Faster Throughput on WCDMA

Fixing unintended attenuation is only one part of what’s new however, the other part of the story is faster cellular connectivity for users on WCDMA/UMTS carriers. Users who are using the 4S on CDMA (like Sprint or Verizon) won’t see a performance difference since this is still the same EVDO Rev.A.

The iPhone 4 used an Intel/Infineon X-Gold 618 which supported HSDPA 7.2 and HSUPA 5.76. The MDM6610 inside the 4S supports HSDPA 14.4 and HSUPA 5.76, alongside a number of 3GPP Rel.7 features which are colloquially known as HSPA+. I talked about this extensively in another piece when there was some confusion about whether or not the 4S is HSPA+ - which it is.

iPhone Cellular Speeds
Property iPhone 3G/3GS iPhone 4 (GSM/UMTS) iPhone 4 (CDMA) iPhone 4S
Baseband Infineon X-Gold 608 Infineon X-Gold 618 Qualcomm MDM6600 Qualcomm MDM6610
HSDPA Cat.8 - 7.2 Mbps Cat.8 - 7.2 Mbps N/A Cat.10 - 14.4 Mbps
HSUPA None - 384 Kbps WCDMA only Cat.6 - 5.76 Mbps N/A Cat.6 - 5.76 Mbps
EVDO N/A N/A 1x/EVDO Rev.A 1x/EVDO Rev.A

The previous X-Gold 618 baseband was a nice improvement over the iPhone 3G/3GS’ X-Gold 608, which lacked HSUPA, but in a world where most WCDMA carriers are at least running HSDPA 14.4, it’s nice to finally have an iPhone with something faster than HSDPA 7.2. I’ve done lots of testing inside my Tucson, AZ market (which is “4G” HSPA+ on AT&T’s coverage viewer) with both the 4 and the 4S, and have built a very good feel for the 4’s performance. As a reminder, if you’re in the USA, those dark blue areas represent HSPA+ coverage areas with AT&T’s upgraded backhaul. In practice these are at least HSDPA 14.4.

 
Left: iPhone 4 Limited to ~6.1 Mbps down, Right: iPhone 4S (same location) hitting ~9 Mbps

With line of sight to an AT&T NodeB inside my HSPA+ market I’m used to seeing a maximum downstream throughput on the iPhone 4 of almost exactly ~6.1 Mbps, which is about right for the 4’s HSDPA 7.2 maximum when you include overhead. The nice straight line in that result should clue you in that downstream throughput on the 4 was being gated by the baseband. On the 4S, in this same location, I’ve been able to get 9.9 Mbps when the cell isn’t loaded at night (I didn't grab a screenshot of that one, for some reason). It’s nice to finally not be gated by the baseband anymore on an iDevice. Having a faster baseband is part of the reason the 4S’s cellular performance is much better, the other half is receive diversity which helps the 4S push these high throughput rates, and also dramatically improve performance at cell edge.

I did some drive testing with the 4 and 4S side by side and targeted areas that I know have pretty poor signal strength. The 4S is shown in yellow, the 4 in blue.

You can see how downstream throughput gets a nice shift up, and the average changes as well, from 2.28 Mbps on the 4 to 2.72 Mbps on the 4S. The maximum in this sample increases from 6.25 to 7.62 Mbps as well. It isn’t a huge shift, but subjectively I’ve noticed the 4S going a lot faster in areas that previously were difficult for the 4.

We’ve also run the usual set of standalone tests on the 4S on AT&T in my market of Tucson, AZ, in Anand’s market of Raleigh, NC, and on Verizon in Raleigh, NC. Though we don’t have a Sprint 4S yet, we hope to do a more serious 4S carrier comparison here in the US when we get one. First up is AT&T which is of course HSPA+ in both of our testing markets.

AT&T HSPA+

Verizon EVDO

iPhone 4S Speedtest Comparison
Carrier AT&T Verizon
  Avg Max Min Avg Max Min
Downstream (Mbps) 3.53 9.94 0.24 0.82 2.05 0.07
Upstream (Mbps) 1.17 1.86 0.009 0.38 0.96 0.003
Latency (ms) 137 784 95 177 1383 104
Total Tests 457 150
Air Interface HSPA+
(HSDPA 14.4/HSUPA 5.76)
EVDO Rev.A

For the CDMA carriers, the 4S shouldn’t (and doesn’t) bring any huge improvement to data throughput because the CDMA 4 had both receive diversity and MDM66x0. For users on GSM/UMTS, however, the 4S does make a difference again thanks to the inclusion of those two new features.

One of the things I noticed was absent on the CDMA iPhone 4 was the 3G toggle. It does indeed make some sense to not include this in a CDMA 1x/EVDO scenario since power draw is about the same between the two air interfaces, however, the absence of this toggle has carried over to the 4S regardless of whether the phone is activated on a CDMA2000 or UMTS/GSM network. That’s right, you can go under Settings -> General -> Network, and there’s no longer any 3G Data toggle which you can disable and fall onto EDGE (2G) with now.

 
Left: iPhone 4S (no 3G toggle), Right: iPhone 4 (3G toggle)

It’s likely that this is absent to accommodate the multi-mode nature of the 4S (and thus the lowest common denominator CDMA mode), however the absence of this toggle makes getting connected in congested areas more difficult. In some markets, (I’m looking at you, AT&T in Las Vegas), EDGE is often the only way to get any connectivity, even without a major convention going on. Not having that 3G toggle makes manually selecting that less-used but more reliable connection impossible now, to say nothing of the potential battery savings that this would afford (and that we sadly can’t test now).

There’s one last tangential question about HSPA+ on the 4S, specifically on AT&T. I’ve left this to the end since it doesn’t impact non-US 4S users, but the last question is whether the 4S is actually on HSPA+. For a while, I was concerned that AT&T would continue using the wap.cingular APN on the 4S which seems shaped to around 7.2 Mbps HSDPA. I’m glad to report that AT&T hasn’t continued using wap.cingular on its 4S data plans, instead using “phone” which is a newer APN that allows for HSPA+ (above 7.2 Mbps) rates. You can check this yourself under PDP Context Info on the 4S in field test.

Improved Baseband - No Deathgrip The A5 Architecture & CPU Performance
POST A COMMENT

201 Comments

View All Comments

  • doobydoo - Friday, December 02, 2011 - link

    Its still absolute nonsense to claim that the iPhone 4S can only use '2x' the power when it has available power of 7x.

    Not only does the iPhone 4s support wireless streaming to TV's, making performance very important, there are also games ALREADY out which require this kind of GPU in order to run fast on the superior resolution of the iPhone 4S.

    Not only that, but you failed to take into account the typical life-cycle of iPhones - this phone has to be capable of performing well for around a year.

    The bottom line is that Apple really got one over all Android manufacturers with the GPU in the iPhone 4S - it's the best there is, in any phone, full stop. Trying to turn that into a criticism is outrageous.
    Reply
  • PeteH - Tuesday, November 01, 2011 - link

    Actually it is about the architecture. How GPU performance scales with size is in large part dictated by the GPU architecture, and Imagination's architecture scales better than the other solutions. Reply
  • loganin - Tuesday, November 01, 2011 - link

    And I showed it above Apple's chip isn't larger than Samsung's. Reply
  • PeteH - Tuesday, November 01, 2011 - link

    But chip size isn't relevant, only GPU size is.

    All I'm pointing out is that not all GPU architectures scale equivalently with size.
    Reply
  • loganin - Tuesday, November 01, 2011 - link

    But you're comparing two different architectures here, not two carrying the same architecture so the scalability doesn't really matter. Also is Samsung's GPU significantly smaller than A5's?

    Now we've discussed back and forth about nothing, you can see the problem with Lucian's argument. It was simply an attempt to make Apple look bad and the technical correctness didn't really matter.
    Reply
  • PeteH - Tuesday, November 01, 2011 - link

    What I'm saying is that Lucian's assertion, that the A5's GPU is faster because it's bigger, ignores the fact that not all GPU architectures scale the same way with size. A GPU of the same size but with a different architecture would have worse performance because of this.

    Put simply architecture matters. You can't just throw silicon at a performance problem to fix it.
    Reply
  • metafor - Tuesday, November 01, 2011 - link

    Well, you can. But it might be more efficient not to. At least with GPU's, putting two in there will pretty much double your performance on GPU-limited tasks.

    This is true of desktops (SLI) as well as mobile.

    Certain architectures are more area-efficient. But the point is, if all you care about is performance and can eat the die-area, you can just shove another GPU in there.

    The same can't be said of CPU tasks, for example.
    Reply
  • PeteH - Tuesday, November 01, 2011 - link

    I should have been clearer. You can always throw area at the problem, but the architecture dictates how much area is needed to add the desired performance, even on GPUs.

    Compare the GeForce and the SGX architectures. The GeForce provides an equal number of vertex and pixel shader cores, and thus can only achieve theoretical maximum performance if it gets an even mix of vertex and pixel shader operations. The SGX on the other hand provides general purpose cores that work can do either vertex or pixel shader operations.

    This means that as the SGX adds cores it's performance scales linearly under all scenarios, while the GeForce (which adds a vertex and a pixel shader core as a pair) gains only half the benefit under some conditions. Put simply, if a GeForce core is limited by the number of pixel shader cores available, the addition of a vertex shader core adds no benefit.

    Throwing enough core pairs onto silicon will give you the performance you need, but not as efficiently as general purpose cores would. Of course a general purpose core architecture will be bigger, but that's a separate discussion.
    Reply
  • metafor - Tuesday, November 01, 2011 - link

    I think you need to check your math. If you double the number of cores in a Geforce, you'll still gain 2x the relative performance.

    Double is a multiplier, not an adder.

    If a task was vertex-shader bound before, doubling the number of vertex-shaders (which comes with doubling the number of cores) will improve performance by 100%.

    Of course, in the case of 543MP2, we're not just talking about doubling computational cores.

    It's literally 2 GPU's (I don't think much is shared, maybe the various caches).

    Think SLI but on silicon.

    If you put 2 Geforce GPU's on a single die, the effect will be the same: double the performance for double the area.

    Architecture dictates the perf/GPU. That doesn't mean you can't simply double it at any time to get double the performance.
    Reply
  • PeteH - Tuesday, November 01, 2011 - link

    But I'm not talking about relative performance, I'm talking about performance per unit area added. When bound by one operation adding a core that supports a different operation is wasted space.

    So yes, doubling space always doubles relative performance, but adding 20 square millimeters means different things to the performance of different architectures.
    Reply

Log in

Don't have an account? Sign up now