AMD Trinity APU - A Notebook Platform Performance Review

** Updated as of 16th May 2012, 9.30am - We've included the technical details of the Trinity APU to go along with the reference notebook performance analysis.

Rise of the Underdog

Today is AMD's big day as it officially launches the AMD Trinity APU - the next generation APU architecture from AMD. Built on the mature 32nm SOI process technology, the Trinity APU will be using a second revision of the Bulldozer architecture and will be a key proponent for ultrathin AMD platform based notebooks. Last month, we gave you our brief experience with the Trinity APUs from the AMD Trinity APU Tech Day event held at Austin, Texas (USA). While that was purely from a standpoint of general usability, it already showcased a marked improvement against their competitor and further cementing the fact that AMD is banking heavily on their heterogeneous system architecture (HSA). As the name suggests, HSA leverages on both the CPU and GPU portions to execute tasks in parallel and thus giving rise to the balanced system architecture that AMD often touts about. This is essentially what an APU is about - an accelerated processing unit which doesn't really concern the user what's really underneath the hood and is able to deliver the right experience utilizing its full set of resources (processing units).

AMD Trinity - What’s in a Name?

So let’s get started with deciphering codenames, specs and all techie matters. What is the Trinity APU about? Trinity is the true second generation APU variety from AMD. Although the Llano APU is the second variety launched after Ontario in the AMD Fusion family of processors, it was still considered first generation APUs. This is despite the fact that the Llano and Ontario APUs are quite different in several aspects. However, both were the first set of APUs to address the netbook and mainstream notebook segments respectively.

Trinity on the other hand is a replacement for Llano, targeting the same market segment, but with far more improved CPU and GPU components to bring about far better performance per watt with the best in-class entertainment and gaming experience. In fact, Trinity got its name for combining three different architectures - the new Piledriver CPU cores, Northern Islands GPU cores (more specifically Caymen’s VLIW4 architecture) and the video processing and display support engine from Southern Islands GPUs. With so many features enrolled, Trinity promises quite a bit and thus it’s only fitting to be called the true second generation APU. 

Peering Inside the Trinity APU

Like the Llano APU before it, the Trinity APU will also be officially known as the A-Series APU - the only difference is the new model numbering scheme which will be denoted by the 4000-series instead of the predecessor’s 3000-series. Featuring the same 32nm SOI manufacturing process and even the same number of cores which number between two to four ‘cores’. However, instead of the old K10 derivative architecture, the AMD Trinity APU takes on an enhanced version of the Bulldozer core.

Now we’re all aware that the Bulldozer CPU core architecture used in the AMD FX processors didn’t fare well mostly because of its ‘dual-core’ module which essentially features dual integer pipelines sharing a floating point unit. As such, a four-core chip will have quad INT units and dual FP units. This unusual design choice was made for more efficient resource sharing in mind as opposed to each processing module having a balanced integer and floating point unit. Obviously AMD wants to make the CPU die more compact with such a setup, but they did also openly admit that the design trade-off is in processing existing x86 instructions in an optimal manner. This basically affects almost all programs until unless updated compilers are used with new applications to better take advantage of its design.

As with many broad-organization-wide decisions, Bulldozer was their future CPU architecture investment which they had to either fix or give it up. They chose the more logical choice to ‘fix’ or improve the architecture which gave birth to the Piledriver CPU architecture that will first be featured on the Trinity APU. Also known as the Enhanced Bulldozer, it was primarily spruced up in terms of new instruction set architecture support (such as AVX, AES and more), several scheduler and pre-fetcher upgrades, improved branch prediction and enhanced cache structure.

Each ‘dual-core’ Piledriver processing module is accompanied by a 2MB L2 cache (which can scale down to as low as 512KB on some entry-level variants), foregoing the L3 cache for mainstream mobile processors as was the case even on Llano. AMD further clarified that negligible improvements were seen in at their design stage to incorporate another cache layer and occupy unnecessary die space. On a typical upper-tier ‘four-core’ A-series APU, this would mean a total of 4MB of L2 cache per CPU. Since it’s still based on the 32nm manufacturing process, the Trinity APU’s 1.303 billion transistors occupy 246mm² as compared to the Llano APU’s 1.178 billion transistors that have a footprint of 228mm².

With the more mature 32nm SOI manufacturing process, AMD now is able to deliver even 17W TDP parts for the mainstream APUs as opposed to 35W being the minimum previously. This will help it better get into the Ultrabook race that Intel began but AMD will pursue those favorable traits with the Ultrathin naming convention. However, there’s only one part qualified for the 17W TDP profile and with such limited specifications, we’re note quite sure how it will compete with Intel’s various ultra low voltage Sandy Bridge and Ivy Bridge 17W processors used in Ultrabooks until we encounter an Ultrathin notebook for our evaluation. The full list of mobile AMD Trinity APU processors are as listed:-

* Take note that every two CPU cores make up 1 Piledriver module. Hence a four-core processor has two Piledriver modules.

AMD’s documentation also revealed that besides the mobile Trinity APUs designed for 17W, 25W and 35W TDP variants, they also mentioned that the future desktop variety will comprise of 65W and 100W TDP variants. This is similar to their existing line-up, but we feel it’s not competitive enough against the competition. AMD’s Trinity solutions are rather mainstream oriented compared to the Intel’s line-up that’s much higher performing for the same TDP profile. AMD will likely have to play the value card once more, but we’ll reserve further comments when we get to test relevant systems.

AMD Turbo Core Technology 3.0

One of the major feature improvements in the Trinity APU is support for the latest Turbo Core technology, now in version 3.0. For those interested to catch-up on how AMD’s Turbo Core works, you can read up all about it in our Phenom II X6 coverage where it first debuted. This original version only had two states of operation where you get turbo clock speeds or you don’t. When half or more of the processor’s cores are idling, the other active half will operate at a speedier predefined clock speed. In version 2.0, a third state was made available so that all the cores get some small boost frequency as long as they are within the TDP of the CPU. This was implemented on the AMD FX.

Since Trinity uses the new Piledriver core which is an enhanced Bulldozer architecture, Turbo Core has been more thoroughly overhauled in this third iteration as it touts automatic bi-directional power management between the GPU and CPU portions of the die. This is a drastic improvement since previous Turbo Core iterations could only ramp up/down the frequency and voltage aspects on the CPU portion.

Turbo Core 3.0 is able to achieve this via thermal mapping of the die as the Trinity APU constantly and dynamically calculates temperature of the CPU cores and the GPU block based on loading level estimation and maps them to obtain an optical operating point to maximize performance from both processing units - all while staying within the safe operating temperature limits. Note that this isn’t based on measured temperature, but calculated based on workloads which AMD has verified in their labs that turbo boosting behavior is fairly consistent and predictable for any given loading level. Thanks to Turbo Core 3.0, Trinity APUs list base and boost clocks for both the processor and the GPU.

The Graphics Quotient

An article on AMD”s Fusion APU processors is far from complete with if we omitted the advances made on its built-in graphics engine. After all, the integration of a powerful graphics core to complement mainstream class CPU cores is the main formula ever since Fusion processing platform began.

On the Trinity APU, the graphics core used is a Radeon HD 7600M class GPU, but unlike the naming suggestion, it’s not based on the newer Southern Islands architecture, but that of the Northern Islands series (Radeon HD 6000 series) - this was first mentioned here. To be specific, it sports the VLIW4 core design first used on the high-end Caymen GPUs (Radeon HD 6900 series) that is more performance optimized per die area and is easier to schedule/manage workloads. This also means that the number of stream processing units (or Radeon cores) would seem less than the previous generation IGPs, but you’ll have to understand that they are different in architecture and capabilities. The Triniy APU’s graphics core is supposed to deliver actual performance that is roughly in the range of Radeon HD 6600-class GPUs.

Some people may have thoughts if the integrated graphics is really discrete class graphics and AMD assured us that the IGP core is equivalent to the identically named mobile discrete GPU. This means the Radeon HD 7660G on the A10-4600M APU performs similar to a mobile Radeon HD 7660M GPU module (if a direct equivalent exists). You’ll notice that the “G” suffix refers to the IGP edition, whereas “M” suffix refers to the mobile GPU variant. Later in our performance results, you’ll notice that this is somewhat true.

However there's more as these integrated GPUs also take some of the best features off the Southern Island GPUs like the AMD HD Media Accelerator (includes UVD 3.0 and Accelerated Video Converter), Video Compression Engine (VCE), DisplayPort 1.2 support and AMD Eyefinity technology for up to quad display capability. To be precise, the integrated graphics core has 3 independent display controllers, while the fourth display is driven by a software controller via DisplayPort 1.2 daisy-chain connectivity.

Overall, you could say that the Trinity integrated graphics core is quite an amalgamation of features across different GPU architectures. Not to forget, highlight features that were first introduced in the Llano’s graphics cores are also present in Trinity such as AMD SteadyVideo, CrossFireX support with an optimal similarly classed Radeon HD mobile GPU module, and Switchable graphics technology. They work pretty much the same as we last discussed in our mobile Llano platform article. For recommended compatible mobile graphics modules, AMD lists that their entire A10, A8 and A6 series of APUs can easily work with these following “Thames” GPU cores - Radeon HD 7670M, 7650M, 7550M, 7510M and 7450M (Seymour). Of course, this guidance is for the system integrators as consumers often have no say in the final product’s specification.

Power Optimization Features and Other Trinity Core Enhancements

As with any new processor (or APU), you can expect a slew of power optimizations that are important to their increased performance-per-watt proposition. We list some of the main optimizations here:-

• New Core C6 (CC6) power state enables powering down individual compute modules. This was present in the desktop AMD FX, but is new to the Fusion processors.
  
   
• Unified Northbridge for better APU power management (power gated when idle)
  
   
• Dynamic DRAM speed changes via memory P-state support
  
   
• Supports 1.25V DDR3 DIMMs
  
   
• PCIe and Display PHY controller power gating
  
   
• SVI-2 Voltage regulator interface for speedier voltage transitions and increase efficiency
  
   
• Video Compression Engine (VCE) to help reduce data footprint and hence reduced workloads that translate to power savings

Further to power optimization matters, AMD also tightened up how the different units within the Trinity APU share data efficiently with a new set of interfaces like FCL, RMB and IOMMU v.2 (the latter of which is key to supporting the latest features offered on discrete GPUs when paired with the integrated graphics engine):-

 AMD’s Mobile Trinity APU Competitive Positioning

By now you should have a clear idea of what is the Trinity APU about and its key differences against its predecessors. We listed a full list of processors launched on the first page and while you can roughly estimate where they rank, you probably may not be sure exactly who are they targeted at. Thankfully, AMD has the following broad slide that points out the estimated target market and key enablers that they bring about in their various classes of A-series APUs:-

Note that the AMD A6, A8 and new A10 class are likely candidates to populate the mainstream notebooks and the supposed Ultrathin notebooks that will compete with Intel’s Ultrabooks. And just like the latter, AMD too is following in their footsteps to prepare more premium labeling that will be used for the Ultrathin notebooks. To give you an idea what you might soon see in stores, here’ yet another slide from AMD:-

However we noted an issue with the branding as there’s no clear differentiation between a first generation AMD A8 APU and the second generation APU. While AMD mentioned of the updated logo color, this is quite subtle and most consumers will not notice it unless told. As such, for the next few months, consumers would be wise to check on the exact APU models that the system and notebooks posses in retail. The foolproof method is look out for the full processor model number such as the A8-4500M APU where the number "4xxx" denotes that it’s a Trinity class APU. Previous generation Llano APUs go by the "3xxx" numbering scheme.

So how exactly do all these second generation A-series APUs compete against the dominant mobile computing platform provider from Intel? Our benchmarks in the next few pages will soon reveal these from our personal testing. But before that, let’s take a look at a slide from AMD showing how they position their Comal platform (which is the AMD Trinity APU for mobile systems) stacks up against Intel’s counterparts.

Given that the compared slide takes Sandy Bridge processor in comparison, we feel that the current Ivy Bridge mainstream CPUs are an even tougher match for them. We estimate that the AMD A10 and A8 would end up competing against Core i5 and Core i3 notebook platforms.

As iterated earlier, a new market segment that AMD wants to target with the Trinity APUs is Intel’s Ultrabooks To tackle this, AMD has able to come up with more ultra low voltage (ULV) processor variety. In the current Comal platform, AMD has one rated for 17W TDP usage and should hopefully fit in an Ultrabook-like form factor, which AMD has coined as Ultrathin (imaginative, we know).

However, will these low-powered chips be able to keep pace with Ultrabooks equipped with Ivy Bridge CPUs? From the looks of things, our benchmarks pitting the AMD A10 APU against an Ivy Bridge equipped machine (HP Pavilion dv6) shows that the AMD chips lose out in pure processing throughput. However when it comes to gaming, the Trinity platform with its integrated GPU came out on top and did not disappoint (when not considering the comparison notebook’s discrete graphics performance.). So what does this all mean for consumers?

In the case where ultra-thin-and-light notebooks are concerned, since most of them do not have space for discrete graphics (unless they are 14-inches or larger), the capabilities of the integrated graphics are far more important and here’s where the AMD Trinity platforms can deliver. The story is a little different when it comes to proper multimedia class notebooks where there’s space to integrate a discrete graphics unit and if the price is right, and Intel CPU and discrete GPU combo is still the better outcome as you’ll seen in our test results later.

But of course, graphical and CPU capabilities are only one part of the story. The multimedia experience on the Trinity platform is also something that will make or break AMD’s “Ultrathin” ambitions. Thus there is a need to introduce the AMD HD Media Accelerator, which is a collection of AMD technologies that boost performance of multimedia tasks. These technologies include the AMD Perfect Picture HD (image and video processing), AMD Quick Stream (online video streaming) and AMD Steady Video (software video stabilization), and all that comes baked into the chip you buy.

To make use of all that exciting multimedia technology, AMD also worked with developers of popular software like Photoshop, VLC, Winzip and more to make sure these apps are written to take advantage of the technology available on the chips. The list of applications that support these functions as well as tapping on to the GP-GPU capabilities is still a growing one and is far from complete, but we’re glad to note that it has ballooned notably as compared to a year earlier.

It will take a while longer to reach a tipping point where almost every other application is written to take advantage of these technologies. Since AMD is adopting open standard APIs, the effort put in by the developers can also be harnessed by competing companies like Intel and NVIDIA. As such, we’ll have to revisit performance aspects all over again, but at least AMD will be in a better position competing with its GPU and CPU know-how than the competition.

Recently, we mentioned that HP was first to announce ultra-portable notebooks (dubbed Sleekbooks) that run on the Trinity platform but sadly won’t be reaching Singapore. However, there will still be other vendors producing AMD APU toting notebooks. In fact one of them will be sending one our way pretty soon. Until then, read on as we tackle how AMD’s reference Trinity notebook platform is able to compete with a variety of tough competitors we’ve got lined up. 

Testing the Reference Mobile AMD Trinity Platform

The new generation Trinity APUs will first populate the notebook segment this quarter, followed by the desktop variants which will likely be introduced sometime next quarter in Q3 2012. Using the mobile Trinity APUs, AMD hopes to spearhead the Ultrathin notebook market - a marketing term to compete against Intel's Ultrabook proposition. While this article is focused on the AMD Trinity APU's performance in the notebook platform, unfortunately what we've got is an AMD reference notebook platform based on the highest performing mobile APU launched today, the AMD A10-4600M. This is 35W TDP rated processor, which is very unlikely to make the grade for an Ultrathin notebook. We're fairly certain that only the low voltage and ultra low voltage variant based A10-4655M (25W) and A6-4455M (17W) would make the grade for these sleek class of notebooks.

Nevertheless, we figure that the high performance reference notebook platform would still be interesting to see if it can compete against Intel Ivy Bridge processor based multimedia notebooks and that's exactly what we're going to compare in this article. Not to forget, we're sure you'll be interested to know how much better the top mobile Trinity APU is ahead of the (almost non-existent) top mobile AMD Llano processor from last year that it supersedes - we've got that compared too. So read on as we pack these interesting comparisons over the next few pages.

Performance Benchmarking

We've managed to put the AMD Trinity reference notebook through a series of benchmarks and pitted the scores against last year's AMD Llano notebook platform, as well as a fairly recent Ivy Bridge notebook, the recently updated HP Pavilion dv6. Do take note that the Trinity reference notebook only makes use of its APU's integrated Radeon HD 7660G graphics. According to AMD, it should perform no different from a discrete mobile Radeon HD 7660M graphics unit and that sounds fairly encouraging to compete with our line-up. The one aspect that might throw off initial performance numbers is that the AMD Trinity notebook is equipped with an SSD drive while the rest of the comparisons are using normal hard disk drives.

Take note that the AMD Llano reference platform is equipped with a discrete graphics module and it even has an option of pairing up the discrete graphics with its integrated graphics core using the CrossFire option. Unfortunately though, we were unable to get more updated drivers to get CrossFire working. Since AMD's desktop drivers will also not work with the mobile Llano platform, we’ve just tested the Llano’s discrete and Integrated graphics capabilities separately.

While we're on the topic of graphics options in each platform, on the Ivy Bridge platform, we've got the HP dv6 that sports an NVIDIA GeForce GT 650M discrete graphics module. The notebook was tested with integrated and discrete graphics options for most scenarios.

To augment these various notebooks mentioned above, we've also decided to throw in another interesting comparison point with the Acer Aspire Timeline Ultra M3 - a notebook with a dual-core mobile Sandy Bridge processor and an NVIDIA GeForce GT 640M discrete GPU. It would be interesting to see how the top mobile Triniy A10-4600M APU copes with this combination.

PCMark 7

PCMark 7 tests a system’s overall performance. Our AMD Trinity notebook was a strong performer here, helped mainly by its excellent System Storage score thanks to its SSD drive. This would also influence most of the other results, so this benchmark isn't really ideal to figure out exactly the performance advantage provided by the Trinity APU. However, when put against a much more powerful quad-core, Ivy Bridge equipped HP DV6, the Trinity was still coping well, averaging a 15% lead in all tests - although again much of that is the result of the Trinity’s SSD compared to the DV6’s HDD.

The results are a little more interesting when compared against the Acer Timeline Ultra M3 as it uses an Intel Sandy Bridge CPU, discrete graphics and a hybrid storage drive. As a result of its improved storage subsystem, you can see it compete better with the SSD equipped AMD Trinity platform. In the end, AMD still has the upper hand, but we can't be totally certain until we get more notebooks to compare them on a more even playing field.

3DMark 11

Our AMD Trinity notebook is equipped with AMD’s Radeon HD 7660G integrated graphics processor, which features new improvements including second generation DirectX 11 support, an improved tessellation unit with enhanced thread management, improved support for Anti-Aliasing and DirectCompute - compared to the Llano platform's AMD Radeon HD 6620G integrated graphics that dates back to the Redwood GPU cores (Radeon HD 5500/5600-class but updated with a UVD3 engine). While IGPs generally trail behind discrete GPUs, as you can see from the results below, there are definitely improvements in this area as the Trinity’s Radeon HD 7660G managed to hold a 20% increase in performance compared to both the IGP and discrete GPU of the Llano platform notebook.

Needless to say, AMD still has an edge over Intel when it comes to integrated graphics, with the Radeon HD 7660G outperforming the HP machine's Intel HD Graphics 4000 by 40% on 3DMark 11’s Entry preset and by a staggering 80% on the Performance preset. In fact, Intel’s HD Graphics 4000 was only competitive with AMD’s previous generation Radeon HD 6620G graphics on the Entry preset, and still fell significantly behind on the Performance preset at 45%.

Against more powerful discrete graphics engines, the new AMD Trinity APU lost out to the NVIDIA GeForce GT 640 and 650M by a large margin (by 50% or more). Unlike discrete graphics that usually have their own speedy video memory, integrated graphics engines rely on the speed of the system memory. This discrepancy and because the benchmark also factor in CPU's performance is a reason for the gap noticed in this benchmark. On the next page, we find out how they fare in real world games

Gaming Performance

For a real-world gauge on just how the Trinity platform fares when running PC games, we've selected three games which we feel are representative and intensive enough to put the new APU to the test.

Far Cry 2

While all of our test machines are capable of DX11 gaming, we starting off with our trusty DX9 Far Cry 2 title to see how things stand. As expected, the HP DV6’s discrete GPU GeForce GT 650M was way out in front, followed by the Acer's GeForce GT 640M. Among the compared IGP units, we were happy to see the Trinity’s HD 7660G attaining playable frame rates at both Medium and Very High settings and also showing a significant performance increase compared to the Llano, almost doubling its average FPS. It was also only trailing  behind the discrete GeForce GT 640M by about 30%, proving that the new APU is capable of matching low to midrange discrete graphics units.

While the DV6’s Intel’s HD 4000 graphics performed better in an actual game (compared to its 3DMark 11 scores) it still trailed the Trinity APU by some 30%, and was unplayable at Very High settings. By comparison, the Trinity suffered in some scenes with lots of foliage or explosions, but was generally still playable even at the Very High quality setting.

Battlefield: Bad Company 2

In our desktop GPU testing, Battlefield: Bad Company 2 has traditionally favored AMD graphics with its DX11 Frostbite 1.5 engine particularly good at stressing a GPU’s geometry shading power. This trend seemed to continue to the IGPs, with both the AMD Trinity's Radeon HD 7660G and previous generation Llano's Radeon HD 6620G IGPs holding a sizeable lead over Intel’s HD Graphics 4000. In fact, the Radeon HD 7660G performed twice as well as Intel's HD Graphics 4000.

Overall, the AMD Trinity’s Radeon HD 7660G was again impressive, being the only IGP to attain playable frame rates at both Medium and High settings. Even the GeForce GT 640M wasn't really any better as seen in the Acer M3 machine's results.

Dirt 3

In Dirt 3, both the DV6’s Intel HD Graphics 4000 and Trinity’s Radeon HD 7660G graphics were playable at medium settings, although Intel’s IGP was noticeably choppier. By comparison, the Trinity APU remained relatively smooth, with an excellent average of 49.2 FPS, a lead of about 38%.

At High settings, only the Trinity IGP was able to attain playable if somewhat laggy frame rates at just under 30 FPS. By comparison, all other IGPs, as well as the Llano’s discrete HD 6630M graphics were unplayable at High settings. This is despite the fact the new APU managed nearly 50 to 70% better performance than the Llano APU.

CPU-Intensive Performance Benchmarking

The previous two pages basically tested the overall performance of the Trinity APUs and stressed on their graphics performance. How would it fare when put to test on CPU intensive tasks? We have two benchmarks to help relate these aspects with a rendering test and a  video encoding test. On these tests, we included the Samsung Chronos 7 which uses an Intel Core i7 Sandy Bridge processor in lieu of our Acer M3 machine that wasn't available.

Cinebench

The Cinebench CPU score is derived by putting the notebooks through a test that forces the machines to render an image using the CPU's raw crunching power. Scores are then allocated based on how fast the image appears (the higher, the better). In this case, you can clearly see that the Ivy Bridge processor is head and shoulders above the processing power of the AMD APUs. Even the Sandy Bridge processor was well ahead, but we sort of expected this outcome. The scores definitely indicate the Trinity platform is only slightly more powerful than the Llano platform in terms of CPU horsepower and the bulk of the APU's capabilities went into integrating a capable integrated graphics core. The renewed Bulldozer core within the Trinity APU didn't really seem to matter here as opposed to the Llano's much older tweaked AMD Phenom II architecture.

Handbrake

Handbrake isn't strictly a benchmarking tool, as it's actually a popular program used to encode movie files for use in other devices such as tablets and handsets. However the time it takes to encode a video file is highly indicative of the CPU's broad performance. Hence we have set up the test to encode a large video file, and have the process timed. When it comes to encoding files, it turns out that the top tier Ivy Bridge platform is still very much ahead of a top tier Trinity platform and the same goes for the previous generation top tier Sandy Bridge processor too.

Take note though, AMD has shared with the media that the next latest revision of Handbrake will come with OpenCL acceleration and that should help it take advantage of the built-in GPU to accelerate such encoding tasks. It is one of the many applications that is is receiving GP-GPU updates in time to come that should give consumers a much needed boost to take advantage of the CPU's and GPU's capabilities combined (and hence this is what the HSA proposition is about). Note that the open nature of the API means that Intel and NVIDIA can benefit from these updates too, but it would then depend how capable are their respective graphics engines. We'll probably run another story when we've more application to showcase these differences brought about.

For now, the Trinity APU is slower than its competition, but it shouldn't be a very big disparity if compared with more mainstream Core i3 and Core i5 processor which the Trinity is designed to go up against.

Performance and Benchmarking

Battery life has thus far been the bane of AMD's notebooks as they've been trying hard to match Intel's fairly impressive notebook up-time figures. From our testing with Llano and now the Trinity APU, we think that AMD has made sufficient efforts to be able to compete well with the competition. However, because both the AMD notebook systems in our hands are reference platforms, it doesn't quite paint an accurate picture, nor do we have appropriate comparisons at this point of time. We'll have to make-do with the results we've now, but it won't be long before we get actual retail notebooks to paint you an accurate representation of Trinity's battery performance.

Perhaps the most interesting battery life comparison is when we get AMD's 17W TDP Trinity processor based Ultrathin notebooks to compare against Intel's Ultrabooks. This is perhaps the most important battleground that AMD has to compete against Intel, but we guess we'll have to save that showdown for another day.  For now, we'll have to compare against a mixed variety of notebooks that we have at hand.

PowerMark

To test the battery life of each platform, we used a new Powermark benchmark to determine just how the Trinity platform stacks up against Intel's top-tier Ivy Bridge processor. We ran the test in "balanced" mode, which basically ran a series of programs ranging from internet browsing, multimedia encoding, movie playback and more which simulate typical notebook usage. Here, you can see that the AMD platforms have more staying power, but however do bear in mind that the dv6 has a much better screen, more powerful speakers and most importantly a far more powerful processor.

We're not exactly content with this outcome because AMD claimed that the Trinity APU system is able to provide an hour more battery life than the previous generation. In our practical testing, that was totally proven incorrect from their own reference platforms.

We're still gathering results from this rather new benchmark, but the balanced mode that we use to run the test seems to produce results that are very similar to our usual video looping test. As such, you might want to also refer to the results on the Acer Timeline Ultra M3 machine's review where we compared a wider variety of Intel notebook systems - from Ultrabooks to multimedia machines.

The results so far seem to indicate that the AMD Trinity and Llano platforms are managing pretty decently (despite the disappointment noted earlier) and can compete on battery life aspects, though not setting new records. One interesting note to consider here is that AMD graphics engine's various video enhancement technologies like utilizing the UVD 3.0 and Video Codec Engine (VCE), to supporting features like AMD Steady Video technology and more do come into play during these tests where applicable. This means that some extra processing does take place unless otherwise these features are disabled in the drivers. Perhaps by doing so, the AMD platforms can pull in even better battery life and compete better against Intel platforms that don't perform these extra processing tasks and are software dependent. We'll also be taking a closer look at these aspects in a dedicated article.

Power Consumption

As highlighted above, the HP dv6 is running a much more powerful platform and hardware specs that consume more power and the outcome below is not surprising. However, if you were to compare with the Acer M3 Ultrabook running a GeForce GT 640M GPU and more comparable specs overall, you can see that the AMD platforms are almost evenly matched with it and that's a great sign. We'll find out if it these nice numbers reproduce themselves when we tackle actual retail notebooks featuring AMD's Trinity APUs.

Concluding Remarks

In a nutshell, AMD's new Trinity APU's performance results are rather interesting and strike a positive note. Certainly the mobile Trinity platform can easily claim 50% or better graphics performance over the Llano predecessor while maintaining similar battery life performance stats. That means the Trinity's performance-per-watt ratio is much improved and can easily compete with some of Intel's Sandy Bridge and even Ivy Bridge equipped processors - with low-end discrete graphics. However, if you're relying on CPU-intensive tasks like rendering, image editing and video encoding tasks, the Intel processors are still nearly two times faster.

Looking ahead, with more applications being enhanced with OpenCL acceleration capabilities, Intel's lead might become less dramatic in conventional CPU-heavy tasks. The level of advantage obtained from these new APIs to take advantage of compute capabilities on both the CPU and GPU will vary and the landscape is an evolving one. Plus, the differentiated hardware architecture design emphasis between AMD and Intel processors also don't yield an easy outcome. AMD's APU has allocated far more die space to its onboard GPU and the results show that clearly; the opposite is true for Intel where the graphics engine, though well integrated within the CPU now, is still more of an afterthought as it still holds the bottom performance bar. Raw processing power on the Intel processors is still far higher.

We'll have to investigate how the performance of these processors evolve in newer applications and with more comparable platforms. Notebook based systems aren't the easiest to perform platform comparisons unlike DIY systems where most parameters can be self controlled.

Performance Fit for Casual Users

What we've shown today is just a performance potential of the new APUs. But considering that what we've shown is for the top SKU, we're somewhat concerned how the other variants will stack up. If you were to peer into the results graphics results carefully, what the Radeon HD 7660G managed to do is make integrated graphics finally usable for a wide variety of games at low to medium graphics quality. We've not yet turned up the graphics detail levels to maximum nor push the resolution up - both of which would make tax the APU by a large degree to easily bog it down. And that's just for graphics performance. If we were to assume that the lower performing variant has general performance to rival the Llano platform, we reckon that the processor's capabilities are clear - a rung below that of Intel. So while it seems that the APU has progressed much, it hasn't made enough of an impact to change night to day.

But is it bad? From a general performance stand point, we think consumers will like what AMD's Trinity APU delivers. It certainly can satisfy mainstream non-demanding users better with its better graphics capabilities than Intel's iGPU-only system solutions. This would mean it can compete better in the entry to low mid-range machines that usually don't sport a discrete graphics module since the Trinity APU can offer better experience for this target group of users. Consequently, the value proposition would have to be strong and from what AMD has been singing, we hope they will be attractive and competitive. More discerning users would  still prefer the extreme performance potential possibilities with a quad-core Intel CPU and discrete GPU combination - it's still the ultimate combo for this year.

The other challenge that lies ahead is notebook and system vendors integrating them to their platforms. Will they be something that consumers look forward to? They've got to be appealing physically from the aesthetics point of view before consumers want to even check these AMD equipped systems. Furthermore, we're also concerned of availability matters. AMD has stumbled once with limited availability of AMD Llano processors, but surely they should be better equipped this time round with a tried and tested manufacturing process? At least that's what AMD would like us to believe too.

In the end, we'll have to wait out till more AMD Trinity APU based notebooks and systems arrive in the market and get evaluated before we can make a better conclusion of AMD's new platform. The Fusion strategy has served them will in the netbook market and we believe they will have equally enticing products for the Ultrathin and mainstream multimedia target group to make a better market push than the Llano platform could ever imagine. 

We can only hope their execution goes as planned and put pressure on the Intel-based machines. Remember, Intel's Ultrabooks aren't yet as affordable as the chip giant wants us to believe, but pressure from AMD should help in one way or another. Likewise, the same should apply to the other notebook segments too. Here's to a more exciting and dynamic notebook market for us consumers!

Our articles may contain affiliate links. If you buy through these links, we may earn a small commission.