Is HummingBird Really Slower than Snapdragon Gen 2? [Independent of JIT] - Vibrant General

I know this topic has been debates over time but I noticed that most people attributed the differences in performance is caused by firmware difference (2.1 vs. 2.2).
Today there's an article release about G2 overlock to 1.42 Ghz. Along with the article I noticed "Native Benchmark" using SetCPU which doesn't uses JIT.
Lower is Better.
G2 Result:
{
"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"
}
Now My Vibrant at 1.2 Ghz:
C: 702.89
Neon: 283.15
The difference between the two phone is so great that I doubt it is due to the 200 MHz difference alone.
As a comparison, my score at regular 1 GHz is:
C: 839.21
Neon: 334.51
There is about 130 ms decrease for 200 Mhz overclock, which is Vibrant is at 1.4 Ghz would put the two CPU really close to each other but with G2 having a slight edge. Remember this test is suppose to be JIT independent running Native Codes. But since the vibrant can only be stable overclocked to 1.3 Ghz (what is available anyways), the newer generation of Snapdragon may just be more efficient than Hummingbird, despite us the galaxy owner believes otherwise.
Another thing to keep in mind though, is that Snapdragon are supposedly to have an edge in Neon instruction Set, so I didn't look into that score too much.

It appears to be true.
It appears Hummingbird is not only slower than the new Generation Scorpions, it also appears the Hummingbird is unable to fully capture the CPU performance gain of the Dalvik JIT compiler in Froyo 2.2
http://www.youtube.com/watch?v=yAZYSVr2Bhc

Dunno Something Is Not Right About This 2.2
The Thing That Really Bugs Me Is 2.2 is Suppose To Allow The Full Functionality Of Our 512MB of Ram..But It Doesn't

Erickomen27 said:
Dunno Something Is Not Right About This 2.2
The Thing That Really Bugs Me Is 2.2 is Suppose To Allow The Full Functionality Of Our 512MB of Ram..But It Doesn't
Click to expand...
Click to collapse
It's not 2.2, its Samsung.

SamsungVibrant said:
It's not 2.2, its Samsung.
Click to expand...
Click to collapse
I agree, they should use ext 4 on their phones.
I don't see why they would stick to their old RFS.

SamsungVibrant said:
It appears Hummingbird is not only slower than the new Generation Scorpions, it also appears the Hummingbird is unable to fully capture the CPU performance gain of the Dalvik JIT compiler in Froyo 2.2
http://www.youtube.com/watch?v=yAZYSVr2Bhc
Click to expand...
Click to collapse
I'm sorry, but could you explain what your youtube link has to do with the topic? I'm curious, as I wasn't any wiser on the question at hand when I watched it.

NEON is architecture extension for the ARM Cortex™-A series processors*
Is Snapdragon an ARM Cortex™-A series processor? NO!
Remember SSE instruction set in Intel, and the war AMD vs Intel?
Welcome back, LOL
*The source for NEON: http://www.arm.com/products/processors/technologies/neon.php

Probably is, but does it really matter?
Sent from my SGS Vibrant.

Scorpion/Snapdragon have faster FPU performance due to a 128 bit SIMD FPU datapath compared to Cortex-A8's 64 bit implementation. Both FPUs process the same SIMD-style instructions, the Scorpion/snapdragon just happens to be able to do twice as much.
http://www.insidedsp.com/Articles/t...ualcomm-Reveals-Details-on-Scorpion-Core.aspx
2.2 isnt going to magically give the galaxy S similar scorpion/snapdragon high scores
just look at droidX and other Cortex-A8 phones that already have official 2.2 ROMS they avr 15-20 linpack scores
This doesn't make the hummingbird a bad CPU at all LOL its stupid benchmarks IMHO not going to show in realword use...maybe when the OS matures and becomes more complex but not now..and even by then we will have dualcore CPU's...its a gimmick for HTC to have the "Fastest CPU"
IMO in real world use they are pretty much on par but then when you look at GPU performance its quit obvious the galaxy S pulls ahead thanks to the 90mts PowerVR SGX540

demo23019 said:
Scorpion/Snapdragon have faster FPU performance due to a 128 bit SIMD FPU datapath compared to Cortex-A8's 64 bit implementation. Both FPUs process the same SIMD-style instructions, the Scorpion/snapdragon just happens to be able to do twice as much.
http://www.insidedsp.com/Articles/t...ualcomm-Reveals-Details-on-Scorpion-Core.aspx
2.2 isnt going to magically give the galaxy S similar scorpion/snapdragon high scores
just look at droidX and other Cortex-A8 phones that already have official 2.2 ROMS they avr 15-20 linpack scores
This doesn't make the hummingbird a bad CPU at all LOL its stupid benchmarks IMHO not going to show in realword use...maybe when the OS matures and becomes more complex but not now..and even by then we will have dualcore CPU's...its a gimmick for HTC to have the "Fastest CPU"
IMO in real world use they are pretty much on par but then when you look at GPU performance its quit obvious the galaxy S pulls ahead thanks to the 90mts PowerVR SGX540
Click to expand...
Click to collapse
Once again quoting ARM HQ website:
NEON technology is cleanly architected and works seamlessly with its own independent pipeline and register file.
NEON technology is a 128 bit SIMD (Single Instruction, Multiple Data) architecture extension for the ARM Cortex™-A series processors, designed to provide flexible and powerful acceleration for consumer multimedia applications, delivering a significantly enhanced user experience. It has 32 registers, 64-bits wide (dual view as 16 registers, 128-bits wide.
Click to expand...
Click to collapse
Scorpion is not ARM Cortex™-A series processor

Fuskand said:
I'm sorry, but could you explain what your youtube link has to do with the topic? I'm curious, as I wasn't any wiser on the question at hand when I watched it.
Click to expand...
Click to collapse
I provided the link, because the first part of the link talks about the JIT compiler which increases CPU performance. I put that there in-case someone has never heard of this before. Thus, when I mentioned the Hummingbird can not take full advantage of the JIT compiler, someone would know what I'm talking about.

demo23019 said:
Scorpion/Snapdragon have faster FPU performance due to a 128 bit SIMD FPU datapath compared to Cortex-A8's 64 bit implementation. Both FPUs process the same SIMD-style instructions, the Scorpion/snapdragon just happens to be able to do twice as much.
http://www.insidedsp.com/Articles/t...ualcomm-Reveals-Details-on-Scorpion-Core.aspx
2.2 isnt going to magically give the galaxy S similar scorpion/snapdragon high scores
just look at droidX and other Cortex-A8 phones that already have official 2.2 ROMS they avr 15-20 linpack scores
This doesn't make the hummingbird a bad CPU at all LOL its stupid benchmarks IMHO not going to show in realword use...maybe when the OS matures and becomes more complex but not now..and even by then we will have dualcore CPU's...its a gimmick for HTC to have the "Fastest CPU"
IMO in real world use they are pretty much on par but then when you look at GPU performance its quit obvious the galaxy S pulls ahead thanks to the 90mts PowerVR SGX540
Click to expand...
Click to collapse
Search the net, people have made real world Videos of galaxy s running 2.2, compared to G2. The G2 is faster in the real world on things like launching aps.

lqaddict said:
Once again quoting ARM HQ website:
Scorpion is not ARM Cortex™-A series processor
Click to expand...
Click to collapse
LOL i never said the scorpion is ARM Cortex™-A
try reading my post again
SamsungVibrant said:
Search the net, people have made real world Videos of galaxy s running 2.2, compared to G2. The G2 is faster in the real world on things like launching aps.
Click to expand...
Click to collapse
LOl if it is faster it might be by the most 1-2 sec if its lucky
sorry its going to take allot more than that to impress me..again its a phone now a highend PC

SamsungVibrant said:
Search the net, people have made real world Videos of galaxy s running 2.2, compared to G2. The G2 is faster in the real world on things like launching aps.
Click to expand...
Click to collapse
Due to different filesystem implementation largely, once there is a workable hack to convert the entire filesystem on the Galaxy S to a real filesystem you can make the comparison of the things like launching apps.
Demo, I didn't mean to come off as a **** I was just pointing out the flaw in the OP benchmark - NEON instruction set execution is flawed. G2 processor is ARMv7 which is the base of Cortex-A8, Cortex-A8 adds the instructions specifically targeted for application, like multimedia, and that's where NEON comes into place.

lqaddict said:
Due to different filesystem implementation largely, once there is a workable hack to convert the entire filesystem on the Galaxy S to a real filesystem you can make the comparison of the things like launching apps.
Click to expand...
Click to collapse
Agreed. +10 char

lqaddict said:
Demo, I didn't mean to come off as a **** I was just pointing out the flaw in the OP benchmark - NEON instruction set execution is flawed. G2 processor is ARMv7 which is the base of Cortex-A8, Cortex-A8 adds the instructions specifically targeted for application, like multimedia, and that's where NEON comes into place.
Click to expand...
Click to collapse
No problem I didn't really take it
Also noticed i overlooks allot of things in OP...blame the ADD

What difference does it make? In real world use the difference is negligible. And in three months our phones will be mid-tier anyway. At least we don't have hinges that will fall apart in two months.

how is the hummingbird not able to fully take advantage of JIT?

Well there is a fix for our phones now. And from what I can tell there no way the g2 can open apps faster than my vibrant with the z4mod. Its smocking fast.by far the fastest I've ever seen this phone. No delays whatsoever. Can't wait till I get froyo with ocuv and this will be unreal. I feel like this phone us a high end pc running android or something. When I say instant it's instant lol.

Kubernetes said:
What difference does it make? In real world use the difference is negligible. And in three months our phones will be mid-tier anyway. At least we don't have hinges that will fall apart in two months.
Click to expand...
Click to collapse
Exactly.
People seem to forget that the SGS line is like 6 months old now, we should be glad they're still doing as well as they are.
Then theres the fact that there aren't many other phones that come with 16gb internal. To me, having 16GB and being able to upgrade to 48 [minus the 2GB that Samsung steals] total is worth way more than starting with 4GB [1.5GB usable] and upgrading to a max of 36 [minus what HTC steals from internal].
But, if you don't like your phone, SELL IT while it's still worth something!

Related

G2 Processor faster than Nexus One?

I constantly see references to the G2 beating the Nexus One in benchmark tests but the benchmarks always include GPU performance as part of the score. It was established before the phone came out that the GPU performance would beat just about everything else (save the Galaxy S), however that is somehow conflated with CPU performance.
Yes, the G2 can and probably will be overclocked once root is achieved but for now, apples to apples comparison:
{
"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"
}
Overall score is higher than Nexus One but CPU score is lower
I don't know that this is all that significant and I'd still take my G2 over a Nexus One, but please stop spreading misinformation by saying how the "Scorpion CPU pwnz the Snapdragon LOLZWTF!!!!1!!"
And I'm sure once the G2 get clocked At its native speed then its gonna put your point to rest...G2 still runs faster than my nexus .. I don't care about numbers what I care about is real world performance
Sent from my T-Mobile G2 using XDA App
Both are Snapdragon SoC (G2 and N1), and both have Scorpion CPU's.. newer one is on a 45nm die while older is on a 65nm (lesser heat emission, lower consumption of power as a result longer battery life on the G2 thanks to the 45nm die).
CPU architecture wise there is not much difference.. the application processing however has been tweaked to perform better overall(thanks to the 45nm die I believe) and obviously as you can see the GPU, Adreno 205 thrashes the older Adreno 200..
Also the newer SoC supports more video codecs(mainly DivX actually) as well. ((MPEG-4, H.264, H.263, VC-1, DivX, DivX 3.11, Sorenson Spark, VP6) vs (MPEG-4, H.264, H.263, VC-1, Sorenson Spark, VP6).
msmith1991 said:
And I'm sure once the G2 get clocked At its native speed then its gonna put your point to rest
Click to expand...
Click to collapse
I constantly see references to the G2 beating the Nexus One in benchmark tests
Click to expand...
Click to collapse
Yes, the G2 can and probably will be overclocked once root is achieved but for now, apples to apples comparison
Click to expand...
Click to collapse
I don't know if I came across as bashing the phone. Maybe it was when I said I still prefer it over a Nexus one?
Regardless, my point wouldn't be "put to rest." For now, the CPU on the Nexus One outperforms the G2 on a benchmark. That is what I was addressing, nothing more.
Superfrag said:
Both are Snapdragon SoC (G2 and N1), and both have Scorpion CPU's.. newer one is on a 45nm die while older is on a 65nm (lesser heat emission, lower consumption of power as a result longer battery life on the G2 thanks to the 45nm die).
CPU architecture wise there is not much difference.. the application processing however has been tweaked to perform better overall(thanks to the 45nm die I believe) and obviously as you can see the GPU, Adreno 205 thrashes the older Adreno 200..
Also the newer SoC supports more video codecs(mainly DivX actually) as well. ((MPEG-4, H.264, H.263, VC-1, DivX, DivX 3.11, Sorenson Spark, VP6) vs (MPEG-4, H.264, H.263, VC-1, Sorenson Spark, VP6).
Click to expand...
Click to collapse
My personal experience matches what you've said exactly. Barely a hiccup running the phone and excellent battery life (relatively speaking at least). Like I said, the only thing I'm addressing is people using benchmarks which include the GPU in the score to say that the G2 processor is faster at 800mhz than the N1 at 1ghz.
Is there anything like Adobe Flash benchmark or test page? G2 is supposed to hardware accelerate Flash.
G2 is pure sex compared to my last phone, motorola cliq hahah
Is there anything like Adobe Flash benchmark or test page? G2 is supposed to hardware accelerate Flash.
Click to expand...
Click to collapse
This is particularly interesting, as the phone seems to load and scroll webpages faster than my Nexus One did with Enom's ROM. Well, the G2 is faster in everything, basically. Not by a lot, but enough to be noticeable.
What is interesting is the obnoxiously higher IO score in the G2 compared to the N1.
I wonder what is entailed in an IO test on a phone...
here is some light reading, and chart lookin....
http://androidandme.com/2010/10/news/3dmarkmobile-gpu-showdown-adreno-205-vs-powervr-sgx540/
But jsut remember that the Galaxy class is still stuck on 2.1 so we are missing some updated drivers and JIT and yes that does make a difference.
Oh I forgot.
Better IO scores I believe is thanks to the fact that HTC decided to to with ext3 partition for its OS, thus having faster read/writes(I'm not sure about this.. I think its thanks to the ext3 partition)
Also yeah, like I said, the CPU architecture is similar, (very similar actually), just the overall application processing is better due to tweaks, and one guy mentioned hardware flash acceleration which I forgot to mention, and obviously the GPU. Basically its a very well refined package of the Snapdragons that were on the N1/Desire.
Next year's 3rd Gen Snapdragons will be dual core Scorpion CPU's, even better tweaking and optimization, and HUGE gpu improvement.
Qualcomm says that Adreno 220 is 4-5 times faster than Adreno 205.
carlitozway57 said:
My personal experience matches what you've said exactly. Barely a hiccup running the phone and excellent battery life (relatively speaking at least). Like I said, the only thing I'm addressing is people using benchmarks which include the GPU in the score to say that the G2 processor is faster at 800mhz than the N1 at 1ghz.
Click to expand...
Click to collapse
Yup, at stock speeds the Nexus One's CPU is faster simply due to clock speed(since they have similar architecture). BUT, thanks to the overall package, tweaking and optimization, and a better GPU, the new 2nd gen Snapdragons are extremely well refined and thus the phone runs MUCH smoother. That's why you see Linpack scores of the G2 and N1 are very similar.
Superfrag said:
Yup, at stock speeds the Nexus One's CPU is faster simply due to clock speed(since they have similar architecture). BUT, thanks to the overall package, tweaking and optimization, and a better GPU, the new 2nd gen Snapdragons are extremely well refined and thus the phone runs MUCH smoother. That's why you see Linpack scores of the G2 and N1 are very similar.
Click to expand...
Click to collapse
agreed
my G2 scores 1675 in quadrant
smoothness comparable to iphone, but just running at 245mhz!
many people got scared of for the 800mhz mark
but I would proudly say they have underestimated that babe
running at 800mhz brings satisfactory battery stamina as well
o>c said:
agreed
my G2 scores 1675 in quadrant
smoothness comparable to iphone, but just running at 245mhz!
many people got scared of for the 800mhz mark
but I would proudly say they have underestimated that babe
running at 800mhz brings satisfactory battery stamina as well
Click to expand...
Click to collapse
Tell me about it, it's the smoothest Android phone out there and fastest too I might add.
Once we root and O/C this baby it'll fly!
Think of the difference between the N1 QSD8250 vs the G2 MSM7230 as a Core 2 Duo E6700 vs Core 2 Duo E7200 for the most part. One is older, supports slightly less instructions, but the overall architecture is the same. They just refined it a bit when they made the switch to 45nm is all.
I think the G2 cpu after using almost every snapdragon phone over the few months including my old nexus 1 this 800mhz cpu reminds me of the intel centrino cpus although they lower clock rates the where fast cpus from what I see and read the G2 as a package is the fastest phone on the martket. The cpu has been optimise and after going through many androids phones this one is by far the smoothest and fastest of them all in my opinion.
Sent from my T-Mobile G2 using XDA App
This is the highest I've been able to get it to
carlitozway57 said:
I don't know that this is all that significant and I'd still take my G2 over a Nexus One, but please stop spreading misinformation by saying how the "Scorpion CPU pwnz the Snapdragon LOLZWTF!!!!1!!"
Click to expand...
Click to collapse
I haven't done any benchmarking but I can definitively say the G2 is far and away snappier and seems to run much smoother than the N1. It's the first time an android phone has seemed to just flow and not get in the way of itself. Granted I stopped using cooked roms when I ditched winmo, so I can't speak for CM6.
Sent from my T-Mobile G2 using Tapatalk
Of course the absolute correct statement would be chipset vs chipset, not processor vs processor, but thats all semantics. You cannot compare processors vs processors when it comes to mobile devices. These are all SOCs. You cannot seperate them.
And the fact is, the G2 as a whole has better performance than the N1 as a whole. The reason why people at the first place keep harping on the fact that G2 is faster than N1 is just to make a point that being 800mhz does not mean that the phone as a whole is slower than the N1, and thus less qualified for a Gingerbread upgrade than the N1 (which in turn proves that the so called rumour of a 1Ghz requirement for Gingerbread is not logical). How can one say that just because N1 has 1Ghz proc, and the G2 has 800mhz proc, the G2 is not as qualified as the N1 to get Gingerbread (again, assuming the 1Ghz requirement) when as a whole phone its faster than the N1. So we can conclude that the 1Ghz requirement is just bull****, and that the G2 is not handicapped for having a 800mhz proc.
Sorry did i confused you? I think i confused myself
I was able to get it 1 point higher lmao
Sent from my T-Mobile G2 using XDA App
Some of you keep thinking the G2 has a snapdragon processor in it. It's a scorpion processor not a 2nd gen snapdragon.

If Nexus S (2?) has similar hardware to Galaxy S, is it easy to port 2.3?

Even though I have followed dev/porting for 8 months starting with HTC Touch, I have little knowledge of how it is actually done. So here's my question to the developers.
We all know that Nexus S (2?) will have Gingerbread 2.3. Looking at the rumored specs and model number, it seems that Nexus S is a slight upgrade from Galaxy S.
Assuming most of the hardware is identical to Galaxy S, how easy is it to port 2.3 to Epic 4G, once Galaxy S becomes available?
Specifically, what is needed to bake a new 2.3 rom? Do you need to reverse engineer like what devs did on HTC WM devices? Or is it a straight port? I suspect it's somewhere in between but want to hear from you.
(If necessary, please move this to android development forum.)
The simple answer to this (which has been answered in other threads already if you looked) is no its won't be easy.
Also, the Nexus S is rumored and pretty much guaranteed to launch with a dual core processor. The rumor is that they delayed the device and gingerbread to implement this, since it will be google's new flagship device and has to be cutting edge. Everyone knows that dual core processors are set to hit the market within the first couple months of 2011 anyway, so releasing an old generation processor in a flagship google phone just makes no sense.
So no, it will not be easy to port from the Nexus S. It will not only have a completely different processor, but will also probably only be a GSM phone.
muyoso said:
but will also probably only be a GSM phone.
Click to expand...
Click to collapse
I agree and think this is the bigger problem (for the Epic at least).
Sure would be nice if the folks at Google would release at one clean Google device for each carrier. I'd be on it in a heart beat.
vansmack said:
I agree and think this is the bigger problem (for the Epic at least).
Sure would be nice if the folks at Google would release at one clean Google device for each carrier. I'd be on it in a heart beat.
Click to expand...
Click to collapse
Yes that would be nice if they did that.I am surprised they are even doing another google phone by the way it sounded when they stopped marketing the nexus 1 that had no interest in doing another google branded phone.
muyoso said:
Also, the Nexus S is rumored and pretty much guaranteed to launch with a dual core processor. The rumor is that they delayed the device and gingerbread to implement this, since it will be google's new flagship device and has to be cutting edge. Everyone knows that dual core processors are set to hit the market within the first couple months of 2011 anyway, so releasing an old generation processor in a flagship google phone just makes no sense.
So no, it will not be easy to port from the Nexus S. It will not only have a completely different processor, but will also probably only be a GSM phone.
Click to expand...
Click to collapse
Those are just rumors, all of the official specs that have come out say it will be a 1.2ghz hummingbird. The dual core 1ghz orion chip is definitely on the horizon but I highly doubt they will be able to get it in at the last minute, and there's a good chance we won't see it in phones until late next year. All rumors of a dual core Nexus S have had no credibility with their sources.
That said, if this turns out to be a dual core phone and gingerbread turns out to be optimized for dual cores, a port will probably be very difficult. But if it's just a 1.2ghz hummingbird then it would just be a matter of getting the CDMA radio working.
LucJoe said:
Those are just rumors, all of the official specs that have come out say it will be a 1.2ghz hummingbird. The dual core 1ghz orion chip is definitely on the horizon but I highly doubt they will be able to get it in at the last minute, and there's a good chance we won't see it in phones until late next year. All rumors of a dual core Nexus S have had no credibility with their sources.
That said, if this turns out to be a dual core phone and gingerbread turns out to be optimized for dual cores, a port will probably be very difficult. But if it's just a 1.2ghz hummingbird then it would just be a matter of getting the CDMA radio working.
Click to expand...
Click to collapse
The reason that I think he has to have a dual core in it is as follows:
If it has a 1.2 ghz hummingbird processor, BFD. It immediately launches and is mediocre. Nothing exciting at all about it.
If it launches as the first dual core phone, it is the top of the line phone worthy of being branded as a Google flagship device.
Also, if it just had a 1.2 ghz hummingbird processor, what is the holdup? It is no different from phones released months ago. Also, if its a 1.2 Ghz processor, it will be eclipsed within a matter of a month or two performance wise by Tegra 2 and dual core snapdragon processors. Basically, it would be an embarassing flagship device. The original Nexus is still to this day a damn good phone that is near the top of the pack of android phones performance wise, and it is a year old.
muyoso said:
The reason that I think he has to have a dual core in it is as follows:
If it has a 1.2 ghz hummingbird processor, BFD. It immediately launches and is mediocre. Nothing exciting at all about it.
If it launches as the first dual core phone, it is the top of the line phone worthy of being branded as a Google flagship device.
Also, if it just had a 1.2 ghz hummingbird processor, what is the holdup? It is no different from phones released months ago. Also, if its a 1.2 Ghz processor, it will be eclipsed within a matter of a month or two performance wise by Tegra 2 and dual core snapdragon processors. Basically, it would be an embarassing flagship device. The original Nexus is still to this day a damn good phone that is near the top of the pack of android phones performance wise, and it is a year old.
Click to expand...
Click to collapse
You are fully aware the the Tegra 2 do not even exceed the hummingbird in gpu or cpu performance right? Im just saying cause it would suck if you didn't know what you're talking about.
Plus gingerbread will have HW acceleration, putting gpu performance on a step for the overall fluidity of the gui. So again... what's faster?
Really? I assumed it would greatly ourperform. Where did u get your facts.
Sent from my SPH-D700 using XDA App
InfDaMarvel said:
Really? I assumed it would greatly ourperform. Where did u get your facts.
Sent from my SPH-D700 using XDA App
Click to expand...
Click to collapse
I'm trying to find the article again, but i know they were close but Tegra 2 did not outperform the hummingbird. Apparently now they optimized the platform more.
Don't get me wrong i love nvidia, that's all i've purchased and stayed with them even thou they still dont have a decent dx11 card that doesnt need 2 power supplies. But they really need to step up and their CEO needs to wattch what he says and deliver more.
Here's a quote "On the 3D side, Nvidia says it has doubled the performance of the initial Tegra, resulting in a peak speed of 90 million triangles per second. This level is well beyond the performance of any mobile processor shipping or even sampling today." Hummingbird has the same exact performance. And CPU performance is a very interesting area. Anyway the GPU performance is almost par with the Hummingbird leading maybe by 3-5%
apatcas said:
I'm trying to find the article again, but i know they were close but Tegra 2 did not outperform the hummingbird. Apparently now they optimized the platform more.
Don't get me wrong i love nvidia, that's all i've purchased and stayed with them even thou they still dont have a decent dx11 card that doesnt need 2 power supplies. But they really need to step up and their CEO needs to wattch what he says and deliver more.
Here's a quote "On the 3D side, Nvidia says it has doubled the performance of the initial Tegra, resulting in a peak speed of 90 million triangles per second. This level is well beyond the performance of any mobile processor shipping or even sampling today." Hummingbird has the same exact performance. And CPU performance is a very interesting area. Anyway the GPU performance is almost par with the Hummingbird leading maybe by 3-5%
Click to expand...
Click to collapse
Since Tegra 2 is dual core and android does not have 2 core support till Gingerbread (Actually I don't think it even supports cortex A9 till gingerbread)..so if they ran 1 core vs 1 core I'd see a hummingbird win against a Tegra 2..but if Tegra 2 is running dual core (and optimized for it) it should win...but by that analogy Orion would then be superior.
gTen said:
Since Tegra 2 is dual core and android does not have 2 core support till Gingerbread (Actually I don't think it even supports cortex A9 till gingerbread)..so if they ran 1 core vs 1 core I'd see a hummingbird win against a Tegra 2..but if Tegra 2 is running dual core (and optimized for it) it should win...but by that analogy Orion would then be superior.
Click to expand...
Click to collapse
Not to mention that Tegra 2 does 1080P video recording. So yes, releasing a google flagship phone that within one month is eclipsed by LG with the first Tegra 2 phone, would be embarrassing. The Nexus 1 set the standard for almost a year, before the Galaxy S line came out. If the Nexus 2 can only set the standard for under a month, that would be stupid. Therefore, it is easy to conclude that the rumors of the Nexus S having a dual core are most likely true. Doesn't mean it has to be the Orion, but it would be awesome if it was.
Tegra 2 is a Cortex A9 CPU... as is the Samsung Orion and the TI OMAP4xxx chips. They accomplish 2.5 instructions per MHz as opposed to the 2 instructions per MHz in the Cortex A8 Hummingbird, and that's not counting improvements to instruction efficiency (getting more done with less instructions.) Add to that improvements such as out of order instruction handling and dual-channel memory support and Cortex A9 chips are head and shoulders above Cortex A8.
The only reason Tegra 2 wouldn't outperform Hummingbird significantly is, as mentioned, lack of dual-core support in current builds of Android, and the nVidia GPU which is, surprisingly, only just about on par with Hummingbird's PowerVR SGX540.
Sent from my SPH-D700 using XDA App
Electrofreak said:
Tegra 2 is a Cortex A9 CPU... as is the Samsung Orion and the TI OMAP4xxx chips. They accomplish 2.5 instructions per MHz as opposed to the 2 instructions per MHz in the Cortex A8 Hummingbird, and that's not counting improvements to instruction efficiency (getting more done with less instructions.) Add to that improvements such as out of order instruction handling and dual-channel memory support and Cortex A9 chips are head and shoulders above Cortex A8.
The only reason Tegra 2 wouldn't outperform Hummingbird significantly is, as mentioned, lack of dual-core support in current builds of Android, and the nVidia GPU which is, surprisingly, only just about on par with Hummingbird's PowerVR SGX540.
Sent from my SPH-D700 using XDA App
Click to expand...
Click to collapse
OK here's facts dudes. Tegra sucks... really please get it.
Tegra250 based Toshiba AC100 Running Neocore Benchmark
http://www.youtube.com/watch?v=uJav9ns6b4o
apatcas said:
OK here's facts dudes. Tegra sucks... really please get it.
Tegra250 based Toshiba AC100 Running Neocore Benchmark
http://www.youtube.com/watch?v=uJav9ns6b4o
Click to expand...
Click to collapse
{
"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"
}
I'm not sure that triangles per second is accurate to describe performance.
Still, if you want an article:
http://alienbabeltech.com/main/?p=17125
The Hummingbird has memory bandwidth limitations that I don't think the Tegra 250 will. Lets wait and see.
apatcas said:
OK here's facts dudes. Tegra sucks... really please get it.
Tegra250 based Toshiba AC100 Running Neocore Benchmark
http://www.youtube.com/watch?v=uJav9ns6b4o
Click to expand...
Click to collapse
Here's the facts dude, and try to get this; Tegra2 does not, in fact, suck. You just posted a video of it being benchmarked on a netbook running Android 2.1 which cannot make full use of Tegra2's dual-core CPU. Secondly, neocore is a GPU test, not a CPU test. We already discussed the fact that the Tegra2 GPU is only just about on par with the SGX540. Thirdly, that test is being run at a signifcantly higher resolution than a mobile device would run, and frankly, considering this, the score isn't bad.
Fail, man, fail.
Sent from my SPH-D700 using XDA App
sauron0101 said:
I'm not sure that triangles per second is accurate to describe performance.
Still, if you want an article:
http://alienbabeltech.com/main/?p=17125
The Hummingbird has memory bandwidth limitations that I don't think the Tegra 250 will. Lets wait and see.
Click to expand...
Click to collapse
I would also like to point out that I wrote the article that sauron0101 just linked. It's also posted on my blog (linked in my signature) posted back in March. Tegra2 does feature dual-channel memory support as part of the Cortex A9 architecture, which is a significant advantage.
Sent from my SPH-D700 using XDA App
Electrofreak said:
I would also like to point out that I wrote the article that sauron0101 just linked. It's also posted on my blog (linked in my signature) posted back in March. Tegra2 does feature dual-channel memory support as part of the Cortex A9 architecture, which is a significant advantage.
Sent from my SPH-D700 using XDA App
Click to expand...
Click to collapse
Any idea how fast the other Arm Cortex A9s are compared to the 250?
- We know that the SGX540 will be in the TI OMAP 4 series; probably not bandwidth limited - I am surprised that they did not opt for the SGX545
- The Samsung Orion series has Mail 400 (unknown performance)
- The Snapdragon (A8 unless Qualcomm opts to keep the name "Snapdragon for its A9 CPUs) will have a new generation of Adreno 300 graphics
Unknown if we will see this on mobile
- Marvell also has a new SOC
Also interesting is Samsung's Netbook roadmap, which uses the same SOCs on a phone:
Sorry if all of this is a bit off topic, but it is worth looking at what everyone has.
Edit: Qualcomm is keeping the Snapdragon name for the A9 processors.
Does no one see i was talking about Gpu perfomance? That's what's gonna matter in Gingerbread. And that's running 1024x600 on that res Galaxy tab is around 53 fps. It's the same thing that Vista started doing with HW accel so u understand.

[TEST] Tegra 2 OC vs Tegra 3

Hello guys! Walking around the net I found some diagrams that describe the benchmarks of Tegra 2 stock vs. 3 relative to other processors. I always wondered if now that the processor is overclocked to 1.7GHz (70% more power) has reduced the distances from the future quad-core (which we remember to be 5 times faster than current SoC as Nvidia has said).
Some developers can make this comparison? It would be interesting to see the benefits
"when looking at the compiler versions and settings used to compile the CoreMark benchmark, the Core 2 Duo numbers were produced via GCC 3.4 and only the standard set of optimizations (-O2), while the Tegra 3 numbers were run on a more recent GCC 4.4 with aggressive optimizations (-O3). Il Sistemista website took a Core 2 Duo T7200 and re-ran the benchmark compiled with GCC 4.4 and the same optimization settings. The results were no longer in favor of NVIDIA, as the Core 2 chip scored about 15,200 points, compared to the Tegra's 11,352."
{
"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"
}
CoreMark benchmark comparing nVidia Tegra 3 @1GHz clock to various real and hypothetical products
CoreMark/MHz index shows how much Coremarks can a particular chip extract given its frequency
WoW....More powerful than Core2Due O_O!!!
Now THATS something to get excited about!!!
Ahmed_PDA2K said:
WoW....More powerful than Core2Due O_O!!!
Now THATS something to get excited about!!!
Click to expand...
Click to collapse
if you read the bold they fixed compiler optimizations and the Tegra did not hold up
At this point it would be really interesting to know how much you have reduced the gap with the super-overclock to 1.7GHz of Tegra (which is really the maximum achievable?). I don't know how to operate the CoreMark 1.0 so I ask someone in the community for check if values ​​are underestimated or substantially improved compared to those offered by Nvidia and especially to see if at this point Tegra 3 can really be worth .
I recently read that Nvidia is pushing a lot of their projects and have already popped the development of Tegra 4. The specifications include a 4 Tegra chip manufacturing to 28Nm. IMHO it would be more correct to wait for this to update the Tegra 2 hardware version (so these benchmarks could be a way to understand this in this context).
I highly doubt many people can get their tegra 2 up to 1.7, I can only get mine stable at 1.4, probably 1.5 and maybe 1.6 if I spent a lot of time messing with the voltages. I think I read somewhere that 1.7ghz produces a lot of heat, like almost in the danger zone but I could be exaggerating it a little.
Sent from my ADR6300 using Tapatalk
In any case it would be nice to make a comparative improvement in all frequencies compared to those values ​​offered by Nvidia, to give an idea of who has the device the hardware's potential. But I think it's a matter of development and optimization as we are seeing here on the forum these days ... the tablet is slowly improving on all fronts
I'm sure that once we have access to the OS code the tf will run like a beast! I had an OG Droid and the difference between stock and modded was mind blowing.
Sent from my ADR6300 using Tapatalk
brando56894 said:
I highly doubt many people can get their tegra 2 up to 1.7, I can only get mine stable at 1.4, probably 1.5 and maybe 1.6 if I spent a lot of time messing with the voltages. I think I read somewhere that 1.7ghz produces a lot of heat, like almost in the danger zone but I could be exaggerating it a little.
Sent from my ADR6300 using Tapatalk
Click to expand...
Click to collapse
I'm one of the few with one that does 1.7GHz no problem.
Its not only the cpu speed of tegra3 that will pwn the transformer1, a 12core gpu and support for dynamic lightning , high profile hd decoding and much more. No way a tegra2 will outperform the tegra3.
Just face it: transformer1 will be pwned big time by transformer2, launched okt/nov 2011....
Not to mention the new Ti cpus coming q1 2012.... they will pwn even more then tegra3...
Sent from my Desire HD using XDA App
Tempie007 said:
Its not only the cpu speed of tegra3 that will pwn the transformer1, a 12core gpu and support for dynamic lightning , high profile hd decoding and much more. No way a tegra2 will outperform the tegra3.
Just face it: transformer1 will be pwned big time by transformer2, launched okt/nov 2011....
Not to mention the new Ti cpus coming q1 2012.... they will pwn even more then tegra3...
Sent from my Desire HD using XDA App
Click to expand...
Click to collapse
Yeah it is true that is stronger, but my thesis is that Tegra 3 is only a technology shift. Although it has 12 computing cores capable of dynamically processing the lights, this CPU produces these values ​​and it would be interesting to relate them to the values ​​of the Tegra 2 overclocked and see if indeed the Tegra 3 can really be a solution to replace Tegra 2. We are faced with two different SoC, the first a dual core, the second a quad core, but probably, common sense tells me that if well exploited this device could give a long hard time to the next model (I remember the story of HTC HD2, which was released back in 2009 and today is one of the longest-running phones through a rearrangement of the software every day). I argue that there is no need for raw power in these devices but the fineness of the calculation, since the batteries are not as performance and putting more capacity probabily they can't make the devices more streamlined in near future.
Someone knows how to use CoreMark to update these values​​?
Of course the tegra two will beat the crap out of the tegra 3, thats like comparing a core 2 duo to a core i7 lol
chatch15117 said:
I'm one of the few with one that does 1.7GHz no problem.
Click to expand...
Click to collapse
Lucky bastard! :-D how many mV are you running it at?
Never tried 1.7Ghz but have done 1.5-1.6 on 3 tabs without messing with voltages. Acutally right now running 1.2Ghz messing with LOWERING the voltages.
Coremark download
If anyone can compile Coremark for run under Android here there is the software to download:
Download the Coremark Software readme file
Download the CoreMark Software documentation
This documentation answers all questions about porting, running and score reporting
Download the Coremark Software
Download the CoreMark Software MD5 Checksum file to verify download
Use this to verify the downloads: >md5sum -c coremark_<version>.md5
Download the CoreMark Platform Ports
These CoreMark ports are intended to be used as examples for your porting efforts. They have not been tested by EEMBC, nor do we guarantee that they will work without modifications.
Here there are result's table:
http://www.coremark.org/benchmark/index.php
If a DEV can compile and Run some test for comparing the results with the new frequencys would be great!
brando56894 said:
I highly doubt many people can get their tegra 2 up to 1.7, I can only get mine stable at 1.4, probably 1.5 and maybe 1.6 if I spent a lot of time messing with the voltages. I think I read somewhere that 1.7ghz produces a lot of heat, like almost in the danger zone but I could be exaggerating it a little.
Sent from my ADR6300 using Tapatalk
Click to expand...
Click to collapse
There is no danger zone.. at least not on my kernel. I haven't posted the one for 3.1 (only 3.01), but there is a built in down clock once the chip reaches temps beyond 47degC... This only happens if the tablet (in laptop form) is left in a bag with the screen on (by accident, sometimes the hinge says its open when its closed).. So if you can run 1.7ghz stable, theres no reason not to run it.. as you'll be protected from heat by the built in thermal throttling.
Default thermal throttling had some seriously bizzare ranges.. 80-120degC..
I wish nvidia's boards were produced like this:
http://hardkernel.com/renewal_2011/main.php
http://www.hardkernel.com/renewal_2011/products/prdt_info.php?g_code=G129705564426
Something modular would be a nice change in the mobile tech segment.. Swapping SoCs would require some serious low-level work tho.. Unless writing to the bootloader and boot partitions were made easy via external interface.. something like that.
Blades said:
There is no danger zone.. at least not on my kernel. I haven't posted the one for 3.1 (only 3.01), but there is a built in down clock once the chip reaches temps beyond 47degC... This only happens if the tablet (in laptop form) is left in a bag with the screen on (by accident, sometimes the hinge says its open when its closed).. So if you can run 1.7ghz stable, theres no reason not to run it.. as you'll be protected from heat by the built in thermal throttling.
Default thermal throttling had some seriously bizzare ranges.. 80-120degC..
I wish nvidia's boards were produced like this:
http://hardkernel.com/renewal_2011/main.php
http://www.hardkernel.com/renewal_2011/products/prdt_info.php?g_code=G129705564426
Something modular would be a nice change in the mobile tech segment.. Swapping SoCs would require some serious low-level work tho.. Unless writing to the bootloader and boot partitions were made easy via external interface.. something like that.
Click to expand...
Click to collapse
And so? are you good to use CoreMark to run some bench on your device and compare with Tegra 3 results? coz here i would test this
devilpera64 said:
"when looking at the compiler versions and settings used to compile the CoreMark benchmark, the Core 2 Duo numbers were produced via GCC 3.4 and only the standard set of optimizations (-O2), while the Tegra 3 numbers were run on a more recent GCC 4.4 with aggressive optimizations (-O3).
Click to expand...
Click to collapse
That statement there tells why benchmarks are extremely useless and misleading.
im one of the few that can reach 1.7 ghz no problem to. never ran it long enough to get hot tho. maybe 30 mins just to run benchmarks. never had fc's or have my system crash

Adreno 225 vs "New" Mali 400MP4

So I have been doing a lot of research looking for what will be better and I am guessing that the Mali 400 is going to out perform the Adreno 225. I wish android had a solid GPU test that would give something close to real world results. But if you take a look at these articles you will see on paper the Adreno is the same as the Apple 4S's Power VR543MP2
http://www.anandtech.com/show/5559/...mance-preview-msm8960-adreno-225-benchmarks/3
Now the International version will have the new Exynos 4 Quad (4412) Quad Core Cortex A9 but the US version is rumored to have a Dual Core Qualcomm Snapdragon MSM8960 with the Adreno 225.
http://www.anandtech.com/show/5811/samsung-galaxy-s-iii-preview
and here are some other benchmarks just to sum up the difference in performance.
http://www.anandtech.com/show/5810/samsung-galaxy-s-iii-performance-preview
The main thing I am worried about is GPU performance the CPUs in just about every phone out right not seem over kill. I want to make sure the phone I buy will be able to run FPSE(Playstation Emulator) and N64oid(N64 Emulator) smooth. FPSE now has an Open GL plugin that needs a hard core GPU to run well. My Galaxy nexus is just not cutting it anymore.
So............... get the international version.
cmd512 said:
So............... get the international version.
Click to expand...
Click to collapse
LTE Speed vs 3G Speed = not worth it.
Don't get me wrong I want a phone with a power house GPU but if the mobile connection is slow its just not worth it. I'm on Verizon and I don't want to move away from their LTE.
Zzim said:
LTE Speed vs 3G Speed = not worth it.
Don't get me wrong I want a phone with a power house GPU but if the mobile connection is slow its just not worth it. I'm on Verizon and I don't want to move away from their LTE.
Click to expand...
Click to collapse
I hear ya, LTE is blazing fast. But on my unbranded SGS2, I get downloads of up to 7.5Mbps, pay $10 a month for unlimited HSPA+ data w/ tethering, and everything is plenty fast for what I do on my phone. So, while LTE is tempting for sure, still doesn't outweigh the other benefits.
Now, if I ever need my phone to seed torrents or something, I'll have to look at LTE then... hah.
cmd512 said:
I hear ya, LTE is blazing fast. But on my unbranded SGS2, I get downloads of up to 7.5Mbps, pay $10 a month for unlimited HSPA+ data w/ tethering, and everything is plenty fast for what I do on my phone. So, while LTE is tempting for sure, still doesn't outweigh the other benefits.
Now, if I ever need my phone to seed torrents or something, I'll have to look at LTE then... hah.
Click to expand...
Click to collapse
Are you with ATT because I can get a line through my work for 20 a month unlimited everything. 7.5 would be enough speed to make me switch and how consistent are these speeds?
How could the just give the us version a dual core? That makes the phone a very slight upgrade to the s2
Sent from my HTC Sensation Z710e using XDA
@ Op
did you see the date of the article regarding "Mobile SoC GPU Comparison" ? its dated february and they are comparing with the sgs2 mali 400 gpu not the one in sgs3. the new mali gpu is already beating all the current lineup of many gpus in many becnhmarks
bala_gamer said:
@ Op
did you see the date of the article regarding "Mobile SoC GPU Comparison" ? its dated february and they are comparing with the sgs2 mali 400 gpu not the one in sgs3. the new mali gpu is already beating all the current lineup of many gpus in many becnhmarks
Click to expand...
Click to collapse
The other articles were just to show the performance of the 225 this article shows how the new Mali will run http://www.anandtech.com/show/5811/samsung-galaxy-s-iii-preview
That article also show that the GS2(Mali400) and the GS3(Mali400MP4) are different in some way.
Zzim said:
Are you with ATT because I can get a line through my work for 20 a month unlimited everything. 7.5 would be enough speed to make me switch and how consistent are these speeds?
Click to expand...
Click to collapse
At work (average congestion), it's consistently 7Mbps+. In areas of great congestion (the mall, etc), it does slow down, but again, for work E-mails, surfing the web, youtube, etc, I've never had issues. Of course, I'm in Austin, TX as well, and I've heard HSPA+ speeds are very much region specific.
If you can get a line through from work with unlimited everything, they may be able to get you onto the smartphone data plan tier, which some folks have gotten up to 10-11+Mpbs. I'm on the $10 a month unlimited non-smartphone plan, so I think AT&T caps it at around 7.5-8Mpbs. Still though, plenty fast for what I do with my phone.
(And, the unlimited tethering is a blessing when you're in airports and stuff. Our US airports blow as there is almost never free WIFI.)
Zzim said:
The other articles were just to show the performance of the 225 this article shows how the new Mali will run http://www.anandtech.com/show/5811/samsung-galaxy-s-iii-preview
Click to expand...
Click to collapse
the scorecharts does shows the new mali400 topping the chart with good margin. what else do you need from a gpu ?
The SGS3's Mali-400 is just overclocked.
Anyways, if the US SGS3 comes with the S4 Pro (which has the new Adreno 320) then the difference in GPU will probably be minor.
dude have you seen those scores ? it beats the 4s graphics which we cant deny has a great gpu...
this is more than just an overclocked mali400 ... it may still be a mali400/mp4 but its not just overclocked its remade and has much higher clocks by the look of it
also im not sure about it coming with the s4 pro with adreno 320... i heard its not ready till end of year at earliest.. the mali-400 was the best android gpu and now its the best mobile gpu out atm
^^
u are right its not only just overclocked,there are some changes in the hardware part which we will know eventually in the upcoming days. i can easily OC my sgs2 mali400 to 400mhz, but you people know it wont give the same result as sgs3 which has much more pixels than s2
urmothersluvr said:
How could the just give the us version a dual core? That makes the phone a very slight upgrade to the s2
Sent from my HTC Sensation Z710e using XDA
Click to expand...
Click to collapse
More cores doesn't = faster...
Look at AMD's bulldozer CPU with 8 cores vs Intel's core i5 with 4 cores...the i5 is faster in basically almost everything except for very specalized applications.
Faster cores > more cores.
The LTE dual core version of the SGS3 will use Krait S4 cores which are faster than A9 Exynos cores.
I wished Samsung did dual core A15s instead of Quad Core A9s.
Daemos said:
More cores doesn't = faster...
Look at AMD's bulldozer CPU with 8 cores vs Intel's core i5 with 4 cores...the i5 is faster in basically almost everything except for very specalized applications.
Faster cores > more cores.
The LTE dual core version of the SGS3 will use Krait S4 cores which are faster than A9 Exynos cores.
I wished Samsung did dual core A15s instead of Quad Core A9s.
Click to expand...
Click to collapse
Let's be clear on this
CPU vs CPU
Dual core S4 is not quicker than Quad Core Exynos
ph00ny said:
Let's be clear on this
CPU vs CPU
Dual core S4 is not quicker than Quad Core Exynos
Click to expand...
Click to collapse
Hmmm I don't know about that...
Zzim said:
Hmmm I don't know about that...
Click to expand...
Click to collapse
I for one certainly do. The Exynos 4412 uses a 32nm fab process as opposed to nearly every other A9 architecture based processor (like the 4+1 T3) and High K metal gate tech which basically means twice the processing power of the Exynos 4410 dual core with about 20% less power consumption and that's on a core against core basis. The 4410 was used in the Galaxy S II. So even if the Exynos 4412 was dual core, it's already natively 20% more battery efficient and twice as powerful than last year's model. Clearly we're talking about a lot more than just quad vs dual and 28nm vs 32 or 40. There is a LOT that has gone into the design of the Exynos. For instance keeping it the same size physically as the dual core model, or accepting 128 bit instructions rather than the paltry 64 bit instructions most other mobile processors are limited to.
Trust me, do your research, a Google search of Exynos 4412 brought up instant results that detail what a beast this chip set is.
Like these:
http://www.phonearena.com/news/Exyn...re-processor-in-the-Samsung-Galaxy-S3_id29615
http://www.phonearena.com/news/Sams...nos-to-appear-in-Samsung-Galaxy-S-III_id29494
And of course the official press release. Read through this and then the benchmarks you pointed out in the OP (I'm linking em anyway) Anandtech's benchmark tests were performed on demo units on display to handled and groped by hundreds of people. There's no telling how many people had used it before they bench marked it and no telling if they were able to do it clean (reboot device, no other apps running). If not than they tested it after some fairly heavy use and it still proved itself a beast.
http://phandroid.com/2012/04/25/sam...ynos-4-quad-for-their-next-generation-galaxy/
http://www.anandtech.com/show/5810/samsung-galaxy-s-iii-performance-preview
Research is your best friend. If you're looking for the most powerful CPU and GPU on a phone right now, this is it. And when the devs get a hold of it, it will become even better and will really be utilized to its full.
Sent from my PG86100 using Tapatalk 2
Gene_Bailey said:
I for one certainly do. The Exynos 4412 uses a 32nm fab process as opposed to nearly every other A9 architecture based processor (like the 4+1 T3) and High K metal gate tech which basically means twice the processing power of the Exynos 4410 dual core with about 20% less power consumption and that's on a core against core basis. The 4410 was used in the Galaxy S II. So even if the Exynos was dual core, it's already natively 20% more battery efficient and twice as powerful. Clearly we're talking about a lot more than just quad vs dual and 28nm vs 32 or 40. There is a LOT that has gone into the design of the Exynos. For instance keeping it the same size physically as the dual core model, or accepting 128 bit instructions rather than the paltry 64 bit instructions most other mobile processors are limited to.
Trust me, do your research, a Google search of Exynos 4412 brought up instant results that detail what a beast this chip set is.
Like these:
http://www.phonearena.com/news/Exyn...re-processor-in-the-Samsung-Galaxy-S3_id29615
http://www.phonearena.com/news/Sams...nos-to-appear-in-Samsung-Galaxy-S-III_id29494
And of course the official press release. Read through this and then the benchmarks you pointed out in the OP (I'm linking em anyway) Anandtech's benchmark tests were performed on demo units on display to handled and groped by hundreds of people. There's no telling how many people had used it before they bench marked it and no telling if they were able to do it clean (reboot device, no other apps running). If not than they tested it after some fairly heavy use and it still proved itself a beast.
http://phandroid.com/2012/04/25/sam...ynos-4-quad-for-their-next-generation-galaxy/
http://www.anandtech.com/show/5810/samsung-galaxy-s-iii-performance-preview
Research is your best friend. If you're looking for the most powerful CPU and GPU on a phone right now, this is it. And when the devs get a hold of it, it will become even better and will really be utilized to its full.
Sent from my PG86100 using Tapatalk 2
Click to expand...
Click to collapse
{
"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"
}
Zzim said:
Hmmm I don't know about that...
Click to expand...
Click to collapse
Outside of floating point tests such as linpack, CPU benches will even have quad core tegra3 well ahead of the dual core S4
Quadrant, Antutu, etc will all show the same exact same performance gap and it's a big one
Let's get this straight
Main selling points for Dual Core S4 setup = battery life from 28nm die size and integrated LTE
Spartoi said:
The SGS3's Mali-400 is just overclocked.
Anyways, if the US SGS3 comes with the S4 Pro (which has the new Adreno 320) then the difference in GPU will probably be minor.
Click to expand...
Click to collapse
you are wrong,mali 400mp4 is a quad core gpu while mali 400 is a dual core gpu.
I want to see mali 400mp4 against sgs543mp4

GLBenchmark HTC ONE only 34 FPS @ Egypt HD 1080P Offscreen

HTC ONE GLBenchmark only scores 34 FPS at 1080P offscreen, this is much lower than the Samsung SHV-E300s which scores 41.3 FPS, both using Snapdragon 600, in the same test. IRC HTC One is using LPDDR2 RAM, so are we seeing a lack of bandwidth compared to the Samsung which may use LPDDR3, which is supported by the S600.
http://www.glbenchmark.com/phonedetails.jsp?benchmark=glpro25&D=HTC+One
how do you know HTC One uses LPDDR2 memory
kultus said:
how do you know HTC One uses LPDDR2 memory
Click to expand...
Click to collapse
http://www.htc.com/uk/smartphones/htc-one/#specs
http://www.anandtech.com/show/6754/hands-on-with-the-htc-one-formerly-m7/2
Turbotab said:
HTC ONE GLBenchmark only scores 34 FPS at 1080P offscreen, this is much lower than the Samsung SHV-E300s which scores 41.3 FPS, both using Snapdragon 600, in the same test. IRC HTC One is using LPDDR2 RAM, so are we seeing a lack of bandwidth compared to the Samsung which may use LPDDR3, which is supported by the S600.
http://www.glbenchmark.com/phonedetails.jsp?benchmark=glpro25&D=HTC+One
Click to expand...
Click to collapse
My first question would be is how they even got a benchmark of the SHV-E300?
Xistance said:
My first question would be is how they even got a benchmark of the SHV-E300?
Click to expand...
Click to collapse
How do any results appear on GLbenchmark?
I believe with GLBenchmark, that if you don't register / login before running the test, it automatically uploads to their server for public viewing, so maybe it was done intentionally, or somebody forgot to login?
fp581 said:
he is spamming all around the htc one just look at his posts plz ban him from posting in any htc forum ever again.
he probably works in sony nokia or samsung
Click to expand...
Click to collapse
Who are you talking about?
sorry wrong person i'll delete that lest one.
but i would love pics of that benchmark for proof
fp581 said:
sorry wrong person i'll delete that lest one.
but i would love pics of that benchmark for proof
Click to expand...
Click to collapse
Dude I was going to go atomic, I admit it I have a terrible temper
I believe the benchmark was run by a German Android site, called Android Next, there is a video on Youtube, the GLBenchmark starts at 2.22
http://www.youtube.com/watch?v=Wl1dmNhhcXs&list=UUan0vBtcwISsThTNo2uZxSQ&index=1
thanks turbo for advanced my knoledge...what a shame they didnt choose LPDDR3 but i think its nt issue these days
Just to temper this news, we must remeber that the HTC One is running at 1.7ghz, while the Samsung device is running at 1.9.
Although 200mhz does not seem like much, it could possibly account for the 7 fps difference when u factor in the difference in UI.
If in fact the Samsung device really has DDR3 ram, and the difference (after accounting for clock speed) is 2-3 fps, I can understand why HTC opted not to include it. Was not worth the extra cost most likely.
Maedhros said:
Just to temper this news, we must remeber that the HTC One is running at 1.7ghz, while the Samsung device is running at 1.9.
Although 200mhz does not seem like much, it could possibly account for the 7 fps difference when u factor in the difference in UI.
If in fact the Samsung device really has DDR3 ram, and the difference (after accounting for clock speed) is 2-3 fps, I can understand why HTC opted not to include it. Was not worth the extra cost most likely.
Click to expand...
Click to collapse
GLBenchmark is a test of GPU performance, and isn't really changed by CPU clockkspeed, but it is affected by bandwidth.
As a test, I downclocked my Nexus 7 from an overclocked 1.6 GHz to just 1.15 GHz, I ran GLBench and got 10 FPS. I then ran at the test again but with CPU at 1.6 GHz, the result, 10 FPS again.
I've benched the N7 with both CPU & GPU overclocked to the same level as Transformer Infinity, which gets 13 FPS, but I always get 10 FPS, the reason my N7 has lower memory bandwidth than the Transformer Infinity, because it use slower RAM and thus has less bandwidth. That is a difference of 30% in FPS, just because of lower bandwidth.
I read that LPDDR3 starts at 800 MHz or 12.8 GB/s in dual-channel configuration, whereas LPDDR2 maxs at 533 MHz or 8.5 GB/s max bandwidth in dual-channel configuration.
Turbotab said:
GLBenchmark is a test of GPU performance, and isn't really changed by CPU clockkspeed, but it is affected by bandwidth.
As a test, I downclocked my Nexus 7 from an overclocked 1.6 GHz to just 1.15 GHz, I ran GLBench and got 10 FPS. I then ran at the test again but with CPU at 1.6 GHz, the result, 10 FPS again.
I've benched the N7 with both CPU & GPU overclocked to the same level as Transformer Infinity, which gets 13 FPS, but I always get 10 FPS, the reason my N7 has lower memory bandwidth than the Transformer Infinity, because it use slower RAM and thus has less bandwidth. That is a difference of 30% in FPS, just because of lower bandwidth.
I read that LPDDR3 starts at 800 MHz or 12.8 GB/s in dual-channel configuration, whereas LPDDR2 maxs at 533 MHz or 8.5 GB/s max bandwidth in dual-channel configuration.
Click to expand...
Click to collapse
In that case the results are quite disappointing.
All these fantastic new phones, and so much disappointment.
Sent from my GT-I9300 using xda premium
Tomatoes8 said:
They could have used faster memory for the same price if they didn't cut off Samsung as a supplier. Makes you wonder where their priorities lie. Making the best products possible or just going with the motions.
Click to expand...
Click to collapse
No one is going to take anything you say here seriously, as you've managed to have 2 threads closed in the last 30 mins. One of those inane posts you made involved you saying that HTC is going to be paying, according to your genius calculation, 20% of their profits to Apple (I forget what insanely unintelligent reason you gave). Yeah, because being able to completely migrate data from 1 completely different phone to another is such a bad idea for a company that wants to push their product.
So, what is the per unit cost of what HTC is paying for RAM now vs. what they could have gotten from Samsung? Exactly, you have no idea. I also didn't hear anything about HTC "cutting off" Samsung as a supplier, but maybe I missed it, so I google'd "htc cut off samsung supplier" and found 2 links...
http://tech2.in.com/news/smartphones/following-apple-htc-cuts-component-orders-from-samsung/505402
http://www.digitimes.com/news/a20121009PD213.html
I'm not sure if you have the capability of reading or not, but I'll spoon feed you this information, ok hunny? I've taken the info from the 1st link, since there is more there.
After Apple Inc slashed its orders for memory chips for its new iPhone from major supplier and competitor, Samsung Electronics Co Ltd, HTC too has reportedly cut down on its smartphone component orders from the South Korean company.
Click to expand...
Click to collapse
So, Apple cut down on memory orders. You know, they are the one's who make the iPhone? Have a logo of an Apple on their products? Steve Jobs was the CEO before he died. Anyway, I'll continue...
According to a report by DigiTimes, HTC has reduced its orders from Samsung, and instead opted to order CMOS image sensors from OmniVision and Sony. The company has also chosen to move part of its AMOLED panel orders to AU Optronics, DigiTimes reported citing ‘sources’.
Click to expand...
Click to collapse
Notice it said that HTC reduced its orders from Samsung, specifically on the image sensors (that's for the camera, if you didn't know) and the screen. You know, the thing on the front of your phone that you touch to make it do things? You know what I mean, right? I encourage you to read this link (or possibly have someone read it to you)...
http://dictionary.reference.com/browse/reduce
The point is that reduce isn't the same as cut off. Cutting off would require HTC not ordering ANYTHING from Samsung. Guess what? The One doesn't use an OmniVision CMOS sensor (don't forget, that's what the camera uses) or an AMOLED screen (the bright part of your phone that shows you info).
Also, this is a far better designed phone, especially in regards to hardware, than anything Samsung has ever produced. I went back to my EVO 4G LTE, mainly because I couldn't stand the terrible build quality of the Note 2. It just feels like a cheap toy. And, IMO, Sense is far better than TW. Samsung may have the market right now because of the Galaxy line of products, but that doesn't mean that HTC is out of the game by any means.
Seriously, attempt to use just a bit of intelligence before opening your mouth and spewing diarrhea throughout the One forums. As the saying goes: "it's better to keep your mouth shut and have people think you're an idiot, then to open your mouth and prove it". Unfortunately for you, it's too late.
I really think Turbo was too hasty to open a new thread for this as we've been discussing this in the mega thread
http://www.glbenchmark.com/phonedetails.jsp?benchmark=glpro25&D=HTC+One
It scores 34fps in Egypt HD 1080p offscreen, while the leaked Samsung s600 device socres 41fps which is perfectly inline with Qualcomm's promised speed (3x Adreno 225)
here is a video of what I suspect the source of the benchmark, because we had no benchmark before it
http://www.youtube.com/watch?v=Wl1dmNhhcXs
notice how the battery is almost at end (HTC bar at this stage means its in the last 25%) also notice the activity in the notification area
more important the post ran more than a few full benchmarks, like quadrant before running GL benchmark, this alone is enough to lower the score, especially since Adreno 320 was known to throttle in the Nexus 4
I think benchmarks scores should not be relied on in such events, especially with hundreds of hands messing with the device, we have learned from the One X launch where videos poped up showing horrible performance from the One X, eventually turned out to be were very far from the final device in ur hands
finally both the One X and Nexus 7 at the same gpu clock, but the first is DDR2 and the second is DDR3, score the same in GL Benchmark
in other words its worrying but it's best to wait for proper testers like Anand
Thread cleaned
...from some serious trolling. There should be no trace from him for some time .
but remember:
{
"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"
}
But...
I just wonder that a Samsung phone uses high end parts from Qualcomm instead of Samsungs processors. But I am not in Samsung devices so far, so I would not judge this
Gz
Eddi
Here's a second video also showing Egypt off screen bench at 34FPS.
https://www.youtube.com/watch?v=wijp79uCwFg
Skip to 3:30
Maedhros said:
Just to temper this news, we must remeber that the HTC One is running at 1.7ghz, while the Samsung device is running at 1.9.
Although 200mhz does not seem like much, it could possibly account for the 7 fps difference when u factor in the difference in UI.
If in fact the Samsung device really has DDR3 ram, and the difference (after accounting for clock speed) is 2-3 fps, I can understand why HTC opted not to include it. Was not worth the extra cost most likely.
Click to expand...
Click to collapse
So you're saying that 200mhz o the CPU can account for 7 fps on a GPU test?
Following what you said, the Nexus 4 should have scored 27 fps? Since it has 200mhz less...
But no, it scored 33.7...only 0.3 fps less than the One!
And you know why? First both use the same GPU (and it's what counts for a graphic test) and second the HTC phones are always slower due to Sense!
So stop *****ing and realize that the One is no god phone
Samsung device is running 4.2.1

Categories

Resources