GoBruins
Well-Known Member
Cool thanks Saps. I'm going to stick with the 1ghz Voodoo then. I'm going for stability at the moment... Battery efficientness is my next goal...
-
After 15+ years, we've made a big change: Android Forums is now Early Bird Club. Learn more here.
Root Dirrk's D01 overclocking kernels with lagfix
- Thread starter jamesnmandy
- Start date
-
- Tags
- samsung fascinate
Cool thanks Saps. I'm going to stick with the 1ghz Voodoo then. I'm going for stability at the moment... Battery efficientness is my next goal...
saps
Extreme Android User
Is that jt's or Dirrk's? FWIW, Dirrks seems to score ever so slightly higher overall in the benchies and he confimed its running 1225mV. I dont know jt's voltage but between the two based on my initial impressions I'd go with Dirrk's. Not many of us awake at this hour huh, West Coast midnight owls
saps
Extreme Android User
Well Dirrks 1000lv also checks out fine and dandy. Now Im spending only about 30-40 minutes with each kernel. I go thru several back to back Quadrant runs, I do Speed PI and Benchmark Pi about 20 times each, I do the short bench, long bench, native bench and stress test in SetCPU. I let the Stress Test run for a few minutes. NeoCore, An3DBench, Linpack. Then I play some converted DVD's and jump on WiFi and watch some you tube stuff. So sure if I conclude its stable after 30 minutes of testing that does not mean some might not have problems later.
Admittedly, I was pretty excited to test this one. Since this one runs on just 1225mV compared to the stock 1300mV at the same 1Ghz this should give you/me/us quite a boost in terms of battery life. In fact, here's my observation. With jt's 1.2kernal and 1325mV I am going thru about 1% battery life every 2 minutes when benching it. With the 1Ghz 1225mV kernel its more like 4-5 minutes per 1% when benching. Not the most scientific and I will stop well short of calling this double the battery life but it seems noteworthy for now.
Admittedly, I was pretty excited to test this one. Since this one runs on just 1225mV compared to the stock 1300mV at the same 1Ghz this should give you/me/us quite a boost in terms of battery life. In fact, here's my observation. With jt's 1.2kernal and 1325mV I am going thru about 1% battery life every 2 minutes when benching it. With the 1Ghz 1225mV kernel its more like 4-5 minutes per 1% when benching. Not the most scientific and I will stop well short of calling this double the battery life but it seems noteworthy for now.
saps
Extreme Android User
OK, Dirrk's 1100LV is stable, I could not get it to crash, however, the performance numbers seem nearly identical to his 1000LV, so Im kinda scratching my head on that one. Like others before me have noted it seems for whatever reason something is amiss. For instance on Linpack, I get 8.1-8.2 on 1Ghz and 9.9 with 1.2Ghz. This should give me 8.9-ish for 1.1Ghz but nope never saw better than 8.2 aka same as the 1Ghz. Similar with Quadrant, never cracked 1700 with either the Ghz or the 1.1. So again good news bad news. Good news is the stability which is most critical for most of us. Bad news is same performance as 1Ghz. Honestly between Dirrks 1000LV and 1100LV Im not sure which I'd pick since they bench nearly identical. On to 1200LV
saps
Extreme Android User
Well boys and girls I think we have a winner. Dirrk's 1200LV.
Linpack was 9.826 so thats consistent with what we should be seeing not sure what happened at 1100. Quadrant broke 2K a few times. The short bench in SetCPU was 161ms, ran the stress test for 4 minutes. I almost have no reason to test his SV kernels now. Maybe that 1100SV will be better than the 1100LV. But I'm going stick my neck out an inch and say Dirrk's 1200LV is they kernel to have of the 5 I've tested so far [jt's 1000 & 1200; Dirrk's 1000LV, 1100LV, 1200LV]. I will admit its a fraction behind jt's 1200 but I could not complete Quadrant with jt's and I did got random lockups. I did 6 back to back runs with Dirrk's 1200LV. And again the numbers are close. 9.82 vs 9.90 in LinPack. In Benchmark Pi jt gave me 2114ms Dirrk was 2166ms. I mean its basically a wash. But the 75mV less makes it a done deal. This should give even better battery life than stock. This was the kernal profile I always wanted ever since I knew the Epic guys had a 1.2Ghz@1250mV and they rave about it. Dirrk's getting some paypal luvin from me. Im not kidding this is the kernel.
So I hope no one minds but Im skipping the 1100SV I see nw point in even trying it. 1200LV is more speed and less voltage. 1200SV might bench a tad higher than 1200LV I likely will test that one tomorrow
Well its getting late here 130AM and I gotta be at work in 6 hours so I hope this info helps.
Man I really am excited about this one. Especially since I hear jt is opting out of higher clocked kernels. But Dirrk already had a 1.4 and even a 1.6 he was trying out. So from 1.2Ghz@1250mV I have to assume 1.4Ghz is achievable on 1300-1325mV and we've already learned 1325mV is a tolerable voltage.
Linpack was 9.826 so thats consistent with what we should be seeing not sure what happened at 1100. Quadrant broke 2K a few times. The short bench in SetCPU was 161ms, ran the stress test for 4 minutes. I almost have no reason to test his SV kernels now. Maybe that 1100SV will be better than the 1100LV. But I'm going stick my neck out an inch and say Dirrk's 1200LV is they kernel to have of the 5 I've tested so far [jt's 1000 & 1200; Dirrk's 1000LV, 1100LV, 1200LV]. I will admit its a fraction behind jt's 1200 but I could not complete Quadrant with jt's and I did got random lockups. I did 6 back to back runs with Dirrk's 1200LV. And again the numbers are close. 9.82 vs 9.90 in LinPack. In Benchmark Pi jt gave me 2114ms Dirrk was 2166ms. I mean its basically a wash. But the 75mV less makes it a done deal. This should give even better battery life than stock. This was the kernal profile I always wanted ever since I knew the Epic guys had a 1.2Ghz@1250mV and they rave about it. Dirrk's getting some paypal luvin from me. Im not kidding this is the kernel.
So I hope no one minds but Im skipping the 1100SV I see nw point in even trying it. 1200LV is more speed and less voltage. 1200SV might bench a tad higher than 1200LV I likely will test that one tomorrow
Well its getting late here 130AM and I gotta be at work in 6 hours so I hope this info helps.
Man I really am excited about this one. Especially since I hear jt is opting out of higher clocked kernels. But Dirrk already had a 1.4 and even a 1.6 he was trying out. So from 1.2Ghz@1250mV I have to assume 1.4Ghz is achievable on 1300-1325mV and we've already learned 1325mV is a tolerable voltage.
saps
Extreme Android User
No. Since they both use the same EXT4 filesystem [voodoo v.2] I went from one of jt's to another and then right to Dirrk's from a jt and then of course from various Dirrk's to the other.
Bottom Line you can just flash them one on top of the other jt or Dirrk it doesnt matter as long as the kernels continue to be based on the voodoo v.2
Let me again say what a treat Dirrk's 1200LV is I said I was going to sleep but I just can't. I do miss using JT's because of the boot screen. I need to figure out how to extract it out of jt's zip and add it to Dirrk's.
saps
Extreme Android User
There might be a bit of difference sure. Its another 25mV less. 1Ghz [and 1.1] run on 1225mV whereas 1.2Ghz runs on 1250mV. Again, remember the phone out of the box runs 1Ghz on 1300mV so either is a decrease. I will have to give it a full days use to accurately gauge how durable this battery is now with this kernel. So while 1Ghz @ 1225mV will be a big step up on battery life from stock so is the 1.2@1250mV. Again that 1.2 is the best combo of battery and performance and I cannot recommend it enough.
Ok seriously though I gotta get some sleep now
D13
Android Expert
open up jt's zip and just drag and drop the bootanimation.zip to your desktop. It is located in /system/media in jt's zip. The put in /data/local
jamesnmandy
Android Enthusiast
damnit....is there a way to take a complete image of our phones that would be inclusive of rom, kernel, filesystem, recovery, sdcard......the actual entire thing, ghosting the entire phone, so that i can try again and not have to deal with Odin and uninstalling and all that crap?
saps
Extreme Android User
FWIW, those numbers specifically seem a tad low. Now who cares right? Well if you do and by all means you dont have to if you phone is now snappier and with better battery life win win. But just for bragging sake, most benchmarks will score slightly higher when you set the SetCPU to performance and run the benchmark a few times especially LinPack and Speed Pi. Now how much a big deal is it, not much. In LinPack I peak out between 9.7-9.8 on Conservative and 9.8-9.9 on performance. Also my first LinPack run is usually between 9.5-9.7 but climbs a bit with subsequent tests. So I guess Im saying 9.0-9.1 seems a tad low even if it was your 1 and only 1st run thru. Same with Speed Pi. Run that sucker 10 times at least you really should be much closer to 500ms
I mean Im probably splitting hairs. End of the day what does it matter right?
The kernel is so stable and battery life seems stellar. I pulled off charger at 730am and now its 900am and Im at 98% and thats with 6 texts sent and received and 2 emails sent and received. A bit of Speed Pi and Linpack just so I confirmed again what I told you above. I mean we cannot judge another on 1.5hours Im sure this thing is not going to average battery life, that would be silly 32% per 24 hours. Im sure with heavier usage it drops but still Im impressed so far.
saps
Extreme Android User
No, not that Im aware of. I feel terrible for you buddy because I know where you are right now frustrated with the process. I was there too in the wee hours of Sunday morning with my SF, virtually bricked but somehow I figured it out. Thats why I imposed my own wait period before trying out new stuff.
I also know you said you'd stay where you are at for awhile and I certainly respect that decision. However, the good news is if [big if I know] if you changed your mind these kernels are the real deal. You can flash back and forth between any of jt's and dirrk's right now. I know I tried out six different ones last night. That D012 1200LV is one I know you didnt try yet and thats what Im on now and man is it solid and fast.
Anyway Im not trying to tempt you back into the OC/voodoo club but just know if you came a knockin' I think you'd have a trouble return trip.
saps
Extreme Android User
Well the adventures are over the moment. I cannot imagine being moved off of the 1200LV. The 1200SV might test 1-2% higher in benchmarks. But Im anticipating 20% better battery life with the LV kernel and thats more than a fair trade for me. Im not giving up Dirrks 1200LV for anything less than a 1400LVThank you saps I'm following your adventures closely
saps
Extreme Android User
Told yaAWESOME. Put this through some preliminary testing and its really fast. Also scoring in the 2000 Quad range and 9.7 on initial MFLOPS, without locking it on performance. Hopefully it remains stable so I can stick with it for a while
Awesome indeed nice numbers, basically right where mine are. It should be stable for you too. I tried to crash mine last night. Set it to Performance in SetCPU and it would not crack. I love it. Seriously if I can get a paypal donation link for Dirrk I'm giving him some $$ he's more than earned it
nitsuj17
Android Expert
wow, with my fascinate coming this week and all the good stuff going on I'm pretty pumped to dive right in! Love being able to have stuff to flash, root etc from the get-go after a lot of waiting around when I first got my DX.
When I got my inc it was an insane few days of trying everything out, but with the awesome potential of the sf, I'm much more excited than I was for my inc. Hopefully (and I'm sure it will) it'll live to my excitement/hype. I'm sure it ll make me glad I jumped ship from the x like A LOT of other irratated former X-users.
When I got my inc it was an insane few days of trying everything out, but with the awesome potential of the sf, I'm much more excited than I was for my inc. Hopefully (and I'm sure it will) it'll live to my excitement/hype. I'm sure it ll make me glad I jumped ship from the x like A LOT of other irratated former X-users.
saps
Extreme Android User
Conservative. That and Performance are the only two that are considered safe and reliable thus far. Don't start messing with profiles either. If you wanna bench it and get maximum numbers throw it into Performance, this also is the best way to stress test it. For everyday use of course just leave it on Conservative. Battery Life is sick on this thing. I've lost 6% in 5 hours
D13
Android Expert
Conservative. That and Performance are the only two that are considered safe and reliable thus far. Don't start messing with profiles either. If you wanna bench it and get maximum numbers throw it into Performance, this also is the best way to stress test it. For everyday use of course just leave it on Conservative. Battery Life is sick on this thing. I've lost 6% in 5 hours
wow, I want this...mine is now coming on thursday cause yesterday was Columbus day, so it got shipped today