[KERNEL] 2.6.29, ~1600MHz, >1000 Quadrant

Search This thread

android53

Senior Member
Jul 23, 2010
721
18
Things to look out for:

High fps readings yet things look visually slow or laggy
Weak vibration occasionally
Animations out of sync or running slow

If these are present then this is a port (or chance recreation) of team whiskey's kernel.

Do some testing people! If you don't get these problems then it could be big news, I see one difference atleast, this lacks the large memory score inflation

Sent from my GT-I9000 using XDA App
 

KellyLewis3985

Senior Member
Jul 28, 2010
487
46
Lake Charles, LA
LoL this guy. All you did is copy someone elses work, which is not always bad but you could at least give credit where credit has been earned. Play a 3d game the graphics are slow and choppy. Manhattan project was shelved due to the internal clock not being able to keep up with the processor. Which in the long run will hurt your phone.

Sent from my SGH-T959 using XDA App
 

adrynalyne

Inactive Recognized Developer
Dec 13, 2008
10,950
6,471
LoL this guy. All you did is copy someone elses work, which is not always bad but you could at least give credit where credit has been earned. Play a 3d game the graphics are slow and choppy. Manhattan project was shelved due to the internal clock not being able to keep up with the processor. Which in the long run will hurt your phone.

Sent from my SGH-T959 using XDA App

LOL back at you. He has been working on this in Irc for quite a while with a lot of trial and error. Prove that work was copied or shut up. Do you really think that team whiskey is some special dev team that does things nobody else can? That's ludicrous at best.

Sent from my SCH-I500 using XDA App
 

android53

Senior Member
Jul 23, 2010
721
18
Need more information to determine if this is just a port, don't jump the gun people

Sent from my GT-I9000 using XDA App
 

TheTodFather

Senior Member
Aug 9, 2010
692
57
Huntsville
LoL this guy. All you did is copy someone elses work, which is not always bad but you could at least give credit where credit has been earned. Play a 3d game the graphics are slow and choppy. Manhattan project was shelved due to the internal clock not being able to keep up with the processor. Which in the long run will hurt your phone.

Sent from my SGH-T959 using XDA App

Dude, your such a troll. Nice picture by the way. Sorry stud muffin but this has nothing to do with team whiskey or should i say morphics code. Before you come in yelling try doing a little research their bub! Just cause a 3d game lags has nothing to do with morphic kernel. Probably cause you running something in the background like jersey shore. LOL j/k Its all good . Chill man...
 

TheTodFather

Senior Member
Aug 9, 2010
692
57
Huntsville
Man those dudes are such douche bags... always jumpin the gun!
 

Attachments

  • 300-jersey-shore-guys.jpg
    300-jersey-shore-guys.jpg
    26.3 KB · Views: 145

Kubernetes

Senior Member
Aug 29, 2010
846
26
LoL this guy. All you did is copy someone elses work, which is not always bad but you could at least give credit where credit has been earned.

You have no idea if that's true or not. The OP could easily have developed this independently. What's important is whether this has the same flaws as Manhattan, because if it doesn't, that's good news for the entire Galaxy S community.
 

KellyLewis3985

Senior Member
Jul 28, 2010
487
46
Lake Charles, LA
Reguardless of what you may think or say, the op knows what I'm saying is true. Show respect to the creator thats all I'm saying.

Sent from my SGH-T959 using XDA App
 

saps

Senior Member
Sep 17, 2010
444
30
1600Mhz is real. This is from wickedbeav's test2 kernel over on the IRC:
snap20101101_175354.png
snap20101101_172124.png

snap20101101_170709.png
snap20101101_170201.png

Battery life seems pretty decent too. I benched it for an hour and used like 15% battery not bad.
 

Diversion

Senior Member
Mar 2, 2010
343
27
I'm seeing NO slowdown in gaming - the opposite.. Angry Birds is silky smooth... much smoother than stock kernal. However, I too am seeing some funkiness in the App drawer.. sometimes it's normal.. sometimes it's jerky.. nothing major tho.

Haven't had a FC yet either.. Although my particular Fascinate seems to overclock a lot better than my fiance's Fascinate.. She'll get FCs at 1200 or more.. but fine at 1100 and under.. whereas I was previously at 1250 and now this 1600 with no issues.
 

krylon360

Retired Recognized Developer
May 10, 2010
2,228
3,109
Seattle
Hey guys. Just thought I would swing by. This thread was tweeted to me.
I just talked with morf, to see if it was his code.
Though it is not, it is very similar, and he wanted to let you guys know, that the same mistake was made. Your phones will read 1.6 and will bench like it's 1.6, but the system will slow down.
 

Kubernetes

Senior Member
Aug 29, 2010
846
26
Can you expand upon that statement / define slow down?

Here's how Viralblack explained it in the Project Manhattan thread:

Imagine if you were running at 10MPH, and while you were running time slowed down but the slowdown didnt effect you. So youre still running at 10MPH but to the rest of the slowed down world, now youre running at 15MPH, even though your speed and direction havent changed.

The changes in the code for one reason or another arent playing nice with the Android system. Instead of pushing the clock up, in most cases, what its doing is simply slowing the internal clock down to the tune of about 30%. This is what is actually providing the benchmark gains that were being recorded, not an actual gain.
 

morfic

Inactive Recognized Developer
Aug 3, 2008
7,211
12,879
San Antonio
www.derkernel.com
Code:
/*pms value table*/
/*APLL(m, p, s), MPLL(m, p, s)*/
static const u32 s5p_sys_clk_mps_tab_1GHZ[][6] = {
        {400, 8, 1, 667, 12, 1},
//        {250, 6, 1, 667, 12, 1},
        {200, 6, 1, 667, 12, 1},
        {200, 6, 1, 667, 12, 1},
        {200, 6, 1, 667, 12, 1},
        {200, 6, 1, 667, 12, 1},
};

I hate to break it to you, you morficed your kernel.
you are running a stable 1.2GHz, if you have time drift or not i didn't even look, but the formula for Fout is MDIV * FIN / (PDIV * 2^1-SDIV)

if MDIV of 250 and PDIV of 6 gives is 1GHz, we can verify that FIN, the input frequency, is 24MHz.

So 24*400/(8*2^(1-1)) gives us 24*400/8 (2^0 is 1, so we do not change PDIV)
24/8 is 3 , 3*400 is 1200

So you like me, missed how they calculate FOUT, i got it finally right when unhelpful linked me to the samsung pdf on xda.

I can run 1.4GHZ (true 1.4GHz if i turn on the phone in the fridge) and a 1.2GHz that scores quadrants around 1050ish, (more than i hear King Klicks 1.3ghz does...)
The 1.6GHz "Manhattan Kernel" had all numbers ad up to a theoretical 1300 quadrant score, 12.x in linpack, so you did not do as bad as me :) but it is all in all only 1.2GHz.

What we need to see if if phones like 266, 4, 1 better than 400, 6, 1
I can run stuff stable at 1.2 or 1.3GHz that do not work at all on other phones.

Also interesting that they give us a G3D divider while they also show in the same document that 3D is in the MSYS domain, while 2D is in the DSYS domain.

Either way, i think there is room to tinker to get more out of the phones, but we don't have it yet. Also a phone with better linpack scores that does not better in quadrant should be of limited use, as our phones are smooth as silk at 1.0 and 1.2 GHz and only games would benefit from more performance......or if google gave us a 3D desktop...YUM! :)

Hope this info helps. I only read the OP after being linked to it and figured if someone had pointed out the 1.2GHz already, the OP would be fixed by now ;)
 

syaoran12

Senior Member
Jun 25, 2010
168
46
Anyone else having an issue downloading from the link?

Sent from my SCH-I500 using XDA App