Snapdragon S4 Pro: Unleashing the beast
0. phoneArena 06 Sep 2012, 01:40 posted on
We've had a lot of movement in the mobile chipset space in the recent years, and by the looks of it, the best is yet to come. With the proliferation of smartphones and tablets, mobile processor makers have been motivated to innovate and deliver faster and more capable chips to market, enabling device manufacturers to create faster and smarter products...
This is a discussion for a news. To read the whole news, click here
1. neutralguy (Posts: 1152; Member since: 30 Apr 2012)
One word for that Quadrant result. "WOW"
2. cheetah2k (Posts: 1576; Member since: 16 Jan 2011)
Meh @ Quadrant.. it seriously doesn't measure real world performance
4. neutralguy (Posts: 1152; Member since: 30 Apr 2012)
true, all of this doesn't show a real world performance. Any benchmark testing doesn't show a real world performance. But it's a good way to test and compare a device to other devices.
26. TylerGrunter (Posts: 1522; Member since: 16 Feb 2012)
And are you aware that any result between 55 and 65 can be ignored in Nenamark?
On those results the only thing that shows is that the GPU is too poweful for the test (or the test too weak for the GPU) and therefore you get limited to 60 FPS.
38. remixfa (Posts: 14255; Member since: 19 Dec 2008)
if you run the tests multiple times your results will also edge up. As the test stays in memory it loads faster, thus increasing the score.
Again, never trust a non-commercial benchmark. They are meant to be as good as they can be, and not indicative of real world conditions.
9. aaronkatrini (Posts: 241; Member since: 06 Jun 2012)
They should put an warning:
"It might kill your kitten, destroy your house and start the mayan apocalypse.
Use it at your own risk. Have Fun! :)"
34. SuperAndroidEvo (Posts: 4833; Member since: 15 Apr 2011)
lol I like that. It's that powerful that it might develop its own AI & try to take over the world one step at a time. lol
I have always said this & I will say it one more time, Qualcomm is the or should I say was the sleeping giant. They were the only game in town for a LONG while, but then came Samsung & Texas Instruments to steal some of its thunder.
This is innovation through competition at its finest. BOTH Samsung & Texas Instruments came VERY strong & took some of the limelight away from Qualcomm, with their respective Exynos & OMAP chips. Well that was the rude awakening Qualcomm needed.
Now because of that we WILL see the quad-core Krait S4 Pro that can be clocked to an insane 1.7GHz with the shockingly POWERFUL Adreno 320 before DEC. of 2012.
What is EVEN more scary to me is that when finally the Exynos 5 & the OMAP 5 come out (mind you first their dual-core, who knows when the quad-cores will come out), Qualcomm will have something EVEN better than they have now with the quad-core Krait S4 Pro.
Qualcomm is so far ahead of their competition that it looks like 2013 will be dominated by Qualcomm just like 2012 was & how it was before when Qualcomm was king.
Qualcomm has done its homework & there are now enjoying the fruits of their labor. Congrats Qualcomm you have regained your title of best & most dominant SoC in the world.
22. Mittal (Posts: 493; Member since: 14 Dec 2011)
I only wished they improve upon the device naming...
Optimus G? Experia T,V,J ?
But yes somehow One X,S, Galaxy S does not sound bad!!
5. Luuthian (Posts: 246; Member since: 09 Sep 2011)
I want it... Just not in an LG phone :(. Their OS is too far removed from stock. I'd much rather pick up a Moto device simply because it runs so much closer (even the Galaxy series is too far removed for my tastes).
This is all rather moot though. I've never seen phone with as much community support of flexibility as the Galaxy Nexus. I'm benchwarming until I see a new Nexus device. All this power is worthless if the phone itself isn't desirable.
23. Mittal (Posts: 493; Member since: 14 Dec 2011)
Mayb but LG 4X HD is the quad core closest to stock compared to OneX or SIII
6. bobfreking55 (Posts: 866; Member since: 15 Jul 2011)
i wonder what a quad core ARM 15 exynos would feel like.
27. juanyunis (Posts: 73; Member since: 31 Aug 2012)
Well, based on this the difference between Exynos Quad Core Cortex A9 and S4 Pro is not that big, being the Exynos an old technology, so i guess when Samsung unveils the new Dual Core A15 Exynos will blow the S4 Pro, being just a dual core. So imaging a A15 Quad Core Exynos with the most powerful Mali GPU, it could power a good video game console.
7. aaronkatrini (Posts: 241; Member since: 06 Jun 2012)
They should put an warning:
"It might kill your kitten, destroy your house and start the mayan apocalypse.
Use it at your own risk. Have Fun! :)"
8. bayhuy (Posts: 320; Member since: 23 Jun 2011)
I'm constantly scoring about 5800-6000 in Quadrant with my Tegra 3 based One X. Antutu is about 12000-12500 ...
12. XPERIA-KNIGHT (unregistered)
you must have the sense 4.1 update with android 4.0.4......right?
10. J1919 (Posts: 32; Member since: 28 Aug 2012)
This reminds me... Why has LG gone so silent since unveiling the OG?
13. XPERIA-KNIGHT (unregistered)
put this on an xperia phone or one of motos razr phones......and im completely sold!
50. buccob (Posts: 2564; Member since: 19 Jun 2012)
I might be one of the only ones waiting this on a Asus PadFone 2... :-P
14. baldilocks (Posts: 743; Member since: 14 Dec 2008)
Hmmm, my stock Tegra 3 One X gets well over 6,400 in Quadrant. You guys need to do your articles and test equipment with the LATEST hardware/software versions.
Your articles become less valid when you use old numbers.
15. XPERIA-KNIGHT (unregistered)
yea dude but how in the hell are you gettin 6,400 with a freaking tegra 3 on the one x is the question??.........you must be doing some serious modifying!
16. darkskoliro (Posts: 1076; Member since: 07 May 2012)
pics or gtfo for that one, but you are right when you say that they need to update the numbers :L Making one x look bad when tegra 3 is quite powerful..
40. rizevnarastek (Posts: 347; Member since: 06 Sep 2012)
Here's a link for the benchmark where in surpasses 6000 on quadrant... no GTFOuting here :)
41. remixfa (Posts: 14255; Member since: 19 Dec 2008)
aaaand why I dont trust quadrant.
See the big increase in green? What test is that? The I/O test. The easiest faked part of the whole test... :(
That just means that HTC is taking a page from Sammy and Sony and buffing the I/O scores..
Also, why does it compare it to a bunch of single core and older dual core models? The only current chip device in that whole mix is the 3 Tegra3 chip based devices at the top.
55. darkskoliro (Posts: 1076; Member since: 07 May 2012)
I know about 6000+ but not 6400! haha
and plus remixfa is an idiot -.- The benchmark compares it to a few set devices, if S3 was doing this benchmark too it would also be compared to those models.
57. remixfa (Posts: 14255; Member since: 19 Dec 2008)
if your going to call someone an idiot, be prepared to back it up with a fact or 2. You have none.
59. darkskoliro (Posts: 1076; Member since: 07 May 2012)
I actually did. But since you cant read, here we go again.
"Also, why does it compare it to a bunch of single core and older dual core models?"
Those models on quadrant are set, there is no "bias" when it is compared to those devices. Every phone that does quadrant are compared with them. Its obvious from your other comments that you are just a hater, so im going to leave it at that. Pointless talking to a brick wall.
17. darac (Posts: 2156; Member since: 17 Oct 2011)
Xperia T and V with dual core clocked lower, and with inferior GPU get 6000 on quadrant and 60 on nenamark.
Now I wonder what Sony could do with this chip..
18. XPERIA-KNIGHT (unregistered)
thanks bro i didnt wanna have to keep breaking it too em that the xperia line up is superior to every phone as of right now but hey its true.......and yea im with you in wanting to see what this chip could score on an xperia phone......... sony ftw again
29. remixfa (Posts: 14255; Member since: 19 Dec 2008)
I'd love to see the S3 running XperiaT get 60fps in NENAMARK2 and 6000 on quadrant out of the box. Please show me proof of that. Your smoking that craziness again. My SGS1 gets 50+FPS in Nenamark1.. that test was made to test the S2, Hummingbird, and other older single core chips.
33. darac (Posts: 2156; Member since: 17 Oct 2011)
Remixfa, you're a great guy..but please stop embarrassing your self about snapdragon chips.
35. darac (Posts: 2156; Member since: 17 Oct 2011)
In case I need to clarify - I was talking about the dual core s4.
You can also find a Note 2 preview with benchmarks(compared to Xperias) there.
36. remixfa (Posts: 14255; Member since: 19 Dec 2008)
Now, that is my faux pas. I saw "xperia" and i defaulted to the S3 configuration, as you notice in my post.
Now, what is the first rule of benchmarks that I keep reminding peeps of?
Never trust a pre-release benchmark. Only ones that come from commercially released sets that can be benched and confirmed by multiple sources. Those benchmarks are not from a commercial device.
Also, benchmarks from the SGS3 that use the exact same chip do not come near those results. There is a noticeable difference between an Exynos SGS3 and the S4 equipped SGS3. If you have 5 phones with the same chip that perform close to each other, and then a 6th phone comes a long with the same chip configuration but blows right past them, then more than likely something is being faked.
That's not a fanboy statement. If you remember, I flamed the living hell out of samsung for messing with the i/o & integer scores on the Crapdragon S3 equipped SGS2, which were WAY higher than everyone elses... including their own Exynos based SGS2. I was the first to notice, and I lead the charge to show it to everyone else.
I knocked sony for the same thing with the experia.
Benchmarks are easily faked. Just like statistics.. if you get a result that is way too high or too low for the norm, its normally tossed as an anomaly.
But we can review this phone when actual commercial benchmarks come out.
42. remixfa (Posts: 14255; Member since: 19 Dec 2008)
also, those scores are over all scores. There is no way with those charts to decern I/O manipulation by comparing it to other S4- adreno 225 models who have not gotten nearly as high as a score.
Think companies dont do it? Look at the benchmarks that Rize just posted for the OneX.
The only difference in the "update" from HTC is a MUCH bigger I/O score.. faaaaaked by HTC.. tisk tisk.
Samsung and Sony were both busted with that on the S3 crapdragon chips. No reason to assume they stopped.
19. Bluesky02 (Posts: 1439; Member since: 05 Dec 2011)
Let's wait for remixfa to bash Qualcomm
21. Bfrenz (Posts: 112; Member since: 26 Aug 2012)
We need Exynos in our GS3. Me barely using LTE. What's ur opinion guys?
49. remixfa (Posts: 14255; Member since: 19 Dec 2008)
I'll take exynos and 21mb/s HSPA+ over S4 crapdragon with Adreno 225 & LTE/HSPA42 any day of the week.
25. Suts_97 (Posts: 124; Member since: 05 May 2012)
Well I prefer Exynos over Qualcomm Snapdragon anytime, there's something special in the Samsung beast that still makes it my favourite over any other chips...And I guess the Exynos 5250 will blow this one away!
28. juanyunis (Posts: 73; Member since: 31 Aug 2012)
Agree, even the dual core exynos A15 with latest mali will blow this one easily.
30. remixfa (Posts: 14255; Member since: 19 Dec 2008)
any true A15 will knock around the Krait. Big difference in chip capabilities
43. d_eagle (Posts: 15; Member since: 06 Sep 2012)
I prefer Qualcomm's chipsets instead, even if they are slightly inferior, because it is one of the most developer-friendly chipmaker. Thanks to Qualcomm my old LG-P500 with armv6 msm7227 can run Jelly Bean (CM10) without any big problem, because they released drivers even for their oldest platform. Other platforms like LG Optimus Black's omap are not so lucky. If my phone didn't use a Qualcomm chipset, it would be stuck at Gingerbread...
44. remixfa (Posts: 14255; Member since: 19 Dec 2008)
Exynos based samsung phones are the most mod friendly android phones on the planet over all. No one has a closer relationship with XDA and cyanogen than Samsung.. especially since Samsung actually EMPLOYS Steve K..AKA Cyanogen himself.
Qualcomm definitely has nice open policies as well, dont get me wrong, but they are not the "most" mod friendly.
The worst is actually Nvidia who demands that companies shell out big bux to get the drivers for the latest android versions which is why very few Tegra2 based devices got the upgrade or will ever get it.
45. d_eagle (Posts: 15; Member since: 06 Sep 2012)
Thanks, I knew Steve Kondik was working for Samsung but i didn't know Exynos was so "open" (i've never had a Samsung phone). Anyway I think I must still go with LG Optimus G, unfortunately I need a new smartphone before christmas and probably the first Exynos 5 smartphone won't see the light before next April-May... :(
47. remixfa (Posts: 14255; Member since: 19 Dec 2008)
such is true. The best hope would have been a small chance of it being in the Note 2. That ship has obviously sailed. The S4 pro quad will be more than enough power for a while.
32. remixfa (Posts: 14255; Member since: 19 Dec 2008)
The scores reference Qualcomm's "perfect conditions" reference platform VS real phone performance. Never trust a pre-commercial release benchmark. They are always lower when they get thrown into commercial devices.
39. rizevnarastek (Posts: 347; Member since: 06 Sep 2012)
Please run One X's benchmark tests agian... after the recent update the device goes out to approx. 6000 on quadrant ;)
46. darac (Posts: 2156; Member since: 17 Oct 2011)
Remixfa, we've been through this before, and yeah we know your argument - the I/O score is fluffed and fake.
The problem is, you are no expert and since I haven't seen actual experts sharing your opinion..we should trust you over just about everyone else?
Everyone can try benchmarking and see how much variation a single phone can have, without any change in the software. Software and the level of it's optimization plays a huge role here, that's a fact.
Also, by your logic, why wouldn't Samsung tweak the international S3 in the same way?
48. remixfa (Posts: 14255; Member since: 19 Dec 2008)
actually, there are lots of "experts" that share opinions about fluffed quadrant numbers and score cheating. Do you think I just magically pulled it out of my ass? I learned it from others while I was learning to mod my phone back when android first came out.
Go do some research on it. I'll give ya some search points. I/O fluffing/faking, Linpack cheating, quadrant cheating, benchmark faking, on and on.
You dont HAVE to trust me over anyone. Do your research. Many people do tend to trust my opinion because its nearly always backed up by facts, links, ect. I am no expert, but I am pretty educated on the matters that I espouse. I am not the programmer that does it, I am more of a trusted news reporter than you can generally count on to be accurate (find that in the mainstream media lol).
And a quick and easy example is provided in THIS very thread. Check out that OneX scoring 6000 points on quadrant. Look at the oneX score just below it. Can you tell me what the ONLY real difference between the lower and higher score is?
I'll give ya a hint. It's 2 letters, seperated by a / .... and they are both vowels. :)
51. rizevnarastek (Posts: 347; Member since: 06 Sep 2012)
Everything is slightly higher but I agree, I/O has the most difference. Again to say that benchamarks are kinda useless... it just shows the OneX was and is better then GS3 ;)
52. remixfa (Posts: 14255; Member since: 19 Dec 2008)
lol. you can dream that all you want. the T3 chip has nothing on the 4412 chip.
54. darac (Posts: 2156; Member since: 17 Oct 2011)
Okay, I have no problem with that being true.
Actually I'd like it to be so since it makes more sense.
As for your research credentials, may I just remind you
that you were the author of statement that the s4 quad will be out in "Q2-Q4 2013".
And on top of it, you still refute to admit what I was saying all along - that chip will be the best one released in phones this year.
A little less bias in your interests would give you way more credit.
56. remixfa (Posts: 14255; Member since: 19 Dec 2008)
I dont always respond to you because you get offensive. However, I had that conversation with SAE a while ago.
It still makes no sense that they would release phones with the pro dual and pro quad side by side. My timetable guess was going by normal business sense of about 6 months between chip releases... which is standard fare. Much like all your hopes and dreams of it coming earlier before the LG/Miezu announcements were all guessing games.
Its obviously releasing this year. And yes, since there will be no E5 for it to compete with, of course it will be the best chip released in 2012. However, if it releases in late nov/dec, I'd really call it a 2013 chip because very few of its actual sales will be in 2012. We did the same maneuver with the T3 chip which was released in 2011 yet we kept calling it a 2012 chip.
I dont have a "bias". And again, if you would go back and read my conversations with SAE, I dont have any issues with the Pro dual or quad. The adreno 320 is more than enough for everyone. Still not as good as a true A15 chip, but its got plenty of horsepower.
No one is right all the time. But I am right way more often than I am wrong.
58. wizzardtech (Posts: 59; Member since: 12 Mar 2012)
We cant change your mind set, we all knew since then that your a fan of
exynos and samsung. You hate ifans just like your no1 rival but you yourself is a fanboy too. Why dont you give some companies a chance like qualcomm. If they have something that is great and benefit the consumer then we will be happy with it because it drives INNOVATION. We are here to thank and support them, not to be judged. We are all be thankful of their technology because it drives competition and innovation and we the consumer are the no1 benefeciary.
60. tehkirby (Posts: 1; Member since: 22 Sep 2012)
My Galaxy S3 has very similar 2D and 3D graphics scores; everything else pretty much is lower of course. These type of benchmarks aren't the best. I know other benchmarks that use FPS to measure graphics performance that put the Adreno 320 on the S4 Pro higher scores than the Galaxy S3. The CPU benchmarks on quadrant are more accurate than the graphics benchmarking.