Apple iPad 2 Preview
by Anand Lal Shimpi, Brian Klug & Vivek Gowri on March 12, 2011 6:01 AM ESTWiFi + AT&T 3G
The iPad 2 uses what boils down to same exact same 802.11a/b/g/n WiFi + Bluetooth 2.1 EDR solution as its predecessor, the BCM4329. The iPad 2 maintains full 5 GHz (802.11a/n) support, just like the previous iPad, and also negotiates a transmit rate of 65 megabits/s just like nearly all single stream 802.11n mobile devices we've seen thus far. Close to the AP and within good coverage radius, the experience is exactly the same as its predecessor. Best case throughput is virtually identical between the two generations.
The black plastic strip at the top still is an RF window for cellular connectivity. The strip is black even if your iPad is white, which is a bit puzzling.
That said, thus far we've noticed that wireless range on the iPad 2 is measurably less than the iPad 1. It's not a terribly huge difference, but definitely noticeable as you drop one or two bars of WiFi signal strength. Throughput goes way down compared to the previous generation. We're still testing everything, but it's just not quite as good. No doubt the iPad 2 has a completely different RF design than the original iPad which accounts for this difference.
Left: iPad 1 across the street, Right: iPad 2 same location
We've got both an AT&T, Verizon, and WiFi model between the three of us (Anand, Vivek, and myself) and will do the usual due diligence testing cellular connectivity.
Following the rather controversial and inexplicable crippling of HSUPA on the Motorola Atrix and Inspire 4G, the first thing I did with the iPad 2 is run a speedtest. HSUPA is definitely enabled on this device and working just fine. Throughput is where it should be in the best and worst signal cases so far, topping out at around 6 Mbps down, 1.5 Mbps up at my house within line of sight to an AT&T Node B. Apparently AT&T is inexplicably only concerned with Android devices crippling the network with upstream traffic.
AT&T iPad 2 Cellular Data Throughput
Anand did some initial testing on the Verizon iPad 2 and averaged 1.54 Mbps down, 0.72 Mbps up in testing with good signal strength, which seems about average. Max speeds we've seen on Verizon so far are 1.64 Mbps down and 0.82 Mbps up.
| iPad 2 - Network Support | |||||
| Verizon Version - CDMA2000-1xEV-DO | 800 / 1900 MHz | ||||
| AT&T Version - Quad-Band UMTS + HSDPA/HSUPA | 850 / 900 / 1900 / 2100 MHz | ||||
| AT&T Version - Quad-Band GSM/EDGE | 850 / 900 / 1800 / 1900 MHz | ||||
| Baseband Hardware | Unknown - Possibly MDM6x00 | ||||
Nobody has disassembled an iPad 2 with 3G connectivity from either Verizon or AT&T, but it seems likely that the two would both leverage a Qualcomm MDM6x00 baseband (where x is 6 for CDMA2000/EV-DO + UMTS/GSM support, 2 for only UMTS/GSM) considering its use in the Verizon iPhone 4.
MicroSIM peeking out of the side on the AT&T version. Also yes, it does come with a SIM ejector tool.
Interestingly enough the AT&T model with FCC ID BCG1396 has UMTS/HSDPA/HSUPA support for quad-band UMTS and GSM. The Verizon FCC ID BCG1397 has reports for just cellular and PCS - 850 and 1900 again but CDMA nomenclature.
Facebook
Twitter
RSS
82 Comments
View All Comments
rom0n - Saturday, March 12, 2011 - link
The reason why Ipad webloading tests may be better is that it is missing flash. It is not rendering the full web page.Unless anandtech is using initial Xoom firmware without flash. The new android 3.0.1 brings Flash 10.2 as well as other performance and stability improvements.
tipoo - Saturday, March 12, 2011 - link
I was going to bring that up too. If the Xoom has Flash enabled, its numbers are more impressive.Anand Lal Shimpi - Saturday, March 12, 2011 - link
The test is run apples-to-apples, there's no Flash enabled :)Take care,
Anand
VivekGowri - Saturday, March 12, 2011 - link
Apples to Androids?arthur449 - Saturday, March 12, 2011 - link
zing!iwod - Saturday, March 12, 2011 - link
I really hope the A9 is true. Since there are reports they are Dual Core A8. ( Yeah i know ARM doesn't have that basic design ). But if we get A9 now, there is every chance the Apple A6 will be A15 plus Power VR 6. ( The same as Sony Ericsson ) on there next iPhone / iPad.But why apple only said 2x CPU Power when clearly Dual Core 1Ghz A9 should be at least 2 - 3 times faster. And test results doesn't always show the IPC improvement in A9.
winterspan - Saturday, March 12, 2011 - link
There is no Cortex-A8 multi-core ... Its either a Cortex-A9 or a modified custom-core like Qualcomm's Snapdragon that is a bastard child of A8 and A9 that supports SMP.Given the fact that despite the Intrinsity/PA SEMI purchases, the iPad/iPhone ARM chips have all looked like standard Samsung ARM parts, I highly doubt the iPad 2 is running a custom core. It is far more likely it is a standard Cortex-A9, with some custom power IC and other stuff.
Regarding the benchmarks, there are many other examples where going from Cortex-A8 to dual-core Cortex-A9 doesn't show >100% improvements...
metafor - Tuesday, March 15, 2011 - link
Because 2x the cores does not translate to 2x the performance under all but the tiniest fraction of circumstances.tipoo - Wednesday, March 16, 2011 - link
They were probably being conservative, as very few apps will scale 100% performance wise across both cores.LeTiger - Saturday, March 12, 2011 - link
"Other hardware chances are the addition of front and rear facing cameras for FaceTime and taking pictures, but unfortunately, they seem to be pieces lifted from the iPod touch and nothing near the iPhone 4's 5 megapixel shooter.""chances" might actually trying to be "changes"?
Just a thought. Thanks again for the write ups lately, It felt ridiculously relieving to get Anand's take on the new mbp and ipad line, I felt like I couldn't trust any other news site's objectivity until I had the scoop with Anandtech :) (and as always, this is the only place I go for legitimate reviews with ACTUAL analysis instead of spec-chimping)
Keep em' coming!