A first look at AMD’s Mullins mobile APU

Three years ago, AMD released Zacate, a new APU architected from the ground up for low-power mobile systems. Every year since then, the company has rolled out an updated lineup of chips with that very same mission. The new arrivals this year are called Mullins and Beema, and they promise to deliver the best blend of performance and power efficiency seen so far in low-power AMD APUs.

We already learned a little about these products at AMD’s APU13 event back in November. A couple of weeks ago, AMD flew us down to its Austin, Texas campus for a presentation that revealed the missing pieces. We found out all about what makes these systems-on-a-chip different (and better) than their predecessors, Temash and Kabini, which were introduced last May.

AMD also gave us some hands-on time with a real, live Windows 8.1 tablet based on Mullins. We were able to run some of our productivity and gaming benchmarks on the system, and we were left with a good sense of how AMD’s next-gen tablet SoCs may compare to the competition. We’re still missing data on battery life, but the performance picture, at least, is fairly clear.

Without further ado, let’s delve into our first look at AMD’s Mullins APU—and its slightly more power-hungry sibling, Beema.

Introducing Mullins and Beema

Mullins and Beema are really two sides of the same coin. That is, the code names refer to the same silicon. AMD uses binning to differentiate Beema, which is aimed at thin-and-light notebooks, from Mullins, which fits inside the tighter power envelopes required for tablet designs. This is the same kind of two-for-one scheme we saw with Temash and Kabini last year—and with previous iterations of AMD’s low-power mobile APUs.

So, what’s changed since the previous generation? From a CPU and GPU architecture standpoint, not much. The diagram above refers to “Puma+” CPU cores, but Mullins and Beema are actually based on the same Jaguar CPU microarchitecture as their predecessors. No architectural changes have been made to boost instructions per clock, we’re told. AMD has nevertheless taken a number of steps to raise clock speeds when appropriate and to boost power efficiency. Some of those improvements extend to the integrated Graphics Core Next GPU. In all, the transistor count has risen from 914 million to 930 million transistors due to “low level changes.” AMD has also made enhancements to the memory and display interfaces, and it’s enabled an ARM Cortex-A5 core that was present but disabled in Temash and Kabini. This ARM core underpins Mullins and Beema’s Platform Security Processor (PSP), which sets up an isolated execution environment for secure applications.

We’ll look at the PSP more closely in a minute. For now, let’s talk a little more about those performance and power-efficiency tweaks.

Thanks to a mix of circuit-power optimizations and process-scaling improvements, Mullins and Beema leak substantially less power than the previous generation. AMD claims to have achieved a 19% leakage reduction across the CPU cores and a 38% leakage reduction in the integrated GPU. Less leakage means less energy wasted as heat, which in turn means AMD was able to raise clock speeds within the same thermal envelopes. The fastest 15W chip from last year’s Kabini lineup ran at 1.55GHz with a 500MHz GPU speed, but the fastest 15W Beema chip runs at 2GHz, can hit 2.4GHz with Turbo Core, and clocks its GPU at 800MHz. That’s quite impressive, considering both processors are manufactured on the same 28-nm process.

Speaking of Turbo Core, Mullins and Beema make more extensive—and more intelligent—use of that feature than their predecessors. While Temash and Kabini supported Turbo Core in the silicon, only one member of the lineup actually implemented it: the A6-1450, which used Turbo to push its CPU clock speed from 1GHz up to 1.4GHz. In that implementation, Turbo didn’t exploit the chip’s temperature monitoring capabilities. AMD instead estimated power consumption using digital activity counters.

Mullins and Beema do things a little differently. Turbo Core is enabled in four of the seven initial models, and it allows for much greater gains than before. For example, Turbo can push the fastest Mullins variant from its base speed of 1.2GHz to a maximum of 2.2GHz. Temperature monitoring comes into play this time, but in a way that differs somewhat from what we’ve seen on other AMD processors. Mullins and Beema’s temperature-based clock speed and voltage regulation even has its own name: STAPM, short for Skin Temperature Aware Power Management.

The theory behind STAPM is simple. The silicon is capable of running at up to 100°C without any reliability problems, but sustaining that temperature won’t do in a tablet, where the heat will propagate from the chip to the chassis and will eventually burn the user’s fingers. Rather than lower the maximum operating temperature, AMD lets the chip run at up to 100°C—but it uses an algorithm to estimate the temperature of the tablet’s “skin,” or outer chassis, and it ramps down the clock speed and voltage to prevent the skin temperature from getting too toasty.

STAPM involves a “sophisticated . . . silicon temperature tracking algorithm,” AMD tells us, but there’s no actual sensor to keep track of skin temperatures directly. Skin temperatures are estimated indirectly using “extensive modeling of different system parameters,” from screen brightness to fan speed. Through that modeling, AMD says, “We can predict what the skin temp would be based on everything we already know is going on in the system.”

The payoff of STAPM, according to AMD, is that the processor can operate at a higher clock frequency for anywhere from four to 20 minutes while the skin temperature slowly creeps up. Since many CPU-intensive productivity tasks involve bursts of activity with long stretches of idle time, STAPM leads to “higher performance most of the time,” AMD says. The company claims a performance increase of as much as 63% in “key workloads.”

That gain need not come at the cost of battery life, either. On the contrary, a faster-running processor can complete tasks more quickly and thus spend more time at idle, which “actually saves energy in many common use-cases,” the company tells us.

STAPM works in combination with another scheme, Intelligent Boost, to maximize power efficiency. As part of Intelligent Boost, the power management micro-controller “tracks application behavior [in] real-time to determine frequency sensitivity,” and the clock-boosting behavior is “adjusted accordingly.” Translation: clock speeds only go up for applications that stand to benefit. In situations where a higher clock speed wouldn’t lead to a substantial performance gain, clock speeds stay low, and less power is consumed.

 

The PSP and other improvements

Mullins and Beema’s Platform Security Processor, or PSP, was already inside last year’s Temash and Kabini silicon, but AMD waited until this generation to enable it—purportedly out of a desire to “more closely align with partner timelines and industry readiness.” The PSP is made up of a 32-bit ARM Cortex-A5 processor core, a cryptographic co-processor, dedicated ROM and SRAM, some extra logic to enable secure booting, and a local memory interface that allows access system memory and resources.

In a nutshell, the PSP establishes an autonomous (and programmable) execution environment for secure applications. The nitty-gritty of it eludes us somewhat, but AMD provided this helpful summation:

We have and are working with a partner to enable the security kernel, which we then built into an enabled reference platform. That reference platform was then provided to [software vendors] with support from the security kernel partner and AMD to enable [software vendors] to adapt their software to use the hooks required to enter the secure environment. X86 applications dive through [the] Trusted Application Environment to get access to the ARM core.

The diagram below provides a higher-level overview of how the PSP comes into play:

Non-secure applications run on the x86 cores, while secure applications like anti-virus software, online banking, and biometric authentication can execute code on the PSP. Any malware running on the system should be constrained to the x86 environment, and it should therefore be unable to tamper with or hijack whatever data the PSP is processing.

AMD says this design enables “enterprise class security” and leverages an industry standard, ARM TrustZone, that also works on ARM-based smartphones. AMD expects at least four or five applications to leverage the PSP by the end of 2014. We’re told the PSP will eventually become “pervasive” across the company’s product line, as well. That should give developers an added incentive to support the feature in their applications.

Mullins and Beema have other tricks up their sleeves, too, in addition to the PSP and the new Turbo mojo.

On the memory controller front, faster DDR3-1866 RAM is supported on Beema variants aimed at mainstream notebooks, and AMD has added low-power mode that’s “optimized for [the] lowest power DDR3-1333.” That mode is supposed to cut power draw by 500 mW. Mullins and Beema don’t support LPDDR3, like Intel’s Bay Trail SoC, but AMD claims it’s “getting most of the gains” of LPDRR3 with its new low-power mode.

Also, the company has used voltage-mod logic to reduce the power consumed by the display interface. AMD estimates the savings from that tweak at 200 mW for high-resolution panels.

Finally, there’s Windows’ connected standby mode—a critical feature for tablets that went untapped by Temash and Kabini. AMD tells us Mullins and Beema “can support connected standby.” However, the company “has not seen high demand from customers,” and it expects “the additional cost associated with the feature to impact how many total systems ultimately implement the feature.” That cost stems from the various platform requirements associated with connected standby.

Our sense is that, while there are no technical obstacles to developing a connected standby driver for Mullins and Beema, AMD isn’t feeling much pressure to do so right now. That could be because the company hasn’t been able to secure tablet design wins for Mullins yet. Connected standby allows tablets to provide always-on connectivity in the same way that smartphones do, so one would imagine tablet makers would be eager to implement the feature, even if there is an added cost. Earlier this year, a CNet News story attributed the lack of 64-bit Bay Trail slates to the prioritization of 32-bit connected standby drivers by Intel and Microsoft.

The Mullins and Beema lineup

As we mentioned earlier, Mullins and Beema are based on the same silicon and differentiated through binning. AMD tells us the chips are “very carefully harvested and separated to fit their target market.” The launch lineup consists of seven offerings in all: three that fall under the Mullins umbrella and four higher-wattage models that bear the Beema code name.

Model Radeon SDP TDP Cores CPU

clocks

L2

cache

size

GPU

ALUs

GPU

clocks

Max

DDR3L

speed

A10 Micro-6700T R6 2.8W 4.5W 4 1.2/2.2GHz 2MB 128 500MHz 1333MHz
A4 Micro-6400T R3 2.8W 4.5W 4 1.0/1.6GHz 2MB 128 350MHz 1333MHz
E1 Micro-6200T R2 2.8W 3.95W 2 1.0/1.4GHz 1MB 128 300MHz 1066MHz

Note the new branding. AMD has tacked “Micro” between the series indicator and the model number to demarcate Mullins. That’s a little clearer than what we saw in the previous generation. Temash and Kabini were only differentiated by the first digits of their model numbers.

Interestingly, the lowest-wattage Mullins model actually has a higher TDP than the lowest-wattage chip from last year’s Temash series. However, Mullins has a lower SDP, or scenario design power. In practice, AMD says, Mullins will be able to power high-performance fanless tablets, which Temash could not.

Mullins can also run at much higher clock speeds than Temash, thanks to Turbo and STAPM. Temash’s processor cores maxed out at a 1GHz base speed and a 1.4GHz Turbo speed, but Mullins can range up to 2.2GHz. Perhaps not coincidentally, the slowest member of the Mullins series has the same CPU clock speeds as the fastest Temash chip.

Model Radeon TDP Cores CPU

clocks

L2

cache

size

GPU

ALUs

GPU

clocks

Max

DDR3L

speed

A6-6310 R4 15W 4 2.0/2.4GHz 2MB 128 800MHz 1866MHz
A4-6210 R3 15W 4 1.8GHz 2MB 128 600MHz 1600MHz
E2-6110 R2 15W 4 1.5GHz 2MB 128 500MHz 1600MHz
E2-6010 R2 10W 2 1.35GHz 1MB 128 350MHz 1333MHz

Turbo Core isn’t as prevalent among these Beema processors—only the fastest of them supports it. Nevertheless, clock speeds have gone up, as has power efficiency. The fastest member of the mobile Kabini family runs at 2GHz, but it’s saddled with a 25W thermal envelope. The new A6-6310 reaches the same base speed and runs even faster thanks to Turbo, all within a tighter 15W TDP.

We didn’t get a chance to benchmark Beema, but we do have Mullins results on the next page. Press on to see them.

 

The Discovery tablet and the test

At AMD’s campus in Austin, we were given a few hours to test the Discovery Project tablet, an 11.6″ slate powered by the fastest Mullins variant. The device featured a 1920×1080 display resolution, 2GB of memory, a 64GB solid-state drive, Bluetooth, Wi-Fi, GPS, and all the other bits and pieces one would expect to find in a Windows 8.1 tablet. AMD described the Discovery tablet as a “fully featured product with everything Mullins has to offer.” This was no product, of course. It was an “internally developed” reference machine intended as a showcase for the press and AMD’s partners.

Still, it worked well, and it didn’t look half bad.

The system was set up with Windows 8.1 and a number of applications intended for us to test. We didn’t use the AMD-provided benchmarks, though. Rather, we came armed with a USB 3.0 solid-state drive containing our own test apps and games. We brought the latest versions of 7-Zip, LuxMark, Musemage, TrueCrypt, the x264 encoder, 3DMark, BioShock Infinite, DiRT Showdown, and Tomb Raider. We also ran the latest SunSpider and Kraken web benchmarks.

We didn’t have time to install a clean version of the operating system, but we did check the “Uninstall” and “Power Options” control panels, among other things, to make sure the test conditions were as clean as our schedule would permit. We also configured the operating system the way we usually do, disabling things like Windows Defender and System Protection, to ensure fair and comparable results.

For comparison, we ran the same tests on Asus’ Transformer Book T100 convertible tablet, which is based on Intel’s Bay Trail processor, and the Kabini whitebook AMD sent us last year. (A small caveat: while the AMD systems both ran Windows 8.1 x64, the Transformer was stuck with Windows 8.1 x86, since it lacks 64-bit support.) For our web and 3DMark tests, we also threw in data from Google’s second-gen Nexus 7 slate and from Nvidia’s Shield handheld, which both ran Android 4.4 KitKat. The Shield may not be a tablet, but it provides a glimpse at the unbounded performance of Nvidia’s Tegra 4 processor, which is helpful as a point of reference.

The results

Memory subsystem performance

The Mullins tablet had the least memory bandwidth of the three systems we tested. That’s not too surprising, because the A10 Micro-6700T is limited to a single channel of 1333MHz memory. By contrast, the A4-5000’s single-channel controller supports faster 1600MHz RAM. The Atom Z3740 is limited to a 1066MHz memory speed, but its second memory channel more than makes up the deficit.

Productivity apps

Low memory bandwidth or not, Mullins performs pretty nicely in these productivity tasks. The A10 Micro-6700T actually matches or outperforms the A4-5000 despite its lower base clock speed and tighter TDP. Credit for that performance should probably go to the new Turbo mechanism, which can push the Micro-6700T from its base frequency of 1.2GHz to as much as 2.2GHz—well above the A4-5000’s 1.5GHz maximum.

Mullins fares quite nicely against the Atom, too. It’s faster in x264 and TrueCrypt, and it’s not very much slower in 7-Zip. Of course, the Atom does have a lower SDP: 2W, compared to 2.8W for the A10 Micro-6700T.

Web browsing

Mullins excels in JavaScript-laden web browsing—a critical application for tablets. Keeping an eye on CPU clock speeds while running these tests, we saw the Micro-6700T hover between 1.8 and 1.9GHz. Those speeds were high enough to give this system-on-a-chip a sizable edge over the A4-5000.

 

The results—continued

GPU computing

Thanks to its integrated GCN graphics processor, Mullins comes prepared for GPU computing workloads. We put those capabilities to the test in LuxMark, a ray-tracing benchmark, and in Musemage, a GPU-accelerated photo editing application.

Mullins is far more competent than Bay Trail in these GPU-computing tests. However, this is one scenario where it fails to catch up to the 15W A4-5000—despite the fact that the two chips have the same maximum GPU speed. In all likelihood, Mullins’ more limited memory bandwidth is to blame.

Graphics and gaming

We’ll start our graphics testing with 3DMark, which has the advantage of running on not just Windows devices, but also Android ones. That provides some very helpful context. Note that, while some of the devices below have different screen resolutions, 3DMark is designed to compensate. It renders scenes offscreen at a fixed resolution (1280×720, in the case of the “Ice Storm” test) and then scales the image to match the display.

No doubt about it, Mullins has a lot of graphics horsepower for a tablet SoC. It’s faster than Bay Trail, and it’s even speedier than the Tegra 4 chip in Nvidia’s Shield handheld, which runs virtually unrestrained thanks to the Shield’s large chassis, active cooling system, and beefy battery.

Again, though, Mullins doesn’t quite match the performance of the 15W A4-5000.

Next up: a few games from our Steam library. We ran BioShock Infinite, Dirt Showdown, and Tomb Raider at 1280×720 using the lowest detail settings available.

No real surprises here. The numbers echo what we saw above: Mullins is obviously more capable than Bay Trail, but it’s slower than Kabini, notably so in some cases.

More to the point, Mullins isn’t quite fast enough to run these recent PC games at playable frame rates, even with the detail and resolution turned down. DiRT Showdown is the smoothest of the bunch, but 30 FPS is still a little sluggish for a racing game. On a Mullins tablet, you’d probably want to stick with older releases, casual or indie titles, and whatever tablet-friendly games are available on the Windows Store.

 

Conclusions

Spending time with the Discovery tablet has answered many of our questions—but it’s also left us with many new ones.

We learned that, for a tablet processor, Mullins is quite fast. It performs comparably in many tests to last year’s A4-5000, a 15W part that landed in mainstream notebooks, and its graphics performance outpaces that of the competition from Intel and Nvidia. (To be fair, the Atom Z3740 we tested isn’t the fastest Bay Trail incarnation—but it does have the same GPU speeds as the flagship Z3770.)

Mullins’ performance would be a great asset for any convertible tablet designed to double as a productivity machine. The performance we saw may not be sustainable for prolonged workloads, since AMD’s STAPM mechanism ramps clock speeds during a relatively small window of time. Then again, as AMD pointed out, many CPU-intensive productivity applications involve short bursts of activity with long idle periods. Mullins should do very well in those cases.

Of course, to be fit for convertibles, Mullins also needs to allow for good battery life. We still have no clear sense of whether it will do so. AMD has told us that Mullins will be “competitive” in that respect, which suggests some level of parity with Bay Trail. But in the absence of hard numbers, it’s difficult to make any predictions.

We also don’t know how successful AMD will be at scoring tablet design wins this year. Last year’s Temash chip was a no-show in Windows 8 and 8.1 slates from major vendors, and AMD wouldn’t say how widely it expects Mullins to be adopted. The company did, however, concede that it will be a “challenge” for it to “be highly visible in a larger number of SKUs.” Connected standby support could factor in there, but one thing is certain: AMD is facing extremely aggressive competition. It’s no secret that Intel has been carving out market share for its Atom chips using so-called “contra revenue” subsidies. Those subsidies might well be the biggest impediment to Mullins’ success.

Happily, Beema may have an easier time working its way into the marketplace. Lenovo has already announced several laptops based on it: the Flex 2 and the B and G series. Those systems are coming in June, and they’ll feature a special, A8-branded version of Beema that AMD offers “for select opportunities.” (That model isn’t part of the standard Beema lineup.)

Additional design wins, if they come, will likely be announced around the time of the Computex trade show in early June, with availability to follow in the summer.

Comments closed
    • Theolendras
    • 6 years ago

    First step to the accomplisment of my bold prediction : PS vita successor will be a PS4 portable.

    • anotherengineer
    • 6 years ago

    So, what will we see first, notebooks with Beema, or notebooks with Kaveri??

    Kabini 2Ghz quad was an option for a notebook, but I guess it’s back to the waiting game 😉

    • USAFTW
    • 6 years ago

    Looks like a very nice design. Here’s hoping it materializes.

    • Lans
    • 6 years ago

    I do wonder how well STAPM will do in practice… Say can it let system run at max boost speed all the time because it seems possible for off the shelf laptop cooler pads to dissipate 5W TDP + maybe another 5W of heat from other parts… More importantly, I think it needs to just work and not require any additional effort from device makers.

    Promising preview but waiting for equally well written for follow up.

    I am tempted to subscribe but just can believe average is so high. I guess I’ll bite the bullet when I am not so busy and can spend more time reading TR. 🙂

    • not@home
    • 6 years ago

    There is a cheese factory near where I live called “Mullin’s Cheese.” Man did this article make me hungry. All I read was “Cheese. Cheese. Cheese.”

      • Haserath
      • 6 years ago

      You made me think of fondu. Cheesesink.

    • cmrcmk
    • 6 years ago

    All I want is Beema or even Kabini on a motherboard featuring 6+ SATA ports for a home file server. Is that too much to ask?

      • mczak
      • 6 years ago

      Yes it is. The chip only has 2 sata ports. Someone could of course add pcie sata adapters (there isn’t actually enough pcie connectivity to get max performance for 4 additional sata 6gbps ports), but this doesn’t fit well into the ultra low cost category…
      Actually you could have enough bandwidth if you’d sacrifice the pcie 4x port. I think the market for such a board just is way too small. You can get something fitting the bill from intel, the silvermont server version (for instance the avoton atom c2550) have a lot more connectivity than their mobile siblings (among 6 sata ports, 2 of them sata III). They also have a lot more pcie lanes, which you can put to good use for an additional USB3 controller you’d no doubt want since it only supports USB2… It also has 4 ethernet interfaces built-in. Doesn’t fit the ultra low cost category though.

        • cmrcmk
        • 6 years ago

        For a home file server, I’m just wanting something with a modest cost and low power consumption. Raw performance isn’t important for something that will be idle 99% of the time. There are a couple of ITX boards with Bobcat CPUs that offer 4 SATA ports, but that’s just not very future proof in a system designed for optimal TB/$. Swapping out hard drives for higher capacity ones is way more expensive than adding an additional hard drive.

          • mczak
          • 6 years ago

          Brazos/Ontario (with Bobcat cores) wasn’t really a SOC and required a dedicated Southbridge (Hudson FCH). That’s why it still could support 6 SATA ports “natively” and hence 6 sata ports only really added the cost of the connectors themselves. But these newer chips (from both intel and amd) are primarily intended for tablets where even 2 sata ports is overkill (in fact the z atoms have them disabled afaik, only the non-atom branded parts enable them).
          You can however get Kabini (AM1) boards with 4 SATA ports, they aren’t all that much more expensive (they tend to be of the generally better equipped variety which means you might also get 2 additional USB 3.0 ports). If you still need more you could add a pcie sata card (which probably isn’t cost effective).

        • shank15217
        • 6 years ago

        i think they support port multipliers too.

    • drfish
    • 6 years ago

    Gah, it just irks me so much that we have awesome chips like this but the laptops they get put in are just miserable 15.6″ 1366×768 pieces of crap. When will someone take this low TDP “good enough” hardware that we’ve had for awhile now and put in into a machine that’s [i<]actually desirable[/i<]?

      • HisDivineOrder
      • 6 years ago

      As soon as people are willing to pay for it (i.e., never).

        • drfish
        • 6 years ago

        You don’t think there’s a market for a 11.6″ MacBook Air build quality machine based on Mullins/Beema/Bay Trail for somewhere in the $500 range? People pay $500 for a 16GB iPad and the Asus T100 is $400 – there has to be a place for such a beast…

          • HisDivineOrder
          • 6 years ago

          MacBook Air build quality requires Macbook Air build cost. That cost is NOT really about the CPU or even the GPU. It’s about the BUILD QUALITY. Most of the “quality” parts out there, including the panels, are not built in sufficient quantity to make them cost efficient to price in any laptop around $500.

          Because few companies can pump out enough laptops to make commodities of scale to work in their favor, Apple gets a favorable price on otherwise custom, high end materials that would cost other companies far more to use. And most of them won’t even bother to TRY to get those custom parts, but stick with the parts that are widely available.

          What’s widely available is NOT “Macbook Air build quality.” That’s true of displays, of keyboards, of the shells, etc.

          That’s where the problem is. It’s not the cost of the CPU or GPU or motherboard. Essentially, it’s not the actual COMPUTER part of the laptop that’s the problem.

          It’s all the other things that make YOU think a given 15.6″ laptop from Best Buy isn’t “Macbook Air build quality.” Those are what make these things impossible for OEM’s to produce readily for less than $700 and at that price how many people are going to buy any?

          None. Especially not when you can go buy a $200 Toshiba laptop with a Celeron 1037U that performs “good enough” for the vast majority of users. Meaning everything YOU are talking about is a NICHE user, further working against commodities of scale.

            • drfish
            • 6 years ago

            I guess all I’m saying is that if Asus can make a $350 quad core convertible imagine what they could do with the same or similar guts and an extra $150 to build it with quality materials. I think something like that is what a lot of professionals would like and could feel good about recommending to others. Anytime someone asks me for advice about which laptop to buy all I can do is sigh and tell them that pretty much all the options out there suck for one reason or another. There’s just no excuse for that when we have the chips to make better stuff.

      • UnfriendlyFire
      • 6 years ago

      I came across a few 15.6″ laptops that had A4 APUs in them. And 4-6 hours of battery life.

      What…

      • Voldenuit
      • 6 years ago

      [quote<]Gah, it just irks me so much that we have awesome chips like this but the laptops they get put in are just miserable 15.6" 1366x768 pieces of crap. When will someone take this low TDP "good enough" hardware that we've had for awhile now and put in into a machine that's actually desirable?[/quote<] [url=http://arstechnica.com/gadgets/2014/04/hands-on-acer-disses-ultrabooks-while-updating-its-budget-pcs/<]Acer Aspire Switch 10[/url<] Convertible, dockable tablet PC with quad core Atom Z3745, Windows 8.1, and get this [b<]10.1IN 768 IPS display[/b<], starting from $379.99.

        • UnfriendlyFire
        • 6 years ago

        There’s an A4 APU laptop that is 15.6″ or 17″, TN 1366×768, 4-6 hours of battery life, and crap keyboard. And it cost a little below $400.

        It’s like if the PC manufacturers don’t even care if AMD dies.

    • ptsant
    • 6 years ago

    Looks like a very promising chip. I can’t easily extrapolate with other tablet devices but the fact that compares favorably with netbook-class hardware is “good enough” for me. I don’t care that much about tablets (already own an iPad and don’t plan to upgrade until it breaks).

    Beema is the real deal, especially for cheap 13.3″ notebooks and HTPC uses. If it delivers, it’s going to be a very big hit. I’ll be waiting for that review with impatience…

    • UnfriendlyFire
    • 6 years ago

    STAPM looks like the principle of “Race to Idle” taken to the extreme.

    On a side note, will tablet manufacturers use AMD’s chips this time? I don’t think I saw any of AMD’s APUs in any of the tablets I came across.

      • NeelyCam
      • 6 years ago

      AMD copied Intel. Again.

        • ronch
        • 6 years ago

        Intel has copied AMD quite a few times as well. AMD64, IMC, and … um… ah, a few more!

          • jdaven
          • 6 years ago

          I’ll help you out Ronch.

          Using model numbers instead of MHz
          Integrating the GPU and going after integrated GPU performance
          Adding cores instead of MHz

          However to be fair, AMD has copied Intel as well:

          x86 of course
          Socket to Slot and then back to Socket
          Lightning bolt to answer Thunderbolt

            • swaaye
            • 6 years ago

            Intel had an x86 CPU codenamed Timna with an on-die IGP scheduled for 2001 but they didn’t release it. Market conditions made it not make sense compared to chipset IGPs, I guess.

            [url<]http://www.cpu-world.com/forum/viewtopic.php?t=11663[/url<]

            • NeelyCam
            • 6 years ago

            [quote<]Integrating the GPU and going after integrated GPU performance[/quote<] Actually, Intel had the CPU and the GPU inside the same package before AMD did - I have one of those (Clarkdale). So, sorry, but integrating the GPU in the same package with the CPU was AMD copying Intel.

            • chuckula
            • 6 years ago

            Sandy Bridge… with the GPU fully on-die… was on the market for several months before Llano even launched too.

            • UnfriendlyFire
            • 6 years ago

            I have a laptop with an Intel GMA IGP.

            You come over here and try to play a 1080p youtube video. YOU CAN’T.

            • rootheday3
            • 6 years ago

            I don’t know what kind of IGP platform you have, but it must be pretty old.

            Intel IGPs on client platforms have run 1080 youtube through HW accelerated without issue since GMA4500 (aka Cantiga/Eaglelake chipset) in 2008. Older netbooks (prior to 2012) were not capable due to the fact that they had older architecture but Baytrail has no issues here and even Clovertrail has hardware acceleration for media that works pretty well.

            Sandybridge (HD2000/HD3000) could do something like 16 simultaneous 1080P H.264/AVC streams in parallel.

            • NeelyCam
            • 6 years ago

            That’s unrelated to this meaningless debate.

            • sschaem
            • 6 years ago

            I think he wanted to say , “yes Intel was first to put a GPU on the same package as the CPU, but it was nearly worthless, and AMD had to show Intel how to do it better” .. and so Intel did.

            • swaaye
            • 6 years ago

            Cyrix did the first on die x86 IGP.

            I don’t think things have really changed much over the years. Today’s IGPs are all still pretty ‘basic needs’. It’s not like they both couldn’t build something like what’s in PS4. It just doesn’t make sense for the mass market that x86 CPUs sell to. AMD and Intel just play little strategy games with the market, within a profitable transistor budget. AMD’s CPU tech stinks so they might as well try to peddle that HSA stuff via a somewhat bigger GPU.

            • rootheday3
            • 6 years ago

            Not sure on timing versus Cyrix, but Intel had a cpu/memory controller/gpu back in 1999 or so- check out [url=”http://en.m.wikipedia.org/wiki/Intel_Timna”]Timna[/url]

            it got tripped up by the RDRAM adoption issue and the memory translator hub to bridge to ddr had bugs that delayed it. But the thing worked.

            • swaaye
            • 6 years ago

            Yeah I posted about Timna already. 🙂

            Cyrix built a chip called MediaGX back in 1997. It was used in some very budgety PCs. It was a 486/Pentium-class CPU (Cyrix 5×86) with integrated audio/video functionality. I had a friend with one and it could barely run Diablo 1.

            • mesyn191
            • 6 years ago

            They couldn’t run regular Win98 either. You had to use a special version that Cryix paid MS to recompile just for them.

            Performance wasn’t too good even for the time period but they were very cheap compared to what AMD and Intel were selling that time.

            • mesyn191
            • 6 years ago

            Supposedly performance was terrible though + Intel would’ve lost money trying to sell it since the die size ended up being too large for the price point they wanted to hit. That and company politics was the reason why so many of the engineers working on the project ended up getting fired.

    • tbone8ty
    • 6 years ago

    Damage labs still needs to test AM1

    • Ninjitsu
    • 6 years ago

    “Beema” means “insurance” in Hindi, so i guess this is AMD’s insurance policy…

    • killadark
    • 6 years ago

    hmm this is weird im actually impressed by the performance they put out i came to this artical with a mindset that it sont be anywhere near the intel offerings but oh well look at that it even compares to last years 15w models well done AMD

    • deruberhanyok
    • 6 years ago

    [quote<]AMD claims to have achieved a 19% leakage reduction across the CPU cores and a 38% leakage reduction in the integrated GPU [...]The fastest 15W chip from last year's Kabini lineup ran at 1.55GHz with a 500MHz GPU speed, but the fastest 15W Beema chip runs at 2GHz, can hit 2.4GHz with Turbo Core, and clocks its GPU at 800MHz. [...] both processors are manufactured on the same 28-nm process.[/quote<] Same build process, no enhancements made to the architecture for IPC, but they've gotten a 33% CPU speed increase (60% if you factor in the boost) and 60% GPU and kept it in the same thermal envelope? I feel like "impressive" isn't a big enough word for that. I wonder if these sort of power-use optimisations can also be applied to their laptop/desktop/server chips? A major increase in CPU/GPU clock speed with zero increase in TDP would be a big boost for their APUs - or, alternately, a major drop in TDP with minor clock speed/IPC increases. I wonder how much these systems would benefit from more memory bandwidth? The tablet seems to be making the most of its single channel, so I wonder if it would scale even higher with a second channel?

      • Terra_Nocuus
      • 6 years ago

      imagine if they could get similar results on their discrete GPUs, they might have to try if 20nm Maxwell is as power efficient as the 750.

      • mesyn191
      • 6 years ago

      They probably got the gains by doing lots of hand tuning to a chip that was supposedly mostly or entirely designed via automated tools.

      AFAIK their desktop/server arch. (*roller) is already fairly hand tuned and actually does get fairly good clocks even on TSMC’s and GF’s bulk process. Neither of which are known for being all that great for high performance CPU’s which is probably why those high clocks come with such a high power cost too.

      What AMD really needs is a true clean sheet design that performs well at moderate clocks so they can get decent power consumption on TSMC’s and GF’s processes both of which will probably remain 2+ yr behind Intel for the foreseeable future. That probably means a relatively wide and complex CPU compared to *roller with large, complex, and low latency high performance caches. I wouldn’t be too surprised if they ended up doing what amounted to a x86 version of POWER8 but with much less cache and a lower TDP.

      IIRC something that sounded an awful lot like that was proposed years ago by one of the lead designers of the K8. I believe the scuttlebutt was AMD ended up rejecting that design and that guy left as a result of the decision.

        • stdRaichu
        • 6 years ago

        If I’m remembering correctly, a lot of the blame for the severely lacklustre performance of bulldozer was put down to an increasing reliance an automated layout tools (read: firing a bunch of the guys that did the hand-tuned layout). I’m wondering if any of this (frankly incredible) performance increase was due to the return of Jim Keller, one of the architects of the K7 and K8 and formerly of DEC Alpha fame – also a chip that got a huge chunk of its performance from rather incredible hand-tuning.

          • mesyn191
          • 6 years ago

          Its already been through several revamps now from the original Bulldozer core.

          • Voldenuit
          • 6 years ago

          IBM relies heavily on automated layout tools for its POWER CPUs, and it hasn’t stopped them from producing kickass server CPUs. Mind you, they are not cost, die size, or TDP constrained, though.

            • mesyn191
            • 6 years ago

            [quote<]Mind you, they are not cost, die size, or TDP constrained, though.[/quote<] Exactly. Not a apples to apples comparison.

      • derFunkenstein
      • 6 years ago

      On the flip side, how awful is Kabini knowing it apparently leaks so badly?

        • deruberhanyok
        • 6 years ago

        Heh, I wondered that, as well. If they’re able to get these enhancements out of it without tweaking the actual IPC, then yes, I think it’s a valid question to ask how much “overhead” from automatic design could have affected the chip’s performance/watt.

        Actually, I’d love to see an article about that, maybe TR could interview one of AMD’s designers?

      • sydbot
      • 6 years ago

      I wonder if AMD will bother (aka get paid) to power-optimize the Xbox One and PS4 chips. Maybe for the first shrink?

        • deruberhanyok
        • 6 years ago

        Maybe!

        Optimizations like this could possibly be done in lieu of a die shrink for XB1 and PS4, too, especially if they’re constrained by whatever GlobalFoundries/TSMC can do.

        If they can get these sort of boosts without increasing TDP, they could also lower TDP but keep performance the same, which would be ideal for a console.

    • Unknown-Error
    • 6 years ago

    AMD actually has a promising product. First time since Brazos. Looks like Jim Keller [b<]IS[/b<] moving things in the right direction. But what does AMD (I assume marketing) idiots do? Gives it a retarded name. [b<]A10 Micro-6700T[/b<] ??? You gotta be kidding me!

    • chuckula
    • 6 years ago

    [quote<]Of course, to be fit for convertibles, Mullins also needs to allow for good battery life. We still have no clear sense of whether it will do so. AMD has told us that Mullins will be "competitive" in that respect, which suggests some level of parity with Bay Trail. But in the absence of hard numbers, it's difficult to make any predictions.[/quote<] Yeah there's the kicker. It's not rocket science to build a chip that's faster than Baytrail. It is harder to build an ultra-mobile chip that's faster than Baytrail while being power competitive with Baytrail though.

      • ermo
      • 6 years ago

      I’d go so far as to say that it’s hard enough to build a mobile chip that is as fast as Bay Trail while being power competitive with it — on a larger 28nm process node.

      The most interesting thing for me was that the NVidia Shield Tegra 4 CPU/GPU combo appeared to be faster than the Z3740 Bay Trail. Does this imply that NVidia missed their chance with Tegra 4? As fast as Bay Trail clock-for-clock — but only while consuming more power? And to add insult to injury, it’s not x86 compatible either (which matters for productivity)?

      It’s not easy to compete with Intel, no matter how you slice and dice it.

    • ronch
    • 6 years ago

    Couldn’t they have waited until Beema is ready before rolling out their desktop parts for their AM1 platform? I mean, not only would it possibly make their new AM1 chips more compelling, it could also avoid potentially irking people who already bought AM1 chips and finding out just a few weeks later that there’s a new chip on the block that looks like what the chips they bought should’ve been in the first place.

    Also, considering how Beema ‘fixes’ Kabini’s shortcomings, as well as how Richland ‘unlocks’ Trinity, and how Piledriver is also what Bulldozer should’ve been from the beginning, it increasingly becomes clear that AMD always seems to leave so many things that need fixing with the first iteration of their chips. Sure, chip companies normally tweak their chips as they go along but in AMD’s case, it seems like it’s more like a matter of ‘fixing’ rather than ‘tweaking’ especially given how their chips look compared to the competition.

      • Fursdon
      • 6 years ago

      I imagine a part of the reason for the AM1 platform is to disperse all their unloved, watt-chewing bin Kabini chips out of their inventory. I would guess we’ll see Beema in some socketed form possibly later this year, probably next. Here’s to hoping they share the same socket too.

      • faramir
      • 6 years ago

      Don’t you worry about Beema on AM1 – you won’t be able to buy one in a socketed form until 2015 at the earliest.

    • NeelyCam
    • 6 years ago

    I like this.

    “Look, ma – no fans!”

    • Vulk
    • 6 years ago

    Don’t those contra-revenue strategies Intel is using fly directly in the face of the settlement they signed with AMD just a few years ago? I know that depending on how you slice the market they’re competing against ARM, but it would seem that if they’re freezing out the only other x86 competitor by selling at a loss, then they’re violating several anti-competition clauses they’ve agreed to here and in the EU.

    Not that billion dollar fines really seem to slow Intel down. But I’d expect to hear about lawsuits if they keep that up, especially in the EU.

      • Ninjitsu
      • 6 years ago

      I’m no lawyer, but from what I could get of their previous anti-trust case, the problem there was that they were paying manufacturers to use their chips instead of AMD’s.

      Right now, not only do they have very little market share, but they’re taking a loss to sell cheap, and it’s still manufacturers deciding.

      I think that’s why there’s no issue this time.

        • ronch
        • 6 years ago

        Whether they pay manufacturers to use their chips or give them heavy discounts or practically give the chips away for free, they all have the same bottom line: more money in the manufacturer’s pockets for using their chips, not counting profits from selling finished products. It just depends how you slice it.

      • UnfriendlyFire
      • 6 years ago

      I think the EU is going to look away since Intel is losing quite a bit of cash, espicially when they have such small marketshare.

      Now if Intel slashed the price of their laptop/desktop CPUs, especially the Iris Pros, by 20%-30%, then I think they might get attention.

      • Zizy
      • 6 years ago

      Well, if they dont have monopoly, they can sell stuff at any price they want. These are tablet CPUs and Intel does not have monopoly there, so, no problem.
      But IF Intel says something like “if you don’t take AMD CPUs, we will give you our stuff even cheaper” they might see another billion or two in fines even without monopoly. As long as they are merely selling them at such a discount nobody even asks for AMD, they should be safe.

        • UnfriendlyFire
        • 6 years ago

        But Intel could be accused of predatory pricing by severely discounting their laptop/desktop CPUs. Especially if they pushed AMD into bankruptcy and then hiked back up their prices.

    • jjj
    • 6 years ago

    So annoying that everybody is going with hardware DRM . You guys pretending not to notice and repeating their spin on it is not very cool either.

      • odizzido
      • 6 years ago

      hardware DRM?

        • The Dark One
        • 6 years ago

        Notice how Video Streaming was sandwiched between AntiVirus and Digital Wallet on the ‘secure’ side of page 2 AMD diagram.

          • odizzido
          • 6 years ago

          I don’t know anything about video streaming sites…..What are they trying to stop with that? Surely any legit site will require login info to stream and any non-legit site will just be able to grab whatever and offer it DRM free anyways?

            • Spittie
            • 6 years ago

            “Sorry, this content is not available for your your rusty pc – buy a Samsung smart TV now and enjoy unlimited* streaming!”

            • odizzido
            • 6 years ago

            Why would they put effort into trying to make it so you cannot watch the content they’ve created?

            • Spittie
            • 6 years ago

            Who knows? People are dumb. Netflix/Amazon and the likes already use DRM for whatever reason.

            • Ninjitsu
            • 6 years ago

            It’s moronic like DVD DRM. Bought a [i<]Band of Brothers[/i<] DVD set the other day, wanted me to switch my region settings to Region 2 or 4, which is Europe and South America. I'm in Region 5, not even sure why they're selling it here. And then people pirate and the lobbyists cry foul.

            • odizzido
            • 6 years ago

            I hate that region crap. I move countries sometimes and that makes things a real pain if I want to get DVDs or any other region locked BS. I don’t want to own multiple DVD players because some fat CEO on a yacht wants to try to make enough for another yacht. It’s theft. fuck him.

      • timon37
      • 6 years ago

      As I understand it (I might be wrong) the tpm itself isn’t necessarily “evil”, it’s the combination of it and proprietary software which makes it evil.
      Basically (among other things) it’s a method for companies to make sure you’re running what they want you to run and stop you from modifying the software. Thus you won’t be able to break any DRM/backdoors/etc they implant.
      It used to be a separate chip plugged into the mother-board now it’s part of the CPU, which is exactly what RMS and others predicted would happen (though it doesn’t take much wisdom to predict that;p).

      Some references:
      [url<]http://www.cl.cam.ac.uk/~rja14/tcpa-faq.html[/url<] [url<]https://www.gnu.org/philosophy/can-you-trust.html[/url<]

    • tbone8ty
    • 6 years ago

    i cant see any reason for manufactures not to spec these chips now.

      • NeelyCam
      • 6 years ago

      Mullins does look strong, but like the article says, we don’t really know how the actual battery life is. ARM Cortex A15s look strong, too, but eat the battery pretty fast.

      I wonder how long these will be in the market before Intel’s 14nm Cherry Trail comes out. Regardless of how well optimized the chip is, competing with something that’s two process nodes ahead [i<]and[/i<] is using FinFETs is really difficult from power efficiency point of view

        • Ninjitsu
        • 6 years ago

        Yeah they have a year…considering AMD has its own version of “Valve time” that may not be a lot…

        I mean, Intel showcased its Bay Trail FFRD at computex [i<]last year[/i<], even [i<]they[/i<] still don't have much (if anything) out for consumers yet.

          • NeelyCam
          • 6 years ago

          Asus Transformer (the one Mullins FFRD is compared to in this article) has been out since last November. If Mullins tablets make it to the market around next Thanksgiving, Cherry Trail tablets might be right around the corner.

            • Ninjitsu
            • 6 years ago

            Well, what I’ve been told is:
            [quote<]From a silicon perspective, from an SoC perspective, currently we're shipping Clover Trail, and we've started sampling Bay Trail since the end of last year...products are getting designed around Bay Trail even as we speak. And following this, the next SoC will be Cherry Trail, which will be the first half of 2015, the early part of 2015.[/quote<] Then he went on to explain how the phone silicon would play into this, with Merrifield based tablets slated for the first half of this year. In the first half they're also sampling Moorefield, with products based on it releasing in the second half. If you want the audio file, I could link it here, though i'm not quite sure if there are any rules against that (wasn't asked to honour an NDA).

            • NeelyCam
            • 6 years ago

            They’re probably talking about Bay Trail on Android ; I haven’t seen Bay Trail Android tablets anywhere yet, while Clover Trail ones are out (e.g., Dell Venue). But I know for a fact that Win8 Asus Transformer tablet has Bay Trail in it, and it’s been out for half-a-year (I bought one around Thanksgiving before promptly returning it).

            That said, please please please link the audio file!!

            • Ninjitsu
            • 6 years ago

            Hmmm. Quite possible, really. Yeah, come to think about it, I do remember Windows 8 Bay Trail tablets being announced. So must be talking about Android. I never asked. Should probably send across a follow up mail.

            The whole file is an hour long, but here’s the bit relevant to this:
            [url<]https://dl.dropboxusercontent.com/u/45160510/Interview.mp3[/url<] I'm a noob at this stuff, forgive me! Haven't interviewed people much.

      • HisDivineOrder
      • 6 years ago

      If you can have a seemingly amazing chip following in a long of seemingly amazing chips that didn’t pan out at all the way they originally seemed from a company teetering on financial ruin with shaky support networks in place or a reasonably performing chip at a low price from a company that still dominates the market, owns its own fabs, and has a huge, sprawling network of support for all its customers…

      Which would you choose? The one with the marginal performance gain undermined either by the company who designed the chip AND/OR the company who actually fabs the chip? Or if the performance gain was not THAT huge for the given audience, wouldn’t you choose the safer, more reliable choice that is far more certain to be up to spec without any “surprises” for you when you go to build your devices around it?

      Because Intel takes full responsibility when its chips fail to fab out right. When AMD’s chips fail, you’ll often have them blaming GloFo (or TSMC) and have GloFo blaming AMD (or throwing their hands up in the air and crying, “WE’RE TRYING OKAY!?”).

      If it was me, I don’t think there’s enough performance difference between Bay Trail and AMD’s latest and greatest in this space to really make me ignore the huge support network Intel provides by the very fact it’s Intel. And sadly AMD doesn’t have the money to invest to really beef themselves back up to pre-Layoffs levels…

        • NeelyCam
        • 6 years ago

        Are you talking about a customer or an OEM/ODM? Because if I’m a customer and the chip is available, I don’t really care what fab the chip was fabbed in as long as it’s fast and efficient and priced well

          • NeelyCam
          • 6 years ago

          Hmm… maybe I should take some of that back, now that I remembered the Bumpgate… I’ve had various graphics cards and AMD CPUs burn out over time, but not once have I had an Intel CPU burn out…

          Then again, burned-out chip = reason to upgrayedd!!

      • sschaem
      • 6 years ago

      One reason… AMD cant give OEM ‘free cash’ like Intel does ?

        • NeelyCam
        • 6 years ago

        Yep, because AMD gave it all to the departing ex-CEOs

          • sschaem
          • 6 years ago

          “Only” 12 million or so. Intel I believe had 3.4 billions in lose for its mobile ‘marketing’
          No way AMD could ever pay OEM that much money , directly or indirectly.

            • NeelyCam
            • 6 years ago

            Wow – sucks to be an AMD CEO. What did Tim Cook make last year? $150mil?

            • sschaem
            • 6 years ago

            1.4 base + 2.8 bonus for 2013, out of 171 billion in revenue. And billions upon billions in profits.

            12 million for bringing a company to the brink of bankruptcy and almost killing its entire future does seem .. Humm. Criminal in comparison..

    • odizzido
    • 6 years ago

    This looks really nice. I have a C50 still right now and that looks like a nice upgrade……assuming that someone releases a 10 or under inch netbook that is decent.

Pin It on Pinterest

Share This