NVIDIA Tegra 2, Samsung Exynos, and Qualcomm Snapdragon the 3rd: the dual-core chipsets and beyond

11comments
NVIDIA Tegra 2, Samsung Exynos, and Qualcomm Snapdragon the 3rd: the dual-core chipsets and beyond
It's been less than six months since we wrote our overview about the future of mobile CPU affairs, and there is already a need for an update, in light of the movers, shakers, and twists of fate, so we ought to follow up. Four major things happened - ARM revealed its Cortex-A15 "Eagle" chipset plans, Samsung came out with its dual-core Orion creation, NVIDIA successfully slipped Tegra 2 under the door of Android handset manufacturers desperate for a standout feature, and Microsoft bowed under peer pressure, announcing it will embrace ARM-based chipsets for Windows too.

While the first three were somewhat expected, the last one is a tectonic shift, and we can't stress enough how important it is that the vast driver support and familiar Windows environment will be marrying the power-efficient ARM silicon with its tiny footprint. It could either be a game-changer, if Microsoft plays it right, or sink Redmond into playing second fiddle to the cool kids on the block Apple and Google.

The Promise

The new wave of multi-core mobile silicon promises 1080p Full HD video recording and playback (first sample here) from camera resolutions up to 18MP, as well as some speed boosters, like support for DDR2 and DDR3 memory, as well as console-level graphics. Well, hello, instant  webpage load times, 3D user interfaces, 3D video capture and playback, simultaneous output to three or more screens, and other eye candy. All on the same battery charge, or less.

Let's get one persistent myth about dual-core chipsets out of the way right now. More cores does not mean sucking more juice out of the battery, if the software is optimized, and power management is done correctly, so don't lose much sleep over it. These two or more 1GHz+ cores are built with a 45nm process, compared to the 65nm process of, say, the first generation 1GHz Snapdragon, as found currently in Windows Phone 7 devices. Shrinking the transistors' size means that more of them can fit on the same place, pumping out higher performance from the same footprint. Since smaller transistors consume less energy, they can also be run at higher frequencies, without your phone starting to drip on your feet from overheating.

Recommended Stories
So you can choose between more transistors, hence more power on the same space, or a similar amount of transistors with a smaller footprint, which leaves space for other stuff, like better graphics, or extra connectivity chips. Not to mention you can fit more of them on the same standard silicon wafer now, which makes the chips cheaper to produce, if we don't count the R&D expenses. And these advantages are present in the single-core chipsets too, produced with the 45nm process, the best representatives of which are Hummingbird on the Samsung Galaxy S, the A4 on the iPhone 4 and the iPad, and the second generation Snapdragons, like those in the upcoming HTC Inspire 4G, or the T-Mobile G2.

When we put two cores on the same footprint, we should theoretically get more efficient system, according to one NVIDIA white paper, which tries to be a public defender of the multi-core case. Look at the chart on the right for their logic. It might be true in reality, or it might turn out half-baked, we only know that the apps that take full advantage of two cores have still to be embraced. And the only mobile OS that supports multi-core arrangements in real time for now is QNX, as found on the BlackBerry PlayBook.

So as far as efficiency goes, dual-core has yet to prove itself, but still, let's see who will be the major players in the upcoming multi-core tsunami. What the mobile chipset industry is going through now, is where desktop CPUs were a few years ago - the GHz war subsided due to overheating, to the advantage of multiple cores, and more efficient designs.

The Players


NVIDIA Tegra 2


All we can say about the NVIDIA Tegra 2 chipset is "well-played, NVIDIA, well-played". The company's dual-core silicon might not be the fastest or most efficient, but it beat everybody to the punch in a time when cell phone manufactures were desperately looking for differentiating factors in the sea of Android handsets. And so it was chosen by Google to become the reference platform for Android 3.0 Honeycomb with the Motorola XOOM tablet. More about it in our recap of the events here.

NVIDIA is currently the only one you can find in commercially available phones and tablets on the market. It pairs two 1GHz ARM Cortex-A9 cores with a low-power version of its GeForce GPU. The energy-saving case is illustrated by NVIDIA with its napkin calculation that two cores doing the same demanding task utilize 40% less energy than a single core, since they don't stress and sweat that much. When the lonely CPU is 100% loaded, it runs on the max 1GHz clock speed, heats up and consumes the maximum 1.1V, while two CPUs distribute the load among themselves, and run on 550MHz, thus running cooler, and drawing just 0.8V in total, resulting in the said 40% gain. What if we manage to max out both cores, though? Then the task will just be finished for a shorter time?



This talk about power efficiency comes from the company that makes quite power-hungry graphics cards. We also saw some promises in the Tegra 2 white paper for 140 hours of audio playback on a charge, while the other Cortex-A9 chipsets with two cores are listing 120 hours. The twist is that they are measuring a charge from a 1000mAh battery as standard, while NVIDIA measures it with a bigger 2000mAh one. If audio playback is an indicator, Tegra 2 could be up to 40% less efficient than the other dual-core competitors, which are yet to make a cameo appearance in a gadget, though.

Maybe that is why the Motorola ATRIX 4G has a huge 1930mAh battery. It promises 9 hours of talk time, which is great for an Android phone, but the single-core Motorola DROID 2 manages even more, and with a 1400mAh battery, so there will be more efficient dual-core chipset designs than Tegra 2 out there, that's for sure.

We'll see what the true values are when production-ready handsets hit the shelves. Other than that we don't doubt NVIDIA when it says it achieves over 2x faster page loading times, and sixty to hundred percent better frame rates in 3D games. The impressive Quadrant scores that the Tegra 2 winter crop is achieving attests to that. Motorola ATRIX 4G scores 2600+, the LG Optimus 2X makes 2391, and the Motorola DROID BIONIC 2284 right out of the box - these are the highest scores we've seen on this generic benchmark, and the test is not even taking into account the second core.


Moreover, NVIDIA says Tegra 3 is almost finished, taking us to quad-core level and CUDA CPU/GPU workload sharing. Work has already started on Tegra 4 as well, but even the 3rd generation won't see the light of day until probably late this year. Tegra 3 is to be announced at the MWC 2011 expo in Barcelona.

Samsung Exynos - the Galaxy Monster

In September of last year Samsung lifted the veil over its own 1GHz Cortex-A9 dual-core puppy, called Orion (what's with the cosmic names, Sammy?). It got renamed to Exynos since, and has been demoed on a tablet prototype. It is now powering the Samsung Galaxy S II.

The Hummingbird generation is the leader in graphics performance, as can be seen from this below benchmark comparison:



The Hummingbird chipset in the Galaxy S and the Apple A4 share a lot of similarities. After all, Samsung used to provide the silicon for the previous generations of the iPhone. The biggest difference is the graphics chip. Apple’s A4 keeps the PowerVR SGX535 that is in the iPhone 3GS, but clocks it up thanks to the 45nm manufacturing process, thus increasing performance to deal with the Retina Display resolution. Hummingbird, however, upgrades to PowerVR SGX540, which is way faster than 535, but still uses acceleration technology developed by Intrinsity, a company which Apple bought last year.

Perhaps that acquisition is the reason why Samsung went with something else than PowerVR for its own dual-core chipset. In February last year it announced that it will use not only ARM’s CPU architecture, but its graphics chip designs as well, called Mali. There are different versions of how fast the quad-core Mali-400 GPU in Orion is. Samsung said in the press release that it provides 5 times the performance of its previous generation chipset, i.e. Hummingbird, which would clock it at peak 450 million triangles per second. For the sake of comparison, that’s almost as much as the current-gen Xbox 360 processes. Guesstimates set Exynos's GPU speeds anywhere between 120 and 450 million triangles per second.

It all might have to do with the frequency that these GPUs are being clocked at. For mobile devices they are usually underclocked so as to keep battery consumption at bay. Sammy boasted a peak performance of the SGX540 in the Hummingbird to be 90 million triangles per second, but it's unclear if those are achievable at its 200MHz clock rate there. It is still the undisputed king of smartphone graphics, though, and there aren’t any games or other software that choke it up yet, so the "millions of triangles" race is a bit pointless for now. Have a look at the concept video for the next version of Mali :

Video Thumbnail


One other feature, which caught our attention, was that Exynos supports DDR3 memory, whereas Tegra 2 supports DDR2. Still desktop-grade memories, but the difference in performance might be significant. Also, since ARM provides both the CPU and GPU design, there might be just as significant cost and power savings for Samsung going the Mali way.

Hopefully we will know pretty soon what Exynos is exactly capable of, when we get our paws on the gear in Samsung’s booth at the MWC expo next month. In the meantime, Sammy is reportedly a heavy buyer of Tegra 2 chipsets. This could mean that while the future in front of Samsung’s Exynos may be bright, it might not be near.

Snappy Dragon

The king of integrated system-on-a-chip solutions will very soon throw down the gauntlet into the ring as well. Qualcomm is just warming us up with its second generation Snapdragon single-core chipsets, made with the 45nm technology. These are currently present in the T-Mobile G2, T-Mobile myTouch 4G, HTC Desire HD, and the upcoming HTC Inspire 4G for AT&T, Sony Ericsson Xperia arc, and the LTE-enabled HTC ThunderBolt. These chips are no slouches either - the ThunderBolt did 1700+ on Quadrant, placing it at the top of the single-core game. Qualcomm’s huge advantage is that it integrates all that the others offer, and rounds it up with baseband connectivity of your choosing, be it 3G, LTE, or HSPA+. 



But since we are spoiled brats, we will pay more attention to the third generation of the Snapdragon chipset, which will have two cores at up to 1.2GHz, and even 1.5GHz, possibly for tablets. Qualcomm demoed all kinds of trickery with them at the CES show, including stereoscopic 3D video with HD resolution, multi-party video conferencing, and fascinating gaming physics on the new Adreno 220 GPU. Unlike Tegra 2, the dual-core arrangement in the 3rd gen Snapdragon can selectively turn off one of the cores completely, thus being more power-efficient.

One of these chipsets even has all 3G radios known to man inside, so it can play nice with all four major US carriers, if needed. For now it is confirmed that the Asus EeePad MeMO Honeycomb tablet will be running a dual-core Snapdragon inside, but it’s not coming until the summer. Have a look at a quick video about the platform:

Video Thumbnail


Qualcomm also has a wonder-chipset up its sleeve, where the future is bright, but not near. The MSM 8960 Snapdragon will have 5x the performance of the 3rd generation Snapdragons with two cores, and will sap just 25% of the power. LTE finally unites with all 3G bands of the world in a single radio there, so we’d rather wait till Q2 of 2011, when devices with this fourth generation Snapdragon ship, rather than pay Verizon’s or AT&T’s early adopters LTE premium  this year.

OMAP4, the Texas Ranger

Extremely promising at first, with a lot of hype surrounding its development, Texas Instruments’ dual-core OMAP4430 chipset got somewhat lost in translation in the last months. Its graphics subset might be last-year’s finest - the PowerVR SGX540 - but it found its way into the BlackBerry PlayBook, so will have an idea about its capabilities very soon.

Texas Instruments already makes some of the more power-efficient designs here, as attested by the Motorola DROID 2 and DROID X, so we will be curious to see the OMAP4 in tandem with the only mobile OS that supports multicore chipsets in real time - RIM’s QNX. Early rumors raised alarm over a short-lived battery, but we are still in prototype stages, so we shouldn’t be drawing final conclusions yet.
.
Video Thumbnail


Next in line here is the OMAP4440, which speeds the two cores up to 1.5GHz, and has the new PowerVR chip, which will allow it to capture 3D stereoscopic images from two 12MP cameras, this time with 1080p Full HD resolution. The OMAP4440 pitch is also an ability to do videoconferencing in 720p HD resolution with up to four chat participants, which is pretty sweet. This chip just began sampling, so volume production is not coming until later this year.

Marvell and Freescale

These two are also hoping to jump from the relative calm of e-readers and the like, to the hustle and bustle of the smartphone/tablet universe. Marvell managed to wow us with a tri-core design, the ARMADA 628. Two of them are humming at 1.5GHz, while the third, at 624MHz, serves for everyday tasks to preserve battery life, then the dynamic duo kicks in for the heavy lifting:

The architecture is analogous to a hybrid muscle car.  The ARMADA 628 is intended to perform like a race car engine on demand, while still delivering the frugal gas-mileage of a hybrid automobile. In real world terms, this enables the ARMADA 628 to play more than 10 hours of full 1080p HD video or 140 hours of music on a single charge while still providing 3 GHz of raw computational horsepower.

Capable of dual-stream 1080p full HD video encoding and decoding at 30fps, the chipset should support 3D visuals at stunning resolutions, similar to TI's OMAP4440. Graphics performance is also boosted up from the current SoC generation - the chipset is rendering 200 million triangles per second, which is more than double what the best in class Hummingbird in the Galaxy S is now offering.

Marvell is traditionally huge in e-readers, so another notable feature is support of up to four displays - two LCDs, HDMI-out, and an ePaper controller are all listed in the mix. Marvell's creation has built-in USB 3.0, which we haven't heard the other contenders to support, and is a very welcome addition to transfer those full HD videos quickly out of the handsets. BlackBerry phones use older versions of Marvell's Armada family, so if we take a wild guess, RIM could catch up quickly on the hardware front.

ARMADA 628 supports all major mobile operating systems, and is now in manufacturers' hands for sampling, which means that we will be seeing devices with it in the second half of 2011 at the earliest. That's about the time when the third Snapdragon generation at up to 1.5GHz should be entering smartphones and tablets, so Marvell might have given Qualcomm a good rush with its announcement.

Not to be outdone by Marvell, Freescale also recently announced its product series, which goes all the way up to the quad-core Freescale i.MX 6 Quad processors. That’s four ARM Cortex-A9 cores running at up to 1.2 GHz per core, folks, very impressive. It doesn’t support full HD in 3D video capture mode like Marvell’s SoC does, just 720p. Technically it might, since the GPU is probably the same latest generation ARM Mali T604 variety, as it is also rated at 200 million triangles per second. Freescale’s i.MX 6 also supports an optional ePaper display controller, like Armada 628, and its differentiating feature is the built-in support for the VP8 codec. That means Google could pick it up as the next Android reference platform, since it is really pushing forward its open WebM video standard, and even cut H.264 support in its Chrome browser recently.

In the previous article we said our money is on NVIDIA as the underdog, which might surprise the big players, and that turned out true. Now our money is on Marvell or Freescale, because USB 3.0 support (Armada 628) and hardware WebM decoding (Freescale i.MX6) are truly novel features. We are slightly leaning towards Freescale, since in the press release the company promises 24 hours of HD video playback on a charge, while the others are offering half of that, but it doesn’t specify from what battery capacity. Still, it is more likely that these chipsets will be picked for placement in tablets, rather than smartphones.

Broadcom - dual-core on a budget

In a refreshing break from the rat race the rest of the chip-making crowd is engaged in, Broadcom announced its dual-core BCM2157 Android chipset, which clocks at... 500MHz. Before you scoff at the speed, bear in mind this will lead to very decent Android handsets at a price that will make you smile.

The SoC is complete with its own 3G, Wi-Fi, GPS and Bluetooth connectivity suite. It also supports multitouch on screens with HVGA resolution, and cameras up to 5MP. All low- to mid-range features nowadays, but we'd imagine Android phones with it might break the $150 barrier downwards.

Besides the low price point, the features that set it apart from the pack, are that the chipset supports dual SIM/dual standby mode, and has a low-power NFC chip embedded, so your affordable phone won't be left out of the mobile payment revolution. The NFC chip is Broadcom's own creation, it can store data locally in its own memory, and draws power only when activated by an NFC reader at POS terminals. Sounds good, moreover the chipset is already sampling with select customers, and will be commercially available this quarter.

ARM vs Intel

All of the above chipsets are based on design architecture licensed from ARM Holdings. Last August the company announced its plans for the future with the Cortex A-15, codenamed "Eagle". The current high-end ARM chips are 32-bit, and able to address only 4GB of RAM, while the Eagle upgrade will allow for a mobile OS to address up to 1TB of memory.

This will make hardware-based virtualization possible, which might mean Android and Windows Phone 7 in one and the same handset, or carriers easily porting a different operating system to the same hardware, as the market conditions shift. Cortex-A15 can be produced with the 32nm/28nm, and even 20nm processes further down the road, meaning even faster chipsets with better power consumption.

Cortex-A15 can reach 2.5GHz with its four cores variant, but these will be mainly for cloud servers and wireless towers; smartphones and tablets will receive the power-sipping 1-1.5GHz iteration, served on one or two cores. ARM also stated that Eagle will be backwards compatible with the existing ARM architecture, so all current mobile operating systems and applications should be running with ease on the new hardware.

Texas Instruments is the first that licensed Eagle, and is already promising a 60% reduction in power requirements, compared to the OMAP chipsets based on Cortex-A9. Eagle, of course, won't enter devices at least until 2013, but just the thought of five times the speed with double the battery life of existing smartphones will keep us warm until then. Other partners that worked with ARM on the Eagle developments have been Samsung and ST-Ericsson, so we will surely be seeing the chipset in Samsung, and maybe Nokia devices, when the time comes.

ARM encroaching onto the server space should raise the hairs on the neck of even the juggernaut Intel. If they can design a mobile chipset with much lower energy footprint, but comparable performance, and way more capable integrated graphics, Intel should be very, very worried.

Well, it isn’t. Intel bought the Germans from Infineon, which make baseband radios, similar to Qualcomm, in an effort to make its own mobile walled garden. The chip giant even managed to spin in a positive way the jaw-dropping announcement that Microsoft will be cheating on Intel’s x86 architecture with the young Prince ChARMing. Intel’s CEO said that it is a good thing, since Microsoft has apparently rolled up its sleeves to create a touch-optimized and scalable Windows 8 from scratch. If Microsoft is aiming to use it in mobile devices, with the same stack, but different interfaces, depending on what gizmo Windows 8 appears in, he said, then Intel will be there with suitable chipsets to power that transition, all the way to smartphones.

Paul Otellini also noted that using Intel's mobile chips will lead to booting multiple operating systems on one device, and he seems to know what he is talking about. Not to forget Intel just settled its last major patent lawsuit with NVIDIA for $1.5 billion, and the deal involved some cross licensing of know-how, so its integrated graphics will only be getting better. If the Windows 7-running Evolve III Maestro, Samsung Sliding PC 7 Series, and Asus Eee Pad EP121 tablets are a harbinger of things to come when Windows 8 lands next year, we wouldn’t be so brave to write Intel off the mobile game. Here are video demos of the Eee Pad EP121 12" powerhouse, followed by one of the cool Samsung Sliding PC 7 Series slate, which lasts 9 hours on a charge:

Video Thumbnail


Video Thumbnail


The Evolve III Maestro, on the other hand, costs as much as an iPad, and triple-boots Windows 7, Android and MeeGO, with finger-friendly interfaces on top of all three, while managing to squeeze eight hours of Windows-ing with the dual-core Intel Oak Trail chipset and sixteen on Android. Next year these might go full day with Windows 8 and Intel's next mobile CPUs, and that’s all we want - a catfight between ARM and Intel, bringing us awesome silicon (not Pam Anderson awesome, but give them time).

What next?


These next year chipsets will be cast by the blacksmiths in the foundries with the 28/32nm methods, inherited from their forefathers, but what’s even further down the pipeline, inquiring minds want to know? Well, the 20nm technology beckons us all, and Samsung just teamed with IBM’s researchers to coin the production methodologies for the next generation of smartphones. ARM is entering a similar collaboration with IBM, but for processors made with the even smaller 14nm technology, all for smartphones and tablets. The step after these is so hard to craft, that Samsung turned the other way, and formed an alliance with Intel and Toshiba, to try and achieve the elusive 10nm process.

These are all probably part of 5-year plans, all the way to 2016, so hopefully we won’t have to update this article every six months or so - wishful thinking, we know. By that time our phones will be foldable, beaming our holograms at videoconferences, and always connected to all the knowledge and people in the world, as well as our robotic vacuum cleaners at home; all the while we charge them wirelessly about once a month, and pay the equivalent of two beers to the carriers. Again, we wish, but if we don’t go back to rubbing stick and stone for fire in 2012, the mobile future is so bright we are squinting.

source: Anandtech, B3D

Recommended Stories

Loading Comments...
FCC OKs Cingular\'s purchase of AT&T Wireless