Related
But why?!?
The Qualcomm CPU should kick ass. (Granted, it might not be faster than the Exynos in SGII but it shouldn't be MUCH slower) The GPU is also on par, if not faster than the Mali and Tegra GPU...
So why do my Benchmark scores kinda suck? My Tegra2 powered 7 inch tablet usually gets way better scores than my Evo3D... On some benchmarks it's even noticably behind LG Optimus 2X and similar phones that it should be AT LEAST on par with, if not faster...
Is it only the ROM that's at fault here or what's going on? What benchmark scores do the rest of you get? Smartbench, Antutu, Quadrant, CF Bench and all the rest of them?
(And yes, the phone is smooth enough in everyday use. And benchmarks don't mean everything, but it DOES mean something...)
It only means anything if all you do is bench testing on your phone. Try play games, watch films and take some picture with youe phone instead.
The mal 400 is way faster than anything else in android phones.
Never the less, nothing comes close to taxing the 3d processor or go. So don't worry about benches.
What more do you really need out of EVO3D benches to see that with the right ROM, and not having 46 widgets and livewallpaper running in the background, you can get some pretty nice scores:
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
Various roms, kernels, no telling really what at the time, but pretty much in line. There's something wrong with the Rom you're on, or your setup if you can't get good benches with this phone, IMHO.
What ROM, kernel, are you running, and what's a 'bad' score in your book?
<<Robotron 2084 = Perfection>>
I love it when people jizz over benchmark scores. Like justj said, try some apps, games, and movies. If they run fantastically, then you're phone is running great, and you don't need to worry about silly benchmark numbers.
Real Use >>>>>>> Benchmarks
mazook98 said:
What ROM, kernel, are you running, and what's a 'bad' score in your book?
<<Robotron 2084 = Perfection>>
Click to expand...
Click to collapse
Ah, thank you... I'll turn off everything but the benchmarking app, and get back to you with the scores... As for "bad score" well, I'd expect it to beat my tablet (Tegra2, 1024*600 resolution) as well as the LG 2X, Atrix, Optimus 3D, and most of the newer phones on the market, really. Which it at present doesn't do. It gets beaten by the Nexus S in some benchmarks!
Video and games and so on run pretty fast and smooth, so overall I'm happy and don't place TOO much weight on benchmarks. I just find it odd, because benchmarks to certain degree DO reflect real world performance.
I'll get back to you with scores and which ROM etc. Off the top of my head, I remember it's running 2.3.4 Rogers ROM...
vszulc said:
Ah, thank you... I'll turn off everything but the benchmarking app, and get back to you with the scores... As for "bad score" well, I'd expect it to beat my tablet (Tegra2, 1024*600 resolution) as well as the LG 2X, Atrix, Optimus 3D, and most of the newer phones on the market, really. Which it at present doesn't do. It gets beaten by the Nexus S in some benchmarks!
Video and games and so on run pretty fast and smooth, so overall I'm happy and don't place TOO much weight on benchmarks. I just find it odd, because benchmarks to certain degree DO reflect real world performance.
I'll get back to you with scores and which ROM etc. Off the top of my head, I remember it's running 2.3.4 Rogers ROM...
Click to expand...
Click to collapse
Didn't know you had a GSM phone.... Those are capped at 80% of potential speed by the Canadian Government.
(Nah, j/k).
Good luck on outbenchmarking the LG2X... (you'll need it). Benchmarks can't save a bad phone, but if they could, that's where they'd start. I've seen multiple screens from multiple Lg2x users running CM7 with Quadrant Standard scores of over 5K. That's insane.
We'll probably never see that on an EVO3D, but it doesn't matter, because ultimately, the LG2X isn't half the phone at the end of the day.
Just my 2 cents.
Edit: If your tablet happens to be a Viewsonic Gtab, well, forget that notion as well. That basically is at the top of every benchmark 'top scores' list.
I've got one.
Is it fast?
Incredibly fast.
Would I trade it right now for a "slower" tablet that was a "better" tablet?
You betcha.
Righty o...
I could always upload screenshots later if you think that might be a good idea, but here are some scores:
Antutu: 4442, which puts it behind both the S2, the Optimus 2X and the Motorola Xoom (!?!)
Nenamark 2 reported a somewhat respectable 25.6.
Smartbench 2011 (Supposedly the only benchmark suite that takes dualcore into consideration): Productivity: 2397 Games: 1650
(Which is behind the Transformer Tab and Atrix in both categories, and even behind the Nexus S and Galaxy S in Games)
Quadrant:1491 (Just ahead of Moto Droid X and Nexus One at 1250!)
Devicemark: 1325 (Equal to Desire HD, but behind the Atrix and LG 2X)
CF Bench: 8740, 1364, 4314 (Way behind both the Atrix and the Sensation?!)
I think that's it as far as the benchmarks I ran. I can upload the screenshots if you think it might help. As for the phone, it's a Rogers phone with the 2.3.4 Gingerbread, HTC software: 1.20.631.2
(BTW: What does "SMP PREEMPT" mean in the Kernel version?!?)
And no, the tablet is an Iconia A100. Zippy little thing, though I don't understand how it can run corners around my Evo with a Tegra CPU and more pixels to push...
vszulc said:
Righty o...
I could always upload screenshots later if you think that might be a good idea, but here are some scores:
Antutu: 4442, which puts it behind both the S2, the Optimus 2X and the Motorola Xoom (!?!)
Nenamark 2 reported a somewhat respectable 25.6.
Smartbench 2011 (Supposedly the only benchmark suite that takes dualcore into consideration): Productivity: 2397 Games: 1650
(Which is behind the Transformer Tab and Atrix in both categories, and even behind the Nexus S and Galaxy S in Games)
Quadrant:1491 (Just ahead of Moto Droid X and Nexus One at 1250!)
Devicemark: 1325 (Equal to Desire HD, but behind the Atrix and LG 2X)
CF Bench: 8740, 1364, 4314 (Way behind both the Atrix and the Sensation?!)
I think that's it as far as the benchmarks I ran. I can upload the screenshots if you think it might help. As for the phone, it's a Rogers phone with the 2.3.4 Gingerbread, HTC software: 1.20.631.2
(BTW: What does "SMP PREEMPT" mean in the Kernel version?!?)
Click to expand...
Click to collapse
There is something seriously wrong. The least you should be getting is around 2000-2500.
Yeah, I turned off background apps, etc. What scores do people with the Rogers ROM usually get? Would backing up and then a full restore to factory settings help?
I'm rooted btw, but didn't install an alternate rom.
No question that the gs2 benches well but after having one for a couple days I missed my 3d..I trade benches for a smooth and usable UI ..but ocd I can beat the gs2 in every bench app there is ..I never felt the supposed increase in speed that the epic has ..the 3d opened my apps faster and was more stable using games that utilized the gyro ie homerun baseball 3d ..almost unplayable on the gs2..our phones have plenty of power either way we have plenty of power on tap
Nope. Quadrant seems about right from what I remember with the Rogers ROM. Not much better with LeeDroid either.
Might have to do with the inability of GSM devs to tweak the kernel and overclock as HTC messed up the board files in the kernel sources they released for the GSM.
This is what I've got with aosp oc...
Sent from my PG86100 using xda premium
vszulc said:
And no, the tablet is an Iconia A100. Zippy little thing, though I don't understand how it can run corners around my Evo with a Tegra CPU and more pixels to push...
Click to expand...
Click to collapse
Oooh... I'm VERY interested in the A100... I'm actually going to Best Buy today, as they finally have them in stock @~300$... But they have the Flyer as well at the same price...
The a100 IS sporting a dual-core Tegra2 A9 processor, is it not?
And since HC 3.2 in a brand-new device is probably optimized better, speed-wise, I, for one, would expect it to out-perform the 3d in a benchmark test, even @ 1000Mhz.
How do you like the A100, overall?
http://db.tt/sy1eZYiy
http://db.tt/dKdJCGpJ
I won...
ssojyeti2 said:
http://db.tt/sy1eZYiy
http://db.tt/dKdJCGpJ
I won...
Click to expand...
Click to collapse
No, you kinda lost, because you made anyone who might be interested download a dropbox file instead of just doing this:
EDIT: WHOOPS, Sorry dude... If I had noticed you were using drocap or whatever I would've understood your need to dropbox it.
Some of those market screenshot apps just do not work. Native screenshot works fine for me..END EDIT.
(whatevs, no biggie, lol.. Nice scores.. Which Rom/kernel combo is that, if you don't mind me asking?
I'm gonna just guess and say that's an OC'd aosp rom.
Yes?
mazook98 said:
How do you like the A100, overall?
Click to expand...
Click to collapse
I actually really like it! When I was in the store, I let all the tablets do a sunspider test, and the A100 actually came out ahead of everybody, even the iPad2, so it IS a zippy little fella.
Anything in particular you'd like to know? The two cameras are just fine, the few times I used them for anything else than a webcam, they did the job fine.
Speakers are great and really loud. The screen is a bit on the dim side, but turning off the auto light in settings take care of that. It's really comfortable to hold, even for extended periods of time, even if the plastic might make it feel and look less than a premium product.
Batterylife is kinda meh, I usually don't get more than 5 hours out of it, but then again it's pretty rare that I take it somewhere where there isn't a charger around. The only thing I can think off that would require your tab to be on for 5 hours is a plane trip, and even most of those have chargers these days
The only caveat I would have, is that as a Evo3D owner a 7 inch screen might be a bit too small. I mean the difference between 7 and 4.3 inches isn't that big in daily use, and after getting the Evo I kinda find the 7 inch screen a bit on the small side. Might just be me though. If I had to choose any tablet today, I think I'd go for the Galaxy. The 10 inch size is just right, and it weighs pretty much the same as the A100, and since getting a tablet I've found out that weight means more in terms of portability than just size.
The only other thing I've found to be annoying would probably also apply to the Flyer: There aren't too many tablet apps available as it is, and the few tablet apps there are, aren't always compatible with 7 inch tabs. There are way more apps and games available for 10 inch tablets.
Oh yeah, one more thing: The A100 is super easy too root. Took me like 5 mins. But the support for it is pretty poor. There's like no ROMs available for it at all, which I find really weird. Not yet anyways, but hopefully it'll gain some devs...
vszulc said:
Oh yeah, one more thing: The A100 is super easy too root. Took me like 5 mins. But the support for it is pretty poor. There's like no ROMs available for it at all, which I find really weird. Not yet anyways, but hopefully it'll gain some devs...
Click to expand...
Click to collapse
BTW, I posted this somewhere else, and you may already know this trick, but if you wanna "revert your UI to Gingerbread" on an HC tablet (I did.... Just not into HC yet), and since you've already rooted, if you change your density to 171 (or above) , it will revert the ui to GB, (something you might be more used to, or like better) , while maintaining compatibility with those 3.0 apps.
<<work in progress>>
Reviews and complaints but didn't see a thread dedicated to synthetics. So here it is! Post your benchmarks and please keep this thread on topic. Thanks guys.
Updated with more benches. Same method. Fresh boot, 3 runs, highest score
Update 2: Linpack seems both far too inconsistant and inaccurate compared with my Droid razr. Will post results if/when it is updated.
Updated with ICEY GOODNESS!!!!
Update 4: overclocked to 1.6ghz
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
Sent from my Transformer Prime TF201 using Tapatalk
How did u get that CF bench score? What is your setup?
Stock with performance mode running. 3 runs after a reset. This thing is a beast!!! My scores ranged from 10-12 k
Sent from my Transformer Prime TF201 using Tapatalk
You know what I don't want to see? This crap! I hate you!! I want my Prime!!!!!
But really, nice post, OP. Only makes me want the damned thing more, but I like seeing what I will hopefully soon have in my hands. That thing is a beast!
Haha sorry I know how much it sucked waiting for mine but I also know how much it sucked seeing all of the negative posts as well. This is a fantastic device with a few initial manufacturing bugs. They will all get ironed out soon and we will have our "Ipad killer" if you must call it that =)
Sent from my DROID RAZR using Tapatalk
Here's mine with no cold boot or anything. Just killed recent apps and ran it.
Sent from my Transformer Prime TF201 using xda premium
I feel obligated to point out some reasons why the usual mobile benchmark suspects return somewhat underwhelming results on the prime.
Quadrant- it seems as though tegra3 has 60fps framerate cap/vsync in place. You won't see framerate in quadrant exceed ~60fps, but it won't really dip below this either, leading me to believe the uncapped framerate would be quite a bit higher.
Linpack- linpack appears to run predictably in single threaded mode, yielding results that are pretty impressive if it is indeed running on 1 1.4ghz core. I have pretty extensive experience with linpack on the server and desktop side, and it typically scales near 98% per core (assuming ample memory bandwidth), as do most pure arithmatic tests. Well, multithreaded mode doesn't seem to like tegra3 at the moment, with scores ranging between 70mflops and 120mflops....120 is probably pretty accurate, but linpack doesn't actually give it credence due to precision. This may just be a matter of updating the app, or it may be a product of the aggressive power-gating in place on tegra3.
Neocore- doesn't run, but would probably cap at 60fps.
Nenamark2- I average 50fps, but bump up against 60fps in several spots. keep in mind the resolution...
And on the topic of resolution, if you look at the results browsers for these apps, the prime may not have the highest score outright, but it definitely has the highest score of a device at 1280x800 (though that's likely to change after CES).
Just some things to keep in mind.
TF201 | XDA Premium
That's kind of why I've taken a liking to cfbench. Seems to take most of those things out of the equation.
Sent from my Transformer Prime TF201 using Tapatalk
I actually like this one lol Sorry couldnt resist what app are you guys using to run this. I figure there might be one that tests better for Tegra since my phone is a Qualcomm device
here is mines. The overall score is basically 12,000. like the op said, this score varies from the 10k-12k range though. I ran this after a fresh reboot and killing off unneccessary apps running. Although the OP scored a lil bit higher overall, my Java score is higher than his
Both cf and antutu.
Sent from my Transformer Prime TF201 using xda premium
SteveG12543 said:
Both cf and antutu.
Sent from my Transformer Prime TF201 using xda premium
Click to expand...
Click to collapse
how come your Java score is so low? did you run these tests in balanced mode? my Java score is basically 3 times higher than yours. I know running the tests, the scores will vary. but by that much?
demandarin said:
how come your Java score is so low? did you run these tests in balanced mode? my Java score is basically 3 times higher than yours. I know running the tests, the scores will vary. but by that much?
Click to expand...
Click to collapse
I have no idea . I never change the power profile from normal. I didn't kill anything right before doing the test either, and it hasn't been rebooted in about 3 days. I'll try rebooting and give it another go.
EDIT: 2764... Lame, idk why it's so low RETURN! Lol jk
Sent from my ADR6425LVW using XDA Premium.
Linpack 90-136
Sent from my Galaxy Nexus using xda premium
skeptikal said:
Here's mine with no cold boot or anything. Just killed recent apps and ran it.
Sent from my Transformer Prime TF201 using xda premium
Click to expand...
Click to collapse
I would say this is already an Ipad killer. If not I dont think my wife would be trying to trade me. She never wanted to on any other tablet.
Sent from my Transformer Prime TF201 using xda premium
Sent from my Transformer Prime TF201 using xda premium
I posted this in a different thread, but will happily move here. I'll finish up my AnTuTu results tonight or tomorrow.
I'll also comment that single-run benchmarks can be *way* off, so the OP definitely has the right idea, multiple run averages are key.
About the following - I didn't mention in the other thread that all the runs were done after manually force-stopping any app I thought might have been running. I didn't thoroughly check for everything, though.
chinly43 said:
Hi all,
As an analytical chemist, it drives me nuts to see things like benchmark results posted without details about things like how many times the test was done, what the error bars are, etc. So, to remedy the situation, here are my results. Higher is better in all circumstances. AnTuTu results will be coming as well, eventually.
Quadrant Standard (5 repetitions per setting):
Code:
Normal - average 3027.4, std dev 24.8
Balanced - average 2754.8, std dev 139.2
PowerSaver - average 2226.8, std dev 302.4
Linpack, single thread (10 repetitions per setting):
Code:
Normal - average 46.08, std dev 3.18
Balanced - average 43.24, std dev 0.44
PowerSaver - average 35.63, std dev 4.48
Linpack, multi thread (10 repetitions per setting):
Code:
Normal - average 69.63, std dev 1.90
Balanced - average 66.75, std dev 3.07
PowerSaver - average 51.90, std dev 7.10
I wonder why my linpack multithread scores were so low? Anyway, these results corroborate earlier posting about the size of the performance hit going between modes.
Click to expand...
Click to collapse
chinly43 said:
I posted this in a different thread, but will happily move here. I'll finish up my AnTuTu results tonight or tomorrow.
I'll also comment that single-run benchmarks can be *way* off, so the OP definitely has the right idea, multiple run averages are key.
About the following - I didn't mention in the other thread that all the runs were done after manually force-stopping any app I thought might have been running. I didn't thoroughly check for everything, though.
Click to expand...
Click to collapse
I basically did the same thing and rebooted before running my CF benchmark test. This info you posted in other thread was relevant to its topic also.
felt pretty special!
Sent from my Transformer Prime TF201 using xda premium
MrHennessey said:
felt pretty special!
Sent from my Transformer Prime TF201 using xda premium
Click to expand...
Click to collapse
Why such a low score? Is that just a cpu test and not a full test?
Sent from my Transformer Prime TF201 using xda premium
i was watching the phonedog.com dogfight between the Gnote and the iP 4s (no surprise the Gnote wins )
anyways, he ran the Quadrant Standard Edition so i went and installed it....his test came out to 2396. i ran the same test and mine came out to 3223. i wonder why there was such a difference? perhaps he hadn't rebooted his in awhile. i have no widgets running, perhaps that is why.
please share yours
~3300 stock, ~3800 overclocked to 1.83GHz
I don't get it. Highest ive gotten so far is like 2745. I have mine rooted with clock mod installed but on stock rom. How are you guys hitting 3,000 e on the stock rom.
Ive done all the clear RAM, reboot, & kill off backround tasks. I need more speeeeeeed! LOL>
nice jjc!
Vcolassi said:
I don't get it. Highest ive gotten so far is like 2745. I have mine rooted with clock mod installed but on stock rom. How are you guys hitting 3,000 e on the stock rom.
Ive done all the clear RAM, reboot, & kill off backround tasks. I need more speeeeeeed! LOL>
Click to expand...
Click to collapse
idk m8. hopefully someone will reply with an idea. i am interested in knowing why we all have diff #'s when we have stock rom.
do you have a widget(s) running?
Rooted, stock, frozen bloat. Fast as hell.
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
Sent from my SAMSUNG-SGH-I717 using XDA App
ARTAQaf said:
nice jjc!
idk m8. hopefully someone will reply with an idea. i am interested in knowing why we all have diff #'s when we have stock rom.
do you have a widget(s) running?
Click to expand...
Click to collapse
Nope no widgets. I'll have to try and freeze some more bloat.
Use setcpu, and use scaling performance so that it's truly running at 1.5GHz. Samsung uses "on-demand" speed scaling, so when you test, there's a chance it might be running at less than 1.5 if it hasn't scaled up to 1.5 yet
I'm running stock ROM (only rooted), CWM b4 recovery, and Da_G overclock kernel
Dont just freeze bloat uninstall it! I never freeze it.
i'm running stock everything except i run an installed Launcher. why would mine be higher than some others? does rooting slow it down?
here is my benchmark.
Using GO launcher, 6 home screens, 11 widgets - WidgetLocker Running in background w/ beautiful widgets.
SauROM w/ DaG kernel overclocked - 1.83 ghz - V6 supercharger script 1000+hp
My setCPU is 1.83 on "performance" setting (yes it makes a difference).
*Also, used RomToolboxPro to change i/o scheduler to Deadline and increase SD read write from 128kb to 3072kb...
Here is mine running Saurom with Da G kernal. Clock is at 1566mhz.
Sent from the Almighty Note!
ARTAQaf said:
i was watching the phonedog.com dogfight between the Gnote and the iP 4s (no surprise the Gnote wins )
anyways, he ran the Quadrant Standard Edition so i went and installed it....his test came out to 2396. i ran the same test and mine came out to 3223. i wonder why there was such a difference? perhaps he hadn't rebooted his in awhile. i have no widgets running, perhaps that is why.
please share yours
Click to expand...
Click to collapse
Couple things here, and by no means am I an Apple fan to ANY degree, but....
4S is close to year old technology, so comparing it to a brand new superphone like the Note (even the Skyrocket which has been to market for a couple months now) is silly. I generally like Phonedog and was genuinely surprised that guy would even compare the two devices. It sets a really silly precedent because 8 months from now, he'll likely make the same comparison between the Note and the Iphone5 which just might be similarly unfair, in the other direction.
Also, Quadrant is very nearly fraudulent in terms of its usefulness in showing the "actual" speed of a device, in a real-world sense, so I take ANY statement that starts or ends with Quadrant benchmarking with a huge truckload of salt.
Just my take...
-Ryan
borijess said:
Here is mine running Saurom with Da G kernal. Clock is at 1566mhz.
Sent from the Almighty Note!
Click to expand...
Click to collapse
Hey did you see somewhere on here, there is a guy whos getting 3999? wonder what he has.
CWM b4, Da_G's Oc Kernel, Saurom, and a lot of bloatware uninstall.
CPU set to "Performance" 1.83ghz = 3900.
DA_g kernel, OC to 1.782, on demand, no reset and no ram clear. Wanted to see what it would if i was to do the test in real world situation (like showing off to friends);-)
Sent from a Galaxy far, far away!
Lol...how many benchmark threads are we going to start....I think there are 4 or 5 now
Sent from my SGH-I717R using XDA App
Running saurom with Da G overclocked kernel clocked @ 1.782
Sent from my SGH-I717R using XDA Premium App
3930 at 1.83
Sent from my SAMSUNG-SGH-I717 using XDA
ARTAQaf said:
nice jjc!
idk m8. hopefully someone will reply with an idea. i am interested in knowing why we all have diff #'s when we have stock rom.
do you have a widget(s) running?
Click to expand...
Click to collapse
I have stock at&t note and get 3200-3300. If I don't clear the RAM, i get 2300-2700. So, just got to task manager, ram tab, and clear the memory, then use quadrant.
3203 Stock the way it comes setup.
Sent from my SAMSUNG-SGH-I717 using Tapatalk
Anyone interested? 1.8ghz was easy peasy. Is this performance gain what you may call negligible? I forgot the file name, otherwise I'd post it. Debating whether or not to do that or not, as this is the first version - plan on going higher. Plus i already work on a kernel for the tf101, might not want to take on too much. Tell me if attached quadrant score is decent. What really makes the difference is usability, I feel that its much quicker - plus it keeps my hands warm, which is great.
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
My goal is 2.0ghz, I was half joking with 'lilstevie' in the IRC chat when I said 2.0ghz was my oc demands for this new tablet - now, I'm kinda serious. Besides, this thing is still kinda slow compared to... my laptop. Also uses a new power reg.. so the max vcore limit could be higher. Yay. GPS is still a kick in the ****. That reminds me, what is the make/model of the GPS? If we can't boost reception, we can certainly try xmit power!
edit: I forgot to post the link for the actual kernel..
http://home.comcast.net/~ibladesi/woahno-it-burns.zip
blended the itsy bitsy asus changes for the big bad .21 kernel (being sardonic).
http://home.comcast.net/~ibladesi/ntrly-wrth-a-ver-bump-v2.zip
I'm going to write a script that will flash the kernel in terminal emulator later on.. thats how I flashed mine.. I'm going to include the texfat and tntfs modules and extract them to the /system/lib/modules directory.. Eventually, I will write a ROM agnostic init script that loads them upon startup.. but now its up to you to load them or write the script.
if you are using netiher exfat or ntfs, you don't need to load them..
if you need more juice for 1.8ghz..
echo 1350 > /sys/kernel/tegra_cap/core_cap_level
or if you want hand warmers..
echo 1400 > /sys/kernel/tegra_cap/core_cap_level
I think, that since this thread was moved.. it should stay here.. Only royalty can post in Development section.
A boost in clock speed could certainly speed up GPS lock down. So a Quadcore 2.0ghz would be beasty!
Nice...be careful not to burn it up!
Sent from my Transformer Prime TF201 using XDA Premium HD app
babymatteo said:
A boost in clock speed could certainly speed up GPS lock down. So a Quadcore 2.0ghz would be beasty!
Click to expand...
Click to collapse
Nah, at any speed the GPS will still be uhh.. lacking. Boosting the signal beyond regulations might help, or might hurt. Similar to a cable modem's downstream being amplified, if you amplify a noisy line - things usually get worse. In the previous post, I said xmit power, I meant rx. Or wait, is it bidirectional? I digress.
nice... good seeing you here been running your **** on my tf101 since the beginning...
Sent from my Transformer Prime TF201 using XDA Premium HD app
the GPS is omnidirectional i think
I'd say go for the overclocking.
edit: is there anyway you can make the 1.7Ghz or 1.8Ghz into an apk form to install on stock roms that's rooted? like Dr.Wowe EzOverclock app that enables 1.6ghz by just installing the app.
Also the GPS is from Broadcom, I believe. the exact name and serial # can be found in GPS teardown thread.
---------- Post added at 06:36 PM ---------- Previous post was at 06:24 PM ----------
Blades said:
Anyone interested? 1.8ghz was easy peasy. Is this performance gain what you may call negligible? I forgot the file name, otherwise I'd post it. Debating whether or not to do that or not, as this is the first version - plan on going higher. Plus i already work on a kernel for the tf101, might not want to take on too much. Tell me if attached quadrant score is decent. What really makes the difference is usability, I feel that its much quicker - plus it keeps my hands warm, which is great.
My goal is 2.0ghz, I was half joking with 'lilstevie' in the IRC chat when I said 2.0ghz was my oc demands for this new tablet - now, I'm kinda serious. Besides, this thing is still kinda slow compared to... my laptop. Also uses a new power reg.. so the max vcore limit could be higher. Yay. GPS is still a kick in the ****. That reminds me, what is the make/model of the GPS? If we can't boost reception, we can certainly try xmit power!
Click to expand...
Click to collapse
that score is decent but I would think you'd score even higher than that. my score on 1.6Ghz overclock is a little lower.
for a beefier overclock, use system tuner pro, then go into CPU tweaks. then set governor to performance. then you could set the min. clock to Max. then now you will be running all 4 cores, at the same time, at 1.8Ghz, Fulltime. I guarantee you will score alot higher than what you posted. it will drain battery faster though..lmao
Here's my score on 1.6Ghz
Blades said:
Anyone interested? 1.8ghz was easy peasy. Is this performance gain what you may call negligible? I forgot the file name, otherwise I'd post it. Debating whether or not to do that or not, as this is the first version - plan on going higher. Plus i already work on a kernel for the tf101, might not want to take on too much. Tell me if attached quadrant score is decent. What really makes the difference is usability, I feel that its much quicker - plus it keeps my hands warm, which is great.
My goal is 2.0ghz, I was half joking with 'lilstevie' in the IRC chat when I said 2.0ghz was my oc demands for this new tablet - now, I'm kinda serious. Besides, this thing is still kinda slow compared to... my laptop. Also uses a new power reg.. so the max vcore limit could be higher. Yay. GPS is still a kick in the ****. That reminds me, what is the make/model of the GPS? If we can't boost reception, we can certainly try xmit power!
Click to expand...
Click to collapse
I'm interested. Be happy to test for you.
Blades said:
Anyone interested? 1.8ghz was easy peasy. Is this performance gain what you may call negligible? I forgot the file name, otherwise I'd post it. Debating whether or not to do that or not, as this is the first version - plan on going higher. Plus i already work on a kernel for the tf101, might not want to take on too much. Tell me if attached quadrant score is decent. What really makes the difference is usability, I feel that its much quicker - plus it keeps my hands warm, which is great.
My goal is 2.0ghz, I was half joking with 'lilstevie' in the IRC chat when I said 2.0ghz was my oc demands for this new tablet - now, I'm kinda serious. Besides, this thing is still kinda slow compared to... my laptop. Also uses a new power reg.. so the max vcore limit could be higher. Yay. GPS is still a kick in the ****. That reminds me, what is the make/model of the GPS? If we can't boost reception, we can certainly try xmit power!
Click to expand...
Click to collapse
Haha there is no such thing as negligible when it comes to overclocking brotha! Take it as high as possible. I have already came up with a cooling mod I eventually want to try. I have to see how to take off the mobo though.
How is your stability? How do you increase vcore? I use set cpu stress test and at 1.7 I eventually freeze.
Sent from my Transformer Prime TF201 using Tapatalk
demandarin said:
I'd say go for the overclocking.
edit: is there anyway you can make the 1.7Ghz or 1.8Ghz into an apk form to install on stock roms that's rooted? like Dr.Wowe EzOverclock app that enables 1.6ghz by just installing the app.
Click to expand...
Click to collapse
The stock kernel allows for 1.6 that's why an apk works. For it to oc to 1.7 or higher it needs a custom kernel.
Sent from my DROID X2 using xda premium
Crank it as high as it can go. Anyone have an idea for how to liquid cool the prime? lol
Stalte said:
Crank it as high as it can go. Anyone have an idea for how to liquid cool the prime? lol
Click to expand...
Click to collapse
You could make a case that does something to this effect. Since the backplate is aluminum all you need is some thermal paste between it and the case and it would work decently well. Ideally you'd want to open the Prime and paste between the case and the processor.
demandarin said:
Here's my score on 1.6Ghz
Click to expand...
Click to collapse
why oh does my 2d score suck
Blades said:
why oh does my 2d score suck
Click to expand...
Click to collapse
wow! you did another test..lmfao this score is alot higher now.
but yeah, my 2D score is more than double yours even on your 1.8Ghz. even my 3D score was higher than yours on the one test in your OP.
is it possibly because I have a NeoGeo Emulator installed? lol They were the originators and best at 2D graphics (its a joke)
do you think it may be possibly because I have force GPU rendering enabled on 2D applications? its in developer settings in prime. I've been running my prime with this enabled & 1.6Ghz overclock. I bet if you enable it and do another test, your 2D score will be alot higher. This setup makes my prime super fast n fluid. lag free(actually was lag free before that). plus I cut off windows and transitions animation. everything near instant now.
Thank you, running stock volt at 1.8 and everything sweet for now ^.^
Enviado desde mi Transformer Prime TF201 usando Tapatalk
Boo crap poop. My prime doesnt seem to like 1.8ghz, even with the additional voltage. Well best of luck to those with the golden chips.
Sent from my Transformer Prime TF201 using Tapatalk
benefit14snake said:
Boo crap poop. My prime doesnt seem to like 1.8ghz, even with the additional voltage. Well best of luck to those with the golden chips.
Sent from my Transformer Prime TF201 using Tapatalk
Click to expand...
Click to collapse
Yup..not all chips are created equal lol
sent from my icylicious transformer prime using tapatalk
Overclocking rocks! Yes, maybe those with open primes could add a passive heatsink to the Tegra 3 CPU. Help cool her down even more. I would think that this would help with the 2.0 ghz goal. Right now its a thin copper tape over it. Add a heatsink with copper tube pasted on chip.
Glad to see you on the prime Blades, love the work you did on the TF101.
Here's my question... Is it possible to make a kernel and just use the companion core as an additional core and have it run all the time? I love having five cores but it almost seems a waste to have one shut down. I understand battery life would go down but it wouldn't bother me if it went from 3-4 days to 1-2 days.
Also, I thought I seen somewhere that manufacturer spec had the companion core at 1.4GHz and the rest at 1.3GHz. Anyway to bump the companion core up to that as well?
I don't know if any of this is possible... Just a thought.
Sent from my ICS lovin 3D!!!
If I've helped you in any way... hit the "Thanks" button.
This is my first kernel build ever (not just for the Nexus 7). Just a starting point for more hacking on the Nexus 7.
Flashable zip package attached.
It should go without saying that anything you do with this package you do at your own risk.
Features
Insecure
init.d support
Overclocked to 1.5GHz
Voltage tweaks at OC clocks
BFQ scheduler
smartassv2 governor
lagfree governor
Made kernel version not suck
Notes
You should root and install busybox into /system/xbin before flashing this kernel. Don't flash this if you don't know what you're doing as I'm too lazy to include n00b instructions.
Credits
Thanks to Virtuous team for their updater-script for this device.
Source
https://github.com/zaventh/nexus7-kernel-grouper
Blog post
http://www.jeffmixon.com/custom-nexus-7-kernel-atlantis-r2-released/
Screens
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
Cheers...
That CPU score is sexy!
Great work! Now I definitely can't wait for my nexus to ship.
Edit: If you could try adding smartassv2. This is a popular governor that I would really enjoy on the nexus 7.
Sent from my Transformer using Tapatalk 2
this is compatible with the stock rom right? ik it has to be rooted btw
Caliyork said:
That CPU score is sexy!
Click to expand...
Click to collapse
Yes it is...
MC153 said:
Great work! Now I definitely can't wait for my nexus to ship.
Edit: If you could try adding smartassv2. This is a popular governor that I would really enjoy on the nexus 7.
Sent from my Transformer using Tapatalk 2
Click to expand...
Click to collapse
Sure. I'll be adding several custom governors to choose from.
jarjar124 said:
this is compatible with the stock rom right? ik it has to be rooted btw
Click to expand...
Click to collapse
Yes, it's compiled from stock kernel sources.
Thanks so much man. Can't wait to get my nexus so I can flash this. Hopefully before the 18th which is my birthday. I had heard though that smartass didn't work so well with more than 1 core though. Could you clarify on that.
Sent from my Rezound
evilstewie23 said:
Thanks so much man. Can't wait to get my nexus so I can flash this. Hopefully before the 18th which is my birthday. I had heard though that smartass didn't work so well with more than 1 core though. Could you clarify on that.
Sent from my Rezound
Click to expand...
Click to collapse
Smartass v1 was made for single core cpu's. Smartass v2 was made for dual core cpu's, so v2 might work better on a quad core? V2 doesn't really work all that well on my single core cpu for the phone I have.
zaventh said:
Yes it is...
Sure. I'll be adding several custom governors to choose from.
Yes, it's compiled from stock kernel sources.
Click to expand...
Click to collapse
Nice work! Can't wait to see the future of this kernel.
Sent from my Transformer using Tapatalk 2
sparksco said:
Smartass v1 was made for single core cpu's. Smartass v2 was made for dual core cpu's, so v2 might work better on a quad core? V2 doesn't really work all that well on my single core cpu for the phone I have.
Click to expand...
Click to collapse
Exactly right. I actually never really liked smartassv2 in general, but I think it is working out fairly well on the N7, actually.
This is after just a little bit of game time, some regular usage, and some sleep time. I may need to tweak some of the higher clock values so it scales up a bit more gracefully. At any rate, there will be plenty of options to play around with.
zaventh said:
Exactly right. I actually never really liked smartassv2 in general, but I think it is working out fairly well on the N7, actually.
This is after just a little bit of game time, some regular usage, and some sleep time. I may need to tweak some of the higher clock values so it scales up a bit more gracefully. At any rate, there will be plenty of options to play around with.
Click to expand...
Click to collapse
So whats the heat production and stability like at 1.5ghz? When you say 1.5ghz, is that like the htc one x where it is actually 1.4ghz but you can go to 1.5ghz when only using one core, or is it all four cores clocked at 1.5 please? Got to admit that a gpu OC would be really nice. This looks great, thanks
any chance of getting some gpu over clocks in then kernal, screw battery life give me mmmhhhzzz!
Is it possible to get scary gov on this? I think it's based on smartassv1 but not really sure. It gives me a good balance between performace and battery. Or perhaps we could make a scaryV2 lol?
Great work for a first Kernel
Is the extra battery drain worth that .2ghz of extra horsepower? Because i've heard it's a perfectly capable device @ 1.3ghz.
lhayati said:
Great work for a first Kernel
Is the extra battery drain worth that .2ghz of extra horsepower? Because i've heard it's a perfectly capable device @ 1.3ghz.
Click to expand...
Click to collapse
Thanks. Naturally, it's a matter of personal preference of course if it is "worth" it to OC any processor. OC'ing the N7 was more of a "because I can" thing than for any practical reason. The CPU is impressive as is.
The N7 is probably the fastest and most responsive mobile device I've ever seen even at stock, mostly thanks to the hardware but also due to Jelly Bean's ability to leverage this kind of raw power.
The actual battery drain between 1.3GHz -> 1.5GHz will all depend on how much time you spend at those higher clock speeds, of course.
zaventh said:
Thanks. Naturally, it's a matter of personal preference of course if it is "worth" it to OC any processor. OC'ing the N7 was more of a "because I can" thing than for any practical reason. The CPU is impressive as is.
The N7 is probably the fastest and most responsive mobile device I've ever seen even at stock, mostly thanks to the hardware but also due to Jelly Bean's ability to leverage this kind of raw power.
The actual battery drain between 1.3GHz -> 1.5GHz will all depend on how much time you spend at those higher clock speeds, of course.
Click to expand...
Click to collapse
Thank you for this kernel. I cant wait for my nexus 7
Sent From A Rotary Phone
Do we know at what speed the ram is clocked?
The transformer infinity uses DDR3 at 1600. The 300t is much lower though
Sent from my GT-I9100 using xda app-developers app
emergant said:
Do we know at what speed the ram is clocked?
The transformer infinity uses DDR3 at 1600. The 300t is much lower though
Sent from my GT-I9100 using xda app-developers app
Click to expand...
Click to collapse
It's been confimred in the Q&A section that the N7 has DDR3 ram :beer:
Thanks. But DDR3-667 as in tf300 or DDR-1600 as in Tf700 or something else
I thought about putting this in QA but thought the dev might know. Sorry if that was a bad call
Sent from my GT-I9100 using xda app-developers app
Whops I meant general not Q&A
Hynix HTC2G83CFR DDR3 RAM
Hope that helps
http://forum.xda-developers.com/showthread.php?t=1748933
Smartass V2 doesn't support scaling upto 4 cores
Its useless for this device!