Ok so I know they share processors and are quite similar so will our phones ever be capable of this...
http://www.youtube.com/watch?feature=youtube_gdata_player&v=GpLju9341v8
I know quadrant is flawed towards some devices, I'm not saying anything about that, I have my shift at 1.8 and I'm only pulling 2100...
Sent from my PG06100 using XDA App
On Evervolv (AOSP GB) I can pull 2700 at 1.8.
EDIT: Just pulled 2823. Have faith, the Shift will dominate
^ Shift Faced
Giggity
Sent from my PG06100 using XDA App
What SetCPU settings are you guys running? I've tried multiple governors and freq combos (I haven't locked or rebooted once with max 1.8) but can't even break 1900 in quadrant! I'm using the latest free version of quadrant available on the market if that matters.
I'm jealous of everyone else's reported scores =( My FPS also don't seem to have improved much over stock in the graphics benchs.
Fps won't change much
Sent from my PG06100 using XDA App
nice nice nice
Related
What is the highest mflops score seen on an epic so far?? I'm getting 10.462 on linpack app running bs 1.3 overclock. Any better? If so what Kernal Rom and MHz u run?
Just a snide comment: "my mflop is bigger than your mflop" - that's the future of this thread lol.
Sent from my SPH-D700 using XDA App
It's not the # of the mflops its the way they're used
Sent from my SPH-D700 using XDA App
A 15.something using bs 1.3 and noob's JIT beta. A bit unstable but still fun.
Sent from my Epic 4G
What's the average score for a evo on cm7 .
Sent from my PC36100 using XDA App
parmend said:
What's the average score for a evo on cm7 .
Sent from my PC36100 using XDA App
Click to expand...
Click to collapse
Just tested at 41.8
Yeah, I would say avg is anywhere from 35 to 42, depending on clock speed. What happens if you run it again right after a couple times?
One time I got and 8, and I was like, WTF? Ran it again and got a 38.
Thanks, I just ran it and it still 16-18. Does the kernal matter. I'm on timat 3.3.7
Sent from my PC36100 using XDA App
yeah on cm7 I got 40 give or take.
stock sense is like 32 for me and overclocked to 1.2ghz i get 38-39.
bfs kernels were hit or mis on my EVO. Kinda clunky on mine.
I'f you're getting low linpack scores try switching from smartass to conservative(just to score higher). Also I get an average of 41 with my phone clocked at 1190.
Sent from my PC36100 using XDA Premium App
Specs in sig, I just got 43.261, average is 42-44.
How do you check?
Arnezie said:
How do you check?
Click to expand...
Click to collapse
Download linpack from the market
mattykinsx said:
Download linpack from the market
Click to expand...
Click to collapse
Thanks
32.059
Stock rooted rom
25.59 Destroyed v2 how do i improve that?
39.6
Mik 4.61
Netarchty-toastmod 4.3.1
Setcpu CPU 1190MHZ Very Stable
Quadrant vs linpack? Which one is more important?
Sent from my PC36100 using XDA App
parmend said:
Quadrant vs linpack? Which one is more important?
Sent from my PC36100 using XDA App
Click to expand...
Click to collapse
I would say definitely linpack. Quadrant is not consistent.
you can get bad scores if you have to many services running. i get 33-35 stock non rooted hw 004. got up to 44+ on my old hw 002(got the restart when overheated problem) when 2.2 first came out, completely stock but since then they have added alot of services that slow it down
if the media scanner is running. i get about 22. so check that in the services if you are getting bad scores
Fricksta said:
25.59 Destroyed v2 how do i improve that?
Click to expand...
Click to collapse
Need more details. What kernel, what CPU settings, etc. I was getting 38-40 on d2.
Just got 40.75 on cm7 with dark horse theme and sz 2.0.0 ondemand 1190
The Linpack app also does not pull enough from CPU to ramp the CPU to max. Set your minimum CPU to the same as your max, run linpack then lower your minimum CPU speed, Interactive is best governor to use for this
all these tests are for is braggin rights...these scores will not be like, oh check it out, my browser loads .3 seconds quicker bc i got 3 more mflops...get out of hee, find a rom you like, and use it, most of them are very good. There is not one rom on this site that blows away the competetion, they are all very very very similiar in speed, nuff said
Anyone check the highest linpack score for the evo shift lately? Dodgejcr seems to have 1920mhz stable enough for testing linpack
Sent from my PG06100 using XDA App
DWake014 said:
Anyone check the highest linpack score for the evo shift lately? Dodgejcr seems to have 1920mhz stable enough for testing linpack
Sent from my PG06100 using XDA App
Click to expand...
Click to collapse
I'd like to see him get that stable on our device.
I had a 64.xxx run while on sense that I uploaded under PG06100 instead of htc evo shift I ran her again on CM7 and am in second place with 63.866
Sent From My Speedy Mc Speederson
Yeah or at least release a kernel for testing. Nice score, mine keeps capping out at 63.752 on nfx 2.0 odex
I just hit 63.946
Sent from my PG06100 using XDA Premium App
So is there a kernal for going past 1.8gHz?
It looks like it may be developer only, I'm thinking it may be very unstable
dsexton702 said:
These are some tweaks created by dsexton702 that offer you a generous performance boost in I/O and giving you the newest fugu mod 3.1. This is a fast growing tweak and needs to be shared with everyone who can benefit from it
download the U.S.A.Sv6
OP: http://forum.xda-developers.com/showthread.php?t=1236779
Click to expand...
Click to collapse
It does work on our phones as I'm using right know...just thought I pass on the good work of these devs...
Sent from my PG06100 using xda premium
What is it like V6SuperCharger?
Sent from my PG06100 using XDA App
It speeds up and opyimuzes androids Input and output....its just version6 of the mod.....thy have a new version but I haven't tried yet..I've used this on my HTC hd2 and mytouch 4g and know my shift .....
Sent from my PG06100 using xda premium
Idk if I posted in right section if not can mod move this please to right one....
Sent from my PG06100 using xda premium
v6 supercharger is useless IMO. Tweak your data setting in ##DATA# with a sense rom and use a good kernel like super scary stable.
VICODAN said:
v6 supercharger is useless IMO. Tweak your data setting in ##DATA# with a sense rom and use a good kernel like super scary stable.
Click to expand...
Click to collapse
You see I did tweak my ##DATA# setting and use Scarykernel on the latest CM7 nightlies. There's a noticable improvement using V6Supercharger over without on my device.
Sent from my PG06100 using XDA App
Same here I have found a noticable speed also in quadrent score also....
Sent from my PG06100 using xda premium
bigfdaddy said:
Same here I have found a noticable speed also in quadrent score also....
Sent from my PG06100 using xda premium
Click to expand...
Click to collapse
What's your antutu score with Scary kernel on toasty at 1900mhz?
Here the score Wich includes v6 mod,scary govener,1900mhz over clock,and SD booster...
Sent from my PG06100 using xda premium
bigfdaddy said:
Here the score Wich includes v6 mod,scary govener,1900mhz over clock,and SD booster...
Sent from my PG06100 using xda premium
Click to expand...
Click to collapse
Proof that this **** is useless:
Scarystable compcache disabled:
RAM 612
CPU integer 1221
CPU float-point 518
2d graphics 276
3d graphics 793
database io 535
sdcard write 9.9MB/s 99
sdcard read 17.5 MB/s 175
total score: 4229
Linpack single thread score = 62.33 MFLOPS, Time 1.35 Seconds, Norm Res 5.68, Precision 2.22044604...
Linpack multi thread score = 57.961 MFLOPS, Time 2.91 seconds Norm Res, 3.24, Precision 2.22044604...
P.S. You should use Toasty gov for benchmarking and run benchmarks on a fresh boot and close as many apps as possible.
BTW:
http://forum.xda-developers.com/showthread.php?t=1296614
---------- Post added at 01:51 AM ---------- Previous post was at 12:54 AM ----------
Benchmark scores do not prove actual real life performance...
They are just numbers ppl like to boast about and doesn't mean your phone is any faster/more responsive with a higher score.
And. Any scores wether antutu/quadrant/or whatever can be faked.
Sent from my Supreme Shift on Wires!
notsointeresting said:
Benchmark scores do not prove actual real life performance...
They are just numbers ppl like to boast about and doesn't mean your phone is any faster/more responsive with a higher score.
And. Any scores wether antutu/quadrant/or whatever can be faked.
Sent from my Supreme Shift on Wires!
Click to expand...
Click to collapse
I wouldn't be wasting my breath and posting fake benchmark scores if I didn't believe it and try this mod.
The only difference I noticed was LPP loading a little bit faster on startup.
VICODAN said:
I wouldn't be wasting my breath and posting fake benchmark scores if I didn't believe it and try this mod.
The only difference I noticed was LPP loading a little bit faster on startup.
Click to expand...
Click to collapse
Well every little bit counts doesn't it?.....ty for trying it tho and posting ya score....
Sent from my PG06100 using xda premium
bigfdaddy said:
Well every little bit counts doesn't it?.....ty for trying it tho and posting ya score....
Sent from my PG06100 using xda premium
Click to expand...
Click to collapse
Np.. just saying everytime I flash a nightly I have to do this.. it's annoying as hell where as I just flash ScaryStable once with each nightly flash and I'm good to go and I notice a HUGE difference in performance.
For what it's worth, the V6 Supercharger (currently running the update9 beta) provides an optimal RAM manager which eliminated my habit of installing AutoKiller w/ every ROM. Since running the Supercharger script, I've consistently ran w/ 175-180MB in free RAM w/ no app FC's. That alone is worth using this mod in my book.
Ty 4 trying.....tying to expand this to more people and let them know about it.I.think its worth all the hype and attention I can give it.
Sent from my PG06100 using xda premium
How come the cpu seems so slow in quadrant standard tests.... Heck, my desire z (800mhz single core) blew through the cpu component of the test faster...This result is typical of all Cm7 roms I've tried. Below my result is... Surprise! Atrix 4g, with superior cpu rating, but inferior i/o. I find this all kind of counter intuitive...
Any insights?
Sent from my MB860 using xda app-developers app
=JKT= said:
How come the cpu seems so slow in quadrant standard tests.... Heck, my desire z (800mhz single core) blew through the cpu component of the test faster...This result is typical of all Cm7 roms I've tried. Below my result is... Surprise! Atrix 4g, with superior cpu rating, but inferior i/o. I find this all kind of counter intuitive...
Any insights?
Sent from my MB860 using xda app-developers app
Click to expand...
Click to collapse
um... your I/O bar is the biggest.
HMZX said:
um... your I/O bar is the biggest.
Click to expand...
Click to collapse
I think he meant inferior CPU and superior i/o.
HMZX said:
um... your I/O bar is the biggest.
Click to expand...
Click to collapse
=JKT= said:
How come the cpu seems so slow in quadrant standard tests.... Heck, my desire z (800mhz single core) blew through the cpu component of the test faster...This result is typical of all Cm7 roms I've tried. Below my result is... Surprise! Atrix 4g, with superior cpu rating, but inferior i/o. I find this all kind of counter intuitive...
Any insights?
Sent from my MB860 using xda app-developers app
Click to expand...
Click to collapse
have you underclocked?
i got double your cpu score on 1.45ghz
Its because cm7 doesn't initialize both cousin in quadrant. Try it on cm9 or cm10 CPU doubles.
Managed to get 4.5k quadrant score at 1.45 on the older builds. 4K with 1.3ghz
Sent from Moto Atrix 4g on Neutrino 2.91
I guess I wasn't clear enough lol
My phone was at stock 1ghz. This was to compare objectively to the bar below the "your device" bar, which was entitled "atrix 4g" which had, yes, superior cpu and inferior i/o scores.
Of course you will get better scores overclocked to 1.45 ghz. Try to understand what I'm asking about. I'll ask more clearly.
Why does a custom rom appear to get a worse cpu score than stock at same clock rate? Shouldn't it be equal or greater?
squee666 said:
Its because cm7 doesn't initialize both cousin in quadrant. Try it on cm9 or cm10 CPU doubles.
Click to expand...
Click to collapse
Is that true? System panel gives independent readouts of the load on each core and they seem to both be working. But if they don't both work in quadrant that would explain it all =) (except why quadrant lol)
Sent from my MB860 using xda app-developers app
Score I got at 1.3ghz faux kernel on am older version of cm9. Yea quadrant reads that there are 2 cores. Just does not use them properly.
Sent from Moto Atrix 4g on Neutrino 2.91
Thanks guys. That does seem to be the issue. Here is a score @ 1gHz with mokee os. Looks pretty obvious that both cores are working here:
Sent from my MB860 using xda app-developers app
Lol its not obvious at all the green bar is I/o mokee is cm7 based you should try a cm9 or cm10 based rom to see CPU difference.
Sent from Moto Atrix 4g on Neutrino 2.91
squee666 said:
Lol its not obvious at all the green bar is I/o mokee is cm7 based you should try a cm9 or cm10 based rom to see CPU difference.
Sent from Moto Atrix 4g on Neutrino 2.91
Click to expand...
Click to collapse
Look at my blue (ie cpu) bar in this image compared to the image in my first post. I believe it is, in fact, quite obvious. Lol... At least that something is significantly improved cpu-wise...
Sent from my MB860 using xda app-developers app