Testing shows 32nm A5 chip improves iPad 2 battery life by as much as 16%
A series of tests performed on the latest version of Apple's iPad 2 show that the move to a 32nm-process A5 chip results in battery life improvements of as much as 15.8 percent.
Alongside the launch of the third-generation iPad, Apple announced in March that it would continue to sell the Wi-Fi and 3G versions of its 16GB iPad 2 for $399.
It was later revealed that some new discounted iPad 2 models made use of special A5 chips manufactured using Samsung's 32nm chip-making process. Samsung previously used a 45nm process for the A5 chip.
Early on Friday, Anandtech published the results of its tests with the iPad, known as the iPad 2,4. The site reported having difficulty finding the newest version of the iPad 2, as there was no way to tell whether the device contained the 32nm or the 45nm chip without opening the box.
"This particular iPad 2,4 sample came from Best Buy, and several attempts to find one elsewhere came up short. All indications seem to point to the iPad 2,4 being relatively rare, which makes sense considering what's inside it," the report read.
In addition to the smaller 32nm process, the new A5 also reportedly uses high-k + metal gate LP transistors. The smaller die and new transistors combine to make the chip more power efficient.
Apple is believed to have used the new Apple TV and $399 iPad 2 as an opportunity to test the smaller chips before adopting the process for higher-volume devices.
The device lasted for 11.7 hours during a web browsing test, compared to 10.1 hours from earlier versions of the iPad 2. By comparison, the third-generation iPad lasted between 8.48 and 9.28 hours, depending on whether it was a Wi-Fi or LTE version. A 3D gaming test using the "Infinity Blade 2" application showed the iPad 2,4 as lasting 7.9 hours, compared to 6.12 hours on the iPad 2 and 5.58 hours on the third-gen iPad. A less graphics-intensive app also showed similar results.
Anandtech also conducted a video playback test that got 15.7 hours of playback out of the iPad 2,4. The iPad 2 lasted for just 13.3 hours, while the new iPad ran for 11.15 hours.
Power consumption testing showed a 28 percent increase in usage from the iPad 2,4 to the iPad 2 when under a heavy CPU load. Author Anand Lal Shimpi speculated that the third-generation iPad is "the perfect target" for the 32nm chip because of its higher power needs.
As for performance, the report didn't find any difference between the iPad 2,4 and the older iPad 2. It did, however, report that the iPad 2,4 was one degree Celsius cooler than the iPad 2 after playing "Infinity Blade 2" for an hour.
"If you're in the market for an iPad 2, the 2,4 is clearly the one to get - if you can find one that is," Lal Shimpi wrote. He expects more of the available iPad 2 inventory will switch over to the iPad 2,4 over time, but he said it's "still pretty tough" to find one for the time being.
The report speculated that Apple will "almost certainly" make use of Samsung's 32nm process for the next-generation iPhone.
"The move to 32nm is going to be good all around it seems, and Samsung appears to be a very capable foundry partner for Apple," Lal Shimpi concluded.
Alongside the launch of the third-generation iPad, Apple announced in March that it would continue to sell the Wi-Fi and 3G versions of its 16GB iPad 2 for $399.
It was later revealed that some new discounted iPad 2 models made use of special A5 chips manufactured using Samsung's 32nm chip-making process. Samsung previously used a 45nm process for the A5 chip.
Early on Friday, Anandtech published the results of its tests with the iPad, known as the iPad 2,4. The site reported having difficulty finding the newest version of the iPad 2, as there was no way to tell whether the device contained the 32nm or the 45nm chip without opening the box.
"This particular iPad 2,4 sample came from Best Buy, and several attempts to find one elsewhere came up short. All indications seem to point to the iPad 2,4 being relatively rare, which makes sense considering what's inside it," the report read.
In addition to the smaller 32nm process, the new A5 also reportedly uses high-k + metal gate LP transistors. The smaller die and new transistors combine to make the chip more power efficient.
Apple is believed to have used the new Apple TV and $399 iPad 2 as an opportunity to test the smaller chips before adopting the process for higher-volume devices.
The device lasted for 11.7 hours during a web browsing test, compared to 10.1 hours from earlier versions of the iPad 2. By comparison, the third-generation iPad lasted between 8.48 and 9.28 hours, depending on whether it was a Wi-Fi or LTE version. A 3D gaming test using the "Infinity Blade 2" application showed the iPad 2,4 as lasting 7.9 hours, compared to 6.12 hours on the iPad 2 and 5.58 hours on the third-gen iPad. A less graphics-intensive app also showed similar results.
Anandtech also conducted a video playback test that got 15.7 hours of playback out of the iPad 2,4. The iPad 2 lasted for just 13.3 hours, while the new iPad ran for 11.15 hours.
Power consumption testing showed a 28 percent increase in usage from the iPad 2,4 to the iPad 2 when under a heavy CPU load. Author Anand Lal Shimpi speculated that the third-generation iPad is "the perfect target" for the 32nm chip because of its higher power needs.
As for performance, the report didn't find any difference between the iPad 2,4 and the older iPad 2. It did, however, report that the iPad 2,4 was one degree Celsius cooler than the iPad 2 after playing "Infinity Blade 2" for an hour.
"If you're in the market for an iPad 2, the 2,4 is clearly the one to get - if you can find one that is," Lal Shimpi wrote. He expects more of the available iPad 2 inventory will switch over to the iPad 2,4 over time, but he said it's "still pretty tough" to find one for the time being.
The report speculated that Apple will "almost certainly" make use of Samsung's 32nm process for the next-generation iPhone.
"The move to 32nm is going to be good all around it seems, and Samsung appears to be a very capable foundry partner for Apple," Lal Shimpi concluded.
Comments
Good news, I hope when can get a cpu+gpu as well as a storage bump for ipad too in before the end of 2012.
Nice, IF you can get one.
Yeah, that probably won't happen.
Can u say "ripoff" by Apple for those who bought the current iPad 3? LOL
Ironted
Quote:
Originally Posted by IronTed
Can u say "ripoff" by Apple for those who bought the current iPad 3? LOL
Ironted
In what way?
Purchasers of the iPad 3 got a better processor, greatly improved video resolution, and a better camera. Just how is that a ripoff simply because the iPad 2 was also improved very slightly, as well?
Quote:
Originally Posted by jragosta
In what way?
Purchasers of the iPad 3 got a better processor, greatly improved video resolution, and a better camera. Just how is that a ripoff simply because the iPad 2 was also improved very slightly, as well?
I agree 100% -- there is no ripoff here. Apple has the right to substitute a faster, longer lasting system anytime it wants to. Also, Apple does not sell based on specs but user experience and there is no way a new iPad users experience would be denigrated by this chip in the iPad 2.
I'd also like to note these two paragraphs for those that don't understand why it's taking so long for Apple to use 32nm when Intel was using it years ago.
[INDENT]Apple decided to try out Samsung's 32nm HK+MG process on the A5 used in the 3rd generation Apple TV and some of the new iPad 2s. The former is a relatively low volume product for Apple, while the latter still moves in significant quantities. To deal with that fact, Apple is continuing to ship the original 45nm iPad 2,1 alongside the new 32nm iPad 2,4. Any hiccups in Samsung's production of the A5 and there are still more than enough iPad 2,1s to go around. The risk of moving to 32nm is effectively mitigated, while the learnings Apple gains from building the 32nm A5 will pay off later this year as Apple ramps up production of a 32nm SoC for use in the next iPhone. It's a very smart strategy, one you would expect from an experienced chip company - not a device vendor. When you consider that Apple employs chip architects who have worked on everything from the Athlon 64 to the Cortex A15, Apple's behavior is no longer that surprising.
[...]
I'd would like to applaud Apple's 32nm migration plan. By starting with lower volume products and even then, only on a portion of the iPad 2s available on the market, Apple maintains a low profile and gets great experience with Samsung's 32nm HK+MG process. It's very clear that this is all in preparation for the next iPhone, which will almost certainly use Samsung's 32nm process and require it in significant volumes. It's obvious that Apple employs some very smart chip heads in Cupertino.[/INDENT]
You could say that Apple is moving forward using [32nm] technology with boldness and prudence -- let the chips fail where they may...
Sure hope they move to 32nm in the new iPad then. For the next iPhone it's pretty much assured given the timeline, but it would help a lot with the iPad too.
Quote:
Originally Posted by tipoo
Sure hope they move to 32nm in the new iPad then. For the next iPhone it's pretty much assured given the timeline, but it would help a lot with the iPad too.
Just a guess here, but the 32nm capacity will probably be maxed with the new iPhone through the end of the year and by then you're just a few months from the next revision of the iPad (wouldn't make sense to change the processor at that point). I'd guess the 2013 iPad revision will include the move to 32nm on the processor...and possibly change the processor in the 2012 iPad to 32nm since they'll be cheaper once the production volume is there. JMHO.
Quote:
"If you're in the market for an iPad 2, the 2,4 is clearly the one to get - if you can find one that is," Lal Shimpi wrote.
But it is not identified as such on the box - you need to open the box to identify it. I predict that lots of iPad 2 dealers will have plenty of "open box specials" in the near future if word gets out about this.
This benefit will scale with the CPU usage intensity. You won't get the same scale of benefit from ready iBooks or listening to podcasts.
Quote:
Originally Posted by SolipsismX
I'm glad they tested this. I certainly asked them enough times. I also agree with their conclusion that the 6th gen iPhone will use the 32nm process.
I'd also like to note these two paragraphs for those that don't understand why it's taking so long for Apple to use 32nm when Intel was using it years ago.
Apple decided to try out Samsung's 32nm HK+MG process on the A5 used in the 3rd generation Apple TV and some of the new iPad 2s. The former is a relatively low volume product for Apple, while the latter still moves in significant quantities. To deal with that fact, Apple is continuing to ship the original 45nm iPad 2,1 alongside the new 32nm iPad 2,4. Any hiccups in Samsung's production of the A5 and there are still more than enough iPad 2,1s to go around. The risk of moving to 32nm is effectively mitigated, while the learnings Apple gains from building the 32nm A5 will pay off later this year as Apple ramps up production of a 32nm SoC for use in the next iPhone. It's a very smart strategy, one you would expect from an experienced chip company - not a device vendor. When you consider that Apple employs chip architects who have worked on everything from the Athlon 64 to the Cortex A15, Apple's behavior is no longer that surprising.
[...]
I'd would like to applaud Apple's 32nm migration plan. By starting with lower volume products and even then, only on a portion of the iPad 2s available on the market, Apple maintains a low profile and gets great experience with Samsung's 32nm HK+MG process. It's very clear that this is all in preparation for the next iPhone, which will almost certainly use Samsung's 32nm process and require it in significant volumes. It's obvious that Apple employs some very smart chip heads in Cupertino.
Well, probably that is not exactly like that.
Samsung is moving all its high-end ARM processor to the new 32nm process that include the Exynos 2 Dual 32nm (from the Exynos 2 Dual with 45nm on the Galaxy SII), the Exynos 4 Quad and the higher-end Exynos 5 Dual.
All of them, except the Exynos 5 Dual are Cortex A9 based and Samsung is slowing down to stop the A9 45nm process to just use the 32nmm process.
The only processor that could continue with 45nm process is the Cortex A8 Exynos 3.
The Apple A5 processor is based on Cortex A9 architecture and because Samsung is moving already all its A9 production to 32nm process, the Apple processor goes include on it.
That is a Samsung decision to switch to 32nm on all its A9 processor and evidently Apple is using it.
Quote:
Originally Posted by SuperTomcat
Well, probably that is not exactly like that.
Samsung is moving all its high-end ARM processor to the new 32nm process that include the Exynos 2 Dual 32nm (from the Exynos 2 Dual with 45nm on the Galaxy SII), the Exynos 4 Quad and the higher-end Exynos 5 Dual.
All of them, except the Exynos 5 Dual are Cortex A9 based and Samsung is slowing down to stop the A9 45nm process to just use the 32nmm process.
The only processor that could continue with 45nm process is the Cortex A8 Exynos 3.
The Apple A5 processor is based on Cortex A9 architecture and because Samsung is moving already all its A9 production to 32nm process, the Apple processor goes include on it.
That is a Samsung decision to switch to 32nm on all its A9 processor and evidently Apple is using it.
Your post makes absolutely ZERO sense. The fabs that produce 45nm chips are not the same as the fabs that produce 32nm chips. Additionally, designs that operate at 45nm cannot magically be placed on 32nm technology. Large teams of engineers will have to place and route the new chip at 32nm. Then they will have to fix new timing bugs, yield bugs etc. A design shrink is no easy task. Its not like buying a new fab magically makes chips designed on a different process technology.
The real point is that Samsung makes money off both customers who use their 45nm and 32nm fabs. They don't care what functionality these chips implement, and definitely don't care if they contain cortex a8 or cortex a9 cores. The cost to make these transitions is tremendous to both samsung, and their vendors, and they will decide what to implement. If anything, samsung might limit competitors from using their newest fabs to design chips that compete with their own designs, as this would give samsung a competitive advantage (note: there is no sign whatsoever that samsung has done this, or would ever do this). They wouldn't force competitors to use a faster tech node.
Phil
Also we have to remember there is a considerable amount of Apple IP in Samsungs processors or was at one time. People should look at the relationship between Apple and Samsung as a partnership. I wouldn't be surprised to find out that Apple has a stake in Samsungs Texas plant. A stake would assure Apple of a reasonable supply of chips from the plant.
Geeknench for the iPad. First one I tested is not on this new 32 Nm.
Intel's new Ivy Bridge x86_64 chips are 22nm. And yet the best you can get from Apple (in terms of ARM) is 45nm or if you're lucky 32nm.
Given their focus on mobile devices, battery life and thinness, I am surprised Apple are not also at the cutting edge of new manufacturing processes.
Quote:
Originally Posted by ascii
Intel's new Ivy Bridge x86_64 chips are 22nm. And yet the best you can get from Apple (in terms of ARM) is 45nm or if you're lucky 32nm.
Given their focus on mobile devices, battery life and thinness, I am surprised Apple are not also at the cutting edge of new manufacturing processes.
What do you want Apple to do? In spite of the differences in process technology, ARM is STILL the best performance per watt available (although the latest Intel Atom chips are catching up - but they're not 22 nm, either).
You're getting confused by looking at the wrong specs and assuming that any chip at 22 nm is automatically better than a chip at 32 nm - and that's always the case. Ivy Bridge would be a horrible chip for the iPad.
Is Apple working with vendors to shrink their chips? Obviously - and they've been doing it regularly. They will probably not catch up to Intel on line widths, but that doesn't mean that Apple's ARM chips are uncompetitive in terms of what matters - performance per watt.
It would be ideal if Intel would make Apple's chips using the latest Intel technology, but it doesn't appear that this will happen, so Apple has to choose the best products available to it - and at present, ARM is the best choice. Not to mention, of course, the risk of getting further into bed with Intel after Intel screwed them on the Ultrabooks.
Quote:
Originally Posted by SuperTomcat
Well, probably that is not exactly like that.
Samsung is moving all its high-end ARM processor to the new 32nm process that include the Exynos 2 Dual 32nm (from the Exynos 2 Dual with 45nm on the Galaxy SII), the Exynos 4 Quad and the higher-end Exynos 5 Dual.
All of them, except the Exynos 5 Dual are Cortex A9 based and Samsung is slowing down to stop the A9 45nm process to just use the 32nmm process.
The only processor that could continue with 45nm process is the Cortex A8 Exynos 3.
The Apple A5 processor is based on Cortex A9 architecture and because Samsung is moving already all its A9 production to 32nm process, the Apple processor goes include on it.
That is a Samsung decision to switch to 32nm on all its A9 processor and evidently Apple is using it.
I doubt very much that it's that simple. Processor design is an iterative process and Apple has undoubtedly been involved in the process all along. At some stage in the process, Samsung and Apple undoubtedly jointly agreed on the process technology because there are tradeoffs. Perhaps Samsung could only deliver x 32 nm chips, but could sell x + y 45 nm chips. Or perhaps they had to charge Apple more for the 32 nm chips because the plant is not fully operational and they are capacity constrained. Or perhaps Apple had to accept a higher level of rejects on 32 nm. I have no idea what exactly was involved with the discussion, but you can be sure that it was not solely a Samsung decision.