0.593 ... 0.589 ... 0.58 .... 0.589[/FONT]
[FONT=Courier New, monospace]Quadrant Standard Edition - Info only[/FONT]
[FONT=Courier New, monospace]- BenchMark - ...... HCV9 ..... WK ..... C-Base .. C-Mkt .. C-Perf[/FONT]
[FONT=Courier New, monospace]Current freq MHz .... 660 ..... 720 ..... 720 ..... 720 ..... 720[/FONT]
[FONT=Courier New, monospace]Memory: - Total always showed 157820 kB[/FONT]
[FONT=Courier New, monospace]Free: .............. 6316 .... 5044 ... 20720 ... 26092 ... 10456[/FONT]
[FONT=Courier New, monospace]Inactive: ......... 77052 ... 68408 ... 58548 ... 51508 ... 64712[/FONT]
First reaction: Run that CM3 test on WateryKyros again. I did, the second run the overall score was 2061 - No real difference. But compare the "Float" test between the test suites. There WateryKyros is more in line with the other tests.
Watery Kyros totally blows the others away in CM3. None of the others scored half as much. Why? Why did HCV9 take a hit in clock speed? I expected to prove that having Market hurts performance. That was not proven by the tests. That surprised me since I've seen too many side effects to keep Market on my everyday system. Also I expected to see BASE be noticeably better than BASE + apps (market, games etc), again data did not prove it so. Maybe the tests are subject to background activity. In Ubuntu when cron runs mlocate in the background I get a lot hard drive noise even though I'm not using the hard drive. Also except for WK CM3 I only ran one test. To be really fair I should run a dozen of each combination and average them, ain't gonna happen.
Having all these ROMs and booting each as needed is trival with gscript, using ClockWorkMod. It's also very simple without gscript. Just remember to rename your ClockWorkMod backup folder to a meaningful name like "Base", "WaterKyros.1.2b.2011.03.29" etc
These tests were all run on the same Kyros MID7015. From one Kyros to the next I expect variation in hardware from production dates, supplier changes, component lot/batch changes etc. YMMV.
Run your own tests comparing ROMs. Post them here. To format a table see Font in the FAQ. Remember than browsers reduce multiple spaces to one space, so I use "." rather than multiple spaces.
http://www.androidtablets.net/forum/misc.php?do=bbcode
All the ROMs are subject to updates so thing should change.
HCV9 - B6 from
http://www.androidtablets.net/forum...-tc21-hcv925-b7-kyros-7015-all-4-02-11-a.html
Watery-Kyros v1.2b from
http://www.androidtablets.net/forum...tery-kyros-v1-2b-custom-firmware-mid7015.html
And the Coby ROM update.zip
From the Coby ROM I tested three variations.
1. Base - almost nothing except a fresh install of Coby's update.zip with the addition of gscript, AndExplorer, SuperUser and AndTerm.
2. Market - Android Market plus Amazon Appstore
3. Performance - for lack of a better name, This is my everyday system, It does not have Market or Amazon Appstore. It has all the apps I have chosen as keepers.
Repeating that makes five system benchmarks:
HCV9-B6
WK1.2b
Coby {Base, Market, Performance}
For each system I used four benchmark tools:
CaffeineMark 3.0 - cm3.apk pulled from HCV9-B6 /system/app
System Benchmark (ABenchMark.apk) from Market
NBench from Market
Quadrant Standard Edition from Market
And the results:
[FONT=Courier New, monospace]CaffeineMark 3.0[/FONT]
[FONT=Courier New, monospace]- BenchMark - ...... HCV9 ..... WK ..... C-Base .. C-Mkt .. C-Perf[/FONT]
[FONT=Courier New, monospace]Sieve score = ....... 798 .... 2016 ..... 758 ..... 855 ..... 830[/FONT]
[FONT=Courier New, monospace]Loop score = ........ 946 .... 3272 ..... 934 .... 1022 .... 1021[/FONT]
[FONT=Courier New, monospace]Logic score = ....... 677 .... 3478 ..... 671 ..... 736 ..... 739[/FONT]
[FONT=Courier New, monospace]String score = ...... 846 .... 1164 ..... 784 ..... 807 ..... 828[/FONT]
[FONT=Courier New, monospace]Float scroe = ....... 676 .... 1656 ..... 610 ..... 733 ..... 463[/FONT]
[FONT=Courier New, monospace]Method score = ...... 708 .... 1811 ..... 655 ..... 763 ..... 764[/FONT]
[FONT=Courier New, monospace]Overall score = ..... 769 .... 2076 ..... 727 ..... 813 ..... 753[/FONT]
[FONT=Courier New, monospace]System Benchmark (aka ABenchMark.apk )[/FONT]
[FONT=Courier New, monospace]- BenchMark - ...... HCV9 ..... WK ..... C-Base .. C-Mkt .. C-Perf[/FONT]
[FONT=Courier New, monospace]Memory: ............. 101 ..... 108 ..... 106 ..... 103 ..... 105[/FONT]
[FONT=Courier New, monospace]CPU Integer: ........ 221 ..... 239 ..... 236 ..... 233 ..... 236[/FONT]
[FONT=Courier New, monospace]CPU Float: .......... 148 ..... 162 ..... 159 ..... 157 ..... 161[/FONT]
[FONT=Courier New, monospace]2D Graphics: ........ 200 ..... 207 ..... 197 ..... 202 ..... 213[/FONT]
[FONT=Courier New, monospace]3D Graphics: ......... 87 ...... 93 ...... 91 ...... 91 ...... 92[/FONT]
[FONT=Courier New, monospace]Database IO: ........ 100 ..... 100 ...... 85 ...... 85 ...... 75[/FONT]
[FONT=Courier New, monospace]SD Card Write: ....... 32 ...... 33 ...... 35 ...... 34 ...... 34[/FONT]
[FONT=Courier New, monospace]SD Card Read: ........ 74 ...... 73 ...... 74 ...... 70 ...... 73[/FONT]
[FONT=Courier New, monospace]Total scores: ....... 963 .... 1015 ..... 983 ..... 975 ..... 989[/FONT]
[FONT=Courier New, monospace]NBench - [/FONT]
[FONT=Courier New, monospace]- BenchMark - ...... HCV9 ..... WK ..... C-Base .. C-Mkt .. C-Perf[/FONT]
[FONT=Courier New, monospace]Memory: ........... 1.054 ... 1.133 ... 1.129 ... 1.111 ... 1.123[/FONT]
[FONT=Courier New, monospace]Integer: .......... 1.919 ... 2.084 ... 2.08 .... 2.045 ... 2.079[/FONT]
[FONT=Courier New, monospace]Float: ............ 0.546 ... 0.593 ... 0.589 ... 0.58 .... 0.589[/FONT]
[FONT=Courier New, monospace]Quadrant Standard Edition - Info only[/FONT]
[FONT=Courier New, monospace]- BenchMark - ...... HCV9 ..... WK ..... C-Base .. C-Mkt .. C-Perf[/FONT]
[FONT=Courier New, monospace]Current freq MHz .... 660 ..... 720 ..... 720 ..... 720 ..... 720[/FONT]
[FONT=Courier New, monospace]Memory: - Total always showed 157820 kB[/FONT]
[FONT=Courier New, monospace]Free: .............. 6316 .... 5044 ... 20720 ... 26092 ... 10456[/FONT]
[FONT=Courier New, monospace]Inactive: ......... 77052 ... 68408 ... 58548 ... 51508 ... 64712[/FONT]
First reaction: Run that CM3 test on WateryKyros again. I did, the second run the overall score was 2061 - No real difference. But compare the "Float" test between the test suites. There WateryKyros is more in line with the other tests.
Watery Kyros totally blows the others away in CM3. None of the others scored half as much. Why? Why did HCV9 take a hit in clock speed? I expected to prove that having Market hurts performance. That was not proven by the tests. That surprised me since I've seen too many side effects to keep Market on my everyday system. Also I expected to see BASE be noticeably better than BASE + apps (market, games etc), again data did not prove it so. Maybe the tests are subject to background activity. In Ubuntu when cron runs mlocate in the background I get a lot hard drive noise even though I'm not using the hard drive. Also except for WK CM3 I only ran one test. To be really fair I should run a dozen of each combination and average them, ain't gonna happen.
Having all these ROMs and booting each as needed is trival with gscript, using ClockWorkMod. It's also very simple without gscript. Just remember to rename your ClockWorkMod backup folder to a meaningful name like "Base", "WaterKyros.1.2b.2011.03.29" etc
These tests were all run on the same Kyros MID7015. From one Kyros to the next I expect variation in hardware from production dates, supplier changes, component lot/batch changes etc. YMMV.
Run your own tests comparing ROMs. Post them here. To format a table see Font in the FAQ. Remember than browsers reduce multiple spaces to one space, so I use "." rather than multiple spaces.
http://www.androidtablets.net/forum/misc.php?do=bbcode
All the ROMs are subject to updates so thing should change.
Last edited: