Post Reply

Snapdragon 600 vs Snapdragon S4 Pro

OP fredcorp6

20th February 2013, 10:00 PM   |  #31  
richteralan's Avatar
Senior Member
Thanks Meter: 758
 
1,728 posts
Join Date:Joined: Aug 2010
More
Re: Snapdragon 600 vs Snapdragon S4 Pro
Quote:
Originally Posted by aimcr7

You dont need to compare N4 with HTC One, just compare it with OG, it almost doubled N4 score. Why? Because benchmark is useless.

And look at your device free ram it says 1.2GB or more if you have few apps installed while on One its only 800MB and no downloaded apps, its because HTC sense is resource hog.

If you want fast and smooth device in real life, use N4. You want benchmark optimized device go for HTC One or upcoming Galaxy SIV.

This device isn't optimized for benchmark, ask google why. Read my post in page one.

Sent from my Nexus 4 using Tapatalk 2

Blame Qualcomm for the benchmark differences between OG and N4.

Qualcomm has some close sourced blobs (reads: optimized drivers) which only allows LG to use on OG not N4.

Sent from my Nexus 4 using xda premium
20th February 2013, 10:08 PM   |  #32  
BarryH_GEG's Avatar
Senior Member
Flag Spokane, Washington
Thanks Meter: 4,282
 
8,822 posts
Join Date:Joined: Jan 2009
Quote:
Originally Posted by fredcorp6

So how does this new chip compare? Does anyone know what's been improved?

S4 Pro has two components, the CPU and GPU. The CPU is identical architecturally to the dual-core S4 with the obvious addition of two additional cores. S4 Pro was the launch SoC for Adreno 320. So it's markedly improved over S4 GPU wise but not as much so on the CPU side. This is what Anandtech said about S4 Pro when it was introduced. S4 Pro is "v2" and the reference he makes to "v3" is S-600. And the 10-15% performance boost assumes identical clock speeds. The 40% improvement in Qualcomm's PR is based on that plus S-600's higher clock speed(s).
At present, this is the same Krait CPU as what we've seen in MSM8960 in phones like the USA versions of the Galaxy S 3 and HTC One X. Later on, Krait v3 will emerge with higher IPC and shorter critical paths (and clocks up to 1.7 or 2 GHz) and a resulting 10-15% boost in performance. For now however we're looking at 1.5 GHz APQ8064 with a Krait v2 inside and Qualcomm's newest scalar GPU architecture with Adreno 320.
Krait 300 is the architecture used in S-600. Here are the benefits from Anandtech. They aren't trivial and improved battery life is one of the advantages. Qualcomm's also said Adreno 320 has been "speed enhanced" in the S-600 but hasn't provided any details. It's most likely been enhanced through additional memory bandwidth.

In usual Qualcomm fashion, we're missing good depth on exactly what these new revisions deliver. This is one area where Qualcomm really needs to emulate Intel: we know more about Haswell than we do about the original Krait.

That being said, here's what we do know. Krait 300 is still built on TSMC's 28nm LP process, just like the original Krait. The pipeline remains unchanged, but Qualcomm is able to squeeze out higher clocks out of the core. It's unclear whether we're simply talking about voltage scaling or a combination of that and improvements to timing, yields and layout. Whereas the current Krait core tops out at around 1.5GHz, Krait 300 will run at up to 1.9GHz.

Another big addition to the architecture is Krait 300 now features a hardware data prefetcher that preemptively grabs data out of main memory and brings it into L2 cache. The original Krait core had no L2 prefetchers.

Single threaded IPC improvements are the name of the game with Krait 300 and like all good evolutions to microprocessor architectures, the new Krait improves branch prediction accuracy. Since there's no increase to pipeline depth, improved branch prediction directly results in improved IPC (and better power efficiency).

Both Qualcomm and ARM have been very vague about what types of instructions can be executed out of order, but Krait 300 can execute more instructions out of their original program order. Building a robust OoOE (Out of Order Execution Engine) is very important to driving higher performance, and being able to reorder more types of instructions directly impacts single threaded performance.

Krait 300 now supports forwarding between pipelines, although it's not clear whether or not the previous architecture lacked any ability to forward data between stages.

Finally Krait 300 improves FP and JavaScript performance. Once again, it's not clear how. I've asked Qualcomm whether there have been any changes to the execution units in Krait 300 to enable these improvements. In general I believe we're looking at around a 15 percent increase in performance at the same clock frequency, for a jump of 20 to 30 pecent overall with the clock increases. This isn't necessarily enough to close the gap between Krait 300 and ARM's Cortex A15, however Krait 300's power profile should be much better. Compared to Atom, the Krait 300 improvements should be enough to at least equal performance if not surpass it, but not necessarily significantly.
The Following 10 Users Say Thank You to BarryH_GEG For This Useful Post: [ View ]
20th February 2013, 10:09 PM   |  #33  
aimcr7's Avatar
Senior Member
Flag Yogyakarta
Thanks Meter: 775
 
1,825 posts
Join Date:Joined: Jun 2010
More
Re: Snapdragon 600 vs Snapdragon S4 Pro
Quote:
Originally Posted by meangreenie

Has anyone checked the logs on quadrant? Maybe N4 is failing a test or some thing, because it doesn't stack up even against other s4pro's. It should score around 7500 stock and about 9000-10000 overclocked and tuned, imo.

Anyway with antutu it does score about right, so something is up between quadrant and n4.

We dont need it
Just dont use quadrant.





Quote:
Originally Posted by richteralan

Blame Qualcomm for the benchmark differences between OG and N4.

Qualcomm has some close sourced blobs (reads: optimized drivers) which only allows LG to use on OG not N4.

Sent from my Nexus 4 using xda premium

Thats my point, but who need benchmark? You like number? Use benchmark. You want a real life performance? Use your phone.

Sent from my Nexus 4 using Tapatalk 2
The Following 2 Users Say Thank You to aimcr7 For This Useful Post: [ View ]
20th February 2013, 10:22 PM   |  #34  
richteralan's Avatar
Senior Member
Thanks Meter: 758
 
1,728 posts
Join Date:Joined: Aug 2010
More
Re: Snapdragon 600 vs Snapdragon S4 Pro
Quote:
Originally Posted by aimcr7

We dont need it
Just dont use quadrant.







Thats my point, but who need benchmark? You like number? Use benchmark. You want a real life performance? Use your phone.

Sent from my Nexus 4 using Tapatalk 2

We still need benchmark. Or we won't be able to know what Qualcomm did with the closed source drivers.

The important thing is we never rely on one single benchmark to determine the performance.

Sent from my Nexus 4 using xda premium
The Following User Says Thank You to richteralan For This Useful Post: [ View ]
21st February 2013, 12:22 AM   |  #35  
wilsonlam97's Avatar
Senior Member
Flag Toronto
Thanks Meter: 126
 
686 posts
Join Date:Joined: Jul 2010
More
Quote:
Originally Posted by Exodian

Why do people care so much about benchmark scores? Does it really matter? The only test that should matter is your eyeball.

Sent from my Nexus 4 using xda premium

This is like driving a small Toyota and telling me I'm an idiot for saying it's not quite fast enough even though I am being clocked by a radar to be at the speed of 120 km/h. But my neighbour Dave gets 250 km/h on his Ferrari easily. So it's no problem that I am much slower? Or should I just be in denial like you?

Yes my small Toyota vehicle is just as good as the Ferrari. Sure lets optimize the Toyota.
Last edited by wilsonlam97; 21st February 2013 at 12:28 AM.
21st February 2013, 01:35 AM   |  #36  
aimcr7's Avatar
Senior Member
Flag Yogyakarta
Thanks Meter: 775
 
1,825 posts
Join Date:Joined: Jun 2010
More
Re: Snapdragon 600 vs Snapdragon S4 Pro
Quote:
Originally Posted by wilsonlam97

This is like driving a small Toyota and telling me I'm an idiot for saying it's not quite fast enough even though I am being clocked by a radar to be at the speed of 120 km/h. But my neighbour Dave gets 250 km/h on his Ferrari easily. So it's no problem that I am much slower? Or should I just be in denial like you?

Yes my small Toyota vehicle is just as good as the Ferrari. Sure lets optimize the Toyota.

Benchmarking is not that simple, we have a bunch of code, we have software difference.
Even the same phone with the same apps launched at the same time in the same place will produce different result, thats why benchmarking is useless, unless you use it in gaming and you see fps in that game and compare it with another phone, it makes more sense for me. :beer:


Sent from my Nexus 4 using Tapatalk 2
21st February 2013, 01:58 AM   |  #37  
Member
Thanks Meter: 8
 
69 posts
Join Date:Joined: Mar 2010
More
Quote:
Originally Posted by Exodian

Why do people care so much about benchmark scores? Does it really matter? The only test that should matter is your eyeball.

Sent from my Nexus 4 using xda premium

For the most part I agree with ya, but I think where benchmarks come in handy is when comparison shopping for a phone(s), especially when you may not have access to an actual device until you've paid for it and it's delivered. Which I'm sure was the case for most of us that have Nexus 4's.

The benchmarks were, for me, I good way to gauge the performance of the N4 against similarly specs devices.

What is absolutely ridiculous is when someone/magazine/blog compares the benchmark results of devices running different OS's- i.e. and Android flagship phone vs. iPhone 5. Make zero sense...
21st February 2013, 02:53 AM   |  #38  
richteralan's Avatar
Senior Member
Thanks Meter: 758
 
1,728 posts
Join Date:Joined: Aug 2010
More
Quote:
Originally Posted by aimcr7

Even the same phone with the same apps launched at the same time in the same place will produce different result,

Sent from my Nexus 4 using Tapatalk 2

Got anything to back that up?
21st February 2013, 03:28 AM   |  #39  
Senior Member
Flag Holyoke
Thanks Meter: 165
 
887 posts
Join Date:Joined: Jun 2011
More
Re: Snapdragon 600 vs Snapdragon S4 Pro
Have you ever run benchmark test back-to-back? Is the score ALWAYS the same?

Sent from my Nexus 4 using xda app-developers app
21st February 2013, 03:32 AM   |  #40  
aimcr7's Avatar
Senior Member
Flag Yogyakarta
Thanks Meter: 775
 
1,825 posts
Join Date:Joined: Jun 2010
More
Re: Snapdragon 600 vs Snapdragon S4 Pro
Quote:
Originally Posted by richteralan

Got anything to back that up?

Try your phone with your friend phone running at the same time and the same installed apps. And in quadrant first and second launch will produce way different result.

Sent from my Nexus 4 using Tapatalk 2

Post Reply Subscribe to Thread
Previous Thread Next Thread
Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes