GT-I9500 vs GT-I9505...Differences are huge!

Status
Not open for further replies.
Search This thread

darrendm

Senior Member
Jan 24, 2009
489
80
London/Leeds/Toronto
Really want to pre-order but feel as if I'm losing out by potentially getting sent the i9505 instead, even though I'm looking at 3G deals from TMob/Orange/Three.
 

SlimJ87D

Senior Member
Jan 16, 2008
2,326
424
I could get either one but I'm going with the SD600 because difference AREN'T huge.

1. Antutu benchmarks are close
2. Battery life is close.
3. There is LTE everywhere I live now.
4. I'll save $150 getting the SD600
5. I've never done anything with y i9300 or even SGH Captivate that pushed the GPU to submission. If you are going for GPU then mmight as well get the Tegra 4 since it will have optimized exclusive games for it.
6. Because there are dozens of more users now getting the SD600 and with Qualcomm releasing the proper source codes, development is going to BOOM on this platform.

So really, the only thing I see with the 4+4 CPU is battery life being a minor advantage of maybe a few minutes.

I think I'm going with the SD600 mostly because it's cheaper for me and LTE is everywhere in USA, CA.
 
Last edited:

bala_gamer

Recognized Contributor
Mar 29, 2011
12,350
7,035
Madurai-Chennai
Xiaomi 13
Xiaomi 13 Pro
I could get either one but I'm going with the SD600 because difference AREN'T huge.


6. Wolfson DAC and Qualcomm's DAC are on par with each other now. From another user here, Berry, I believe the SD600 is using the same DAC as the HTC Butterfly


wolfson wm5102 is found inside s4 9500 and is confirmed by supercurio which would be defenitely better than qualcom variant
 

daleski75

Senior Member
Jul 21, 2008
5,769
6,469
Northampton
I think Samsung have purposely made both phones almost exactly the same performance wise because if the performance delta was much bigger from the snapdragon to the octa-core and 70% of the phones produced will be the snapdragon variant who would buy them??

Everyone would simply wait to try and get hold of the octa-core variant.

The question is would any of you pay a premium for the octa-core even IF it gives you nothing extra over the snapdragon version?
 
Last edited:

katamari201

Senior Member
Oct 18, 2011
449
160
All the differences are very minor, like how the GS3 had the dual-core krait and the quad-core exynos and the two versions didn't differ much performance wise. You'll see slightly better benchmarks and battery on the exynos version, but that's about it. The snapdragon one will have LTE and be far more compatible with apps and games. You will also pay much more for the octa exynos one, which will be rare to find as the majority will be the snapdragon version. Samsung is gonna update the more widespread version first, which is the snapdragon one.
 
  • Like
Reactions: Shni Shnaw

harise100

Senior Member
Oct 20, 2011
1,015
290
The Exynos variant has the same graphics chip as i(diots)Phone. Most games are developed and optimised for this chip. So I think it should be the better choice for gamers. (like me :D)
 
  • Like
Reactions: forelli and soraxd

morning thunder

Senior Member
Oct 5, 2010
73
19
Kuwait
I could get either one but I'm going with the SD600 because difference AREN'T huge.

1. Antutu benchmarks are close

2. Battery life is close.
3. There is LTE everywhere I live now.
4. I'll save $150 getting the SD600
5. I've never done anything with y i9300 or even SGH Captivate that pushed the GPU to submission. If you are going for GPU then mmight as well get the Tegra 4 since it will have optimized exclusive games for it.
6. Because there are dozens of more users now getting the SD600 and with Qualcomm releasing the proper source codes, development is going to BOOM on this platform.

So really, the only thing I see with the 4+4 CPU is battery life being a minor advantage of maybe a few minutes.

I think I'm going with the SD600 mostly because it's cheaper for me and LTE is everywhere in USA, CA.

Not close at all
Exynos = 27417
S600 = 23607
Source:
http://www.gsmarena.com/galaxy_s4_exynos_5_octa_benchmarks_top_the_charts-news-5769.php
 

SlimJ87D

Senior Member
Jan 16, 2008
2,326
424

Are you kidding me? You really think you'd notice that tiny difference from a score that is reported from Antutu? What's it going to do? Open an app a few milliseconds quicker?

When you say the word "HUGE" difference, there really isn't a huge difference.

But again, I've bought into the whole LTE thing because again LTE is now EVERYWHERE I travel to and live. So when I'm surfing the web or tethering for my work laptop, I'll see a much larger increase in internet speeds than I do with the speeds that are similar to my i9300.

I believe my i9300 got around 4.5 mbps at it's best while my GF's SGH-i747 got around 17 mbps, this is in OC, CA USA. This is a HUGE difference.

But yeah, for the countries that are getting the Octa and a working LTE chip, they're the real winners. I don't see why they wouldn't go with the i9500 version.
 

SlimJ87D

Senior Member
Jan 16, 2008
2,326
424
wolfson wm5102 is found inside s4 9500 and is confirmed by supercurio which would be defenitely better than qualcom variant

Here's some stuff from GSMArena

Audio quality
Both smartphones did greatly in our dedicated audio quality test. Last week when we reviewed the HTC One we found it to have the best audio output of any smartphone we have tested over the past few years, but the Samsung Galaxy S4 didn't go down without a fight.
In fact, we were shocked to find out that the Samsung flagship had a cleaner output than the HTC smartphone. While both are perfect when attached to an active external amplifier, the Galaxy S4 remained just as great when we plugged in a pair of headphones, while the HTC One took a hit to its stereo crosstalk.
Still, it's a difference that's very hard to detect without lab equipment, so for purposes practical the two are on par. What is not so hard to hear is the difference in volume levels - the HTC One is notably louder than the Samsung Galaxy S4 in both test scenarios.
And here go the results so you can see for yourselves.
Test Frequency response Noise level Dynamic range THD IMD + Noise Stereo crosstalk
HTC One +0.11, -0.20 -82.7 82.7 0.0023 0.026 -80.9
HTC One (headphones attached) +0.02, -0.32 -82.4 82.3 0.0045 0.057 -68.3
Samsung Galaxy S4 +0.25, -0.16 -82.5 82.4 0.0033 0.023 -81.5
Samsung Galaxy S4 (headphones attached) +0.27, -0.15 -82.3 82.2 0.0053 0.029 -79.5

gsmarena_a001.png


gsmarena_a001.png


The HTC One is one of the best DACs they have ever tested. The SGS4 is close to comparable than it. They mentioned that there's more power in the HTC One. It might be a tad more crisp possibly due to an amplifier someone mentioned earlier. But they are still comparable. So the DACs still might be on par.

But note that Supercurio doesn't agree witht he way GSMArena performs their test.

Still a mystery to me.
 

Bataga

Senior Member
Jan 2, 2012
2,191
932
Georgia
I wouldn't (and don't) care what supercurio says he will do as far as development goes. Look at how he punked out with the nexus 10. Acid audio is better than voodookie imo so...

Sent from my SAMSUNG-SGH-I717 using xda premium
 

SlimJ87D

Senior Member
Jan 16, 2008
2,326
424
I wouldn't (and don't) care what supercurio says he will do as far as development goes. Look at how he punked out with the nexus 10. Acid audio is better than voodookie imo so...

Sent from my SAMSUNG-SGH-I717 using xda premium

Yeah, i have mad respect for him but after he went off on guhvanoh on his Siyahkernel it kind of made my scratch my head a bit.

I think the Wolfson DAC will still be a bit better, but GSMArena has praised the SD600 DAC, the SGS4 isn't as good as the HTC One or Butterfly but it's still good and marginally close in performance to them.
 

Bataga

Senior Member
Jan 2, 2012
2,191
932
Georgia
I will likely still use my note for music since it sounds perfect with the acid sound mod and my audio system. I liked what superman did with the color adjustments in the past but i read somewhere that the s4 has it built in for users to adjust to a more natural look.

Sent from my SAMSUNG-SGH-I717 using xda premium
 

Zahid Ali

Senior Member
Mar 12, 2013
170
23
32
Lahore
Here's some stuff from GSMArena

Audio quality
Both smartphones did greatly in our dedicated audio quality test. Last week when we reviewed the HTC One we found it to have the best audio output of any smartphone we have tested over the past few years, but the Samsung Galaxy S4 didn't go down without a fight.
In fact, we were shocked to find out that the Samsung flagship had a cleaner output than the HTC smartphone. While both are perfect when attached to an active external amplifier, the Galaxy S4 remained just as great when we plugged in a pair of headphones, while the HTC One took a hit to its stereo crosstalk.
Still, it's a difference that's very hard to detect without lab equipment, so for purposes practical the two are on par. What is not so hard to hear is the difference in volume levels - the HTC One is notably louder than the Samsung Galaxy S4 in both test scenarios.
And here go the results so you can see for yourselves.
TestFrequency responseNoise levelDynamic rangeTHDIMD + NoiseStereo crosstalk
HTC One+0.11, -0.20-82.782.70.00230.026-80.9
HTC One (headphones attached)+0.02, -0.32-82.482.30.00450.057-68.3
Samsung Galaxy S4+0.25, -0.16-82.582.40.00330.023-81.5
Samsung Galaxy S4 (headphones attached)+0.27, -0.15-82.382.20.00530.029-79.5

gsmarena_a001.png


gsmarena_a001.png


The HTC One is one of the best DACs they have ever tested. The SGS4 is close to comparable than it. They mentioned that there's more power in the HTC One. It might be a tad more crisp possibly due to an amplifier someone mentioned earlier. But they are still comparable. So the DACs still might be on par.

But note that Supercurio doesn't agree witht he way GSMArena performs their test.

Still a mystery to me.

Wolfsons chip is in octa version not in sd600. Gsm arena has tested SD600 version 19505

Sent from my LT30p using xda app-developers app
 

SlimJ87D

Senior Member
Jan 16, 2008
2,326
424
Wolfsons chip is in octa version not in sd600. Gsm arena has tested SD600 version 19505

Sent from my LT30p using xda app-developers app

I know. But I don't think you read my post thoroughly. GSMArena praised the SD600 audio in the HTC One and in a comparison they're pretty close with the HTC One being a bit louder than the SGS4. but the point is that it's pretty good.

Please reread my post.
 
  • Like
Reactions: Zahid Ali
H

hot_spare

Guest
I know. But I don't think you read my post thoroughly. GSMArena praised the SD600 audio in the HTC One and in a comparison they're pretty close with the HTC One being a bit louder than the SGS4. but the point is that it's pretty good.

Please reread my post.

Still makes no sense. Wolfson has no relation to the test conducted on 600. Whether the 600 DAC is good or bad has nothing to do with the Wolfson.
 
Status
Not open for further replies.

Top Liked Posts

  • There are no posts matching your filters.
  • 22
    I could get either one but I'm going with the SD600 because difference AREN'T huge.

    1. Antutu benchmarks are close
    2. Battery life is close.
    3. There is LTE everywhere I live now.
    4. I'll save $150 getting the SD600
    5. I've never done anything with y i9300 or even SGH Captivate that pushed the GPU to submission. If you are going for GPU then mmight as well get the Tegra 4 since it will have optimized exclusive games for it.
    6. Because there are dozens of more users now getting the SD600 and with Qualcomm releasing the proper source codes, development is going to BOOM on this platform.

    So really, the only thing I see with the 4+4 CPU is battery life being a minor advantage of maybe a few minutes.

    I think I'm going with the SD600 mostly because it's cheaper for me and LTE is everywhere in USA, CA.
    16
    and please stop calling it octa. it is no octa core cpu.
    these are two quad cpus which can't work together. only one of them can be active at the same time.
    and i'm not sure if it can run A15 optimized code like the snapdragon does because the little quad is an A7 which probably can't.
    that would be another point for the snapdragon.

    if i would get one, then the snapdragon variant.
    It is an 8 core CPU and all 8 cores are able to run simultaneously, it doesn't matter how the initial software is limited because the hardware is not, as much as people like pissing on the Octa name you don't go calling it a dual quad just because the software that's being shipped in in this implementation can't use it.

    A15 optimised code on Snapdragon? Sorry what? The Krait cores are nowhere even near an A15, so all those A15 compiler flags and optimizations aren't even able to be fully used on the Snapdragon versions. And on big.LITTLE, you target the compiler for the A15 cores since the A7 cores are completely ISA compatible, just severely shaved down.

    A little less misinformation from a respected developer here please...
    11
    It was my impression that the Snapdragon (or Krait really) architecture is actually derived from the ARM A15 reference? Except with whatever customizations Qualcomm has added/changed. Is that incorrect? And at any rate, even if Krait-300 isn't on par with the A15 reference, it's still gotta be pretty close. Granted, the Krait-400 is the real competitor to the A15, but so far, from what I've seen the A15 can't be more than 10-15% more efficient per clock over the current Krait.

    And yea, the A15's pipeline depth is greater, but from a number of other perspectives they seem similar. And again, the Krait-300 has gotta be a fair bit better than the Krait-200.

    Finally, as far as running all 8-cores simultaneously... Have you played with an Exynos version to say for sure? Can it stay within it's thermal envelop even with all 8 cores running at their full speed? Has anyone actually seen the kernel drivers Samsung is going to employ? And further, what about that "cache coherent interconnect"? Does it really have the bandwidth/speed to handle all 8 cores running at full tilt? And more interestingly, given that I've yet to see a picture of the actual die (guess will wait for Chipworks), how certain are you that they each given core in this first hardware implementation can actually be gated completely independently of it's companion core (one A7 to one A15)?
    Krait is not based on A15; period. People have been claiming this ever since the S4 (Krait v1) last year.

    Since now we actually have an A15 to compare them to, the performance difference is night and day. If the Octa is already 12% faster than the S600, with an 18% clock disadvantage, then it puts the IPC discrepancy at 25% between the Krait 300 and the A15. And please remember that we're talking about the Krait 300 here, which had huge performance improvements over the previous iterations like the Snapdragon S4's, so imagine how blown away those would be in a direct comparison.

    It's not about pipeline depth only, the A15 simply has more execution units, bigger and fatter branch prediction, and is simply more aggressive in most regards. Ironically, Krait 300's biggest upgrade was the TLB and other lookup table upgrades that the A9r4 also supposedly got retroactively upgraded from the A15 design.

    To claim or talk about Snapdragons using A15 optimized code-paths and compiler optimizations is inherently misinformed or jaded by Qualcomm marketing.

    There is an actual picture of the die, check my other thread on the topic. I already mention it there that the companion core is just a logical pair-up that has no direct physical representation. I don't really understand what you want to say about gating them independently, of course you can, as that's a basic requirement to even run them efficiently in core migration mode. Samsung already confirmed that the chip is able to run all possible running modes back in January; not that they needed to, ARM documented this quite some time ago.

    I wouldn't worry about the CCI, that's something designed to scale up much wider than "merely" 4 A15's and 4 A7's. You're hitting main memory bottlenecks much earlier than that even without running all cores.

    Power consumption is something I don't know about, but adding the A7's on top of the A15's running at full power is like pissing in the ocean and trying to measure the increased volume, it's going to be very minor in terms of power.

    Most quads nowadays enable all of their cores once in a full moon, it's irrelevant in terms of realistic power management. big.LITTLE and especially a many-core implementation like the Octa targets a very different aspect of the multi-core CPU paradigm.
    10
    Was Samsung said on the Unpacked what CPU will S4 have? No.
    Yes.
    Was Samsung said what frequency will the CPU clock will have? No.
    Yes.
    Was Samsung said that there will be a 3G and 4G version of S4? No, they said S4 will support 4G/LTE globally period. No more details.
    Yes.

    People started saying S4 Octa will be clocked at 1.6 GHz as that's what Samsung said. When Samsung said that?
    See above. It's all in the official press event statement.

    Would help out this forum tremendously if people would actually read and get their facts straight instead of doing blind claims.
    10
    How does that even make sense to you? Ignoring information from the chip fab side of samsung. the SoC supports LTE. Samsung Mobile has not said WHY theres a distinction, for you to perpetuate older information as though its gospel, when the ONLY official word we have is that the E5O supports the tech is counter-productive. it doesnt help people looking for updated information.
    Because LTE is not provided by the Exynos chip! It's provided by an external modem which has absolutely nothing to do with it. The S4 variant with the Exynos (i9500) will NOT have an LTE modem.

    Let's repeat: The chip supports LTE, it doesn't provide LTE, that's up to the modem that's attached to it.
    Very true, but it is just as misleading to say that the Octa version will be able to use the LTE networks when all that has been released is that the chip is "capable" of doing so. I think it is safe to say nobody knows at this point. So until the Octa version is either confirmed as using the LTE bands or confirmed as not using them maybe it would be a good idea for everyone to stop the speculation until there is proof. I've seen the same speculation in several S4 forum posts now and all it is doing is misleading people.
    No it's not misleading, what is misleading is that there has been an incredibly large amount of people the last few weeks who have been falsely accusing / coming to the conclusion and implying that the Exynos cannot support LTE even though LTE support has absolutely nothing to do with the SoC. Their tweet simply corrected this, they're correct in what they're claiming.

    It's a massive problem of misinformation and lack of education of both users side and all these incredibly gullible and incompetent armchair-reporter news sites who think they're doing some kind of journalism here.