ARM Cortex A9 cores are relatively tiny and it's not too difficult to put a handful of them down on an SoC. NVIDIA did just that with its Tegra 3 featuring four ARM Cortex A9s. NVIDIA's design approach is unique in the industry in that it uses TSMC's 40nm LPG process rather than a strict LP process like many of its competitors. The G transistors are reserved for the CPU cores, while LP transistors are used everywhere else. Typically you'd build the entire SoC out of LP transistors as it's a more cost effective way to build a low power chip. NVIDIA believes that it can more efficiently hit its higher CPU frequency targets using LPG than if it used a standard LP process. G transistors are typically leakier but can scale to higher frequencies at lower voltages than LP transistors on the same process node. The tradeoff is typically that low frequency operation takes place at higher power than in an LP design.

To have its cake and eat it too, NVIDIA added a fifth Cortex A9 to Tegra 3 built using LP transistors (hence the LPG process, G transistors in a sea of LP). This fifth core would only be active during the lightest of workloads and be limited in frequency (500MHz). If your phone/tablet is idle, the four higher performance A9s power down and only the fifth companion core remains active. As you unlock your phone/tablet and get to work, the fifth core eventually hands off the work to one or more of the higher clocked A9s. This dynamic architecture is what NVIDIA called vSMP. 

Today  NVIDIA's vSMP is getting a rebrand. As of now it's called NVIDIA's 4-PLUS-1. Expect to see a few different Tegra 3 phones announced next week at MWC brandishing NVIDIA's new 4-PLUS-1 trademark.

Intel 313 Series SSDs (also known as "Hawley Creek") were originally scheduled for Q4'11 release but obviously they were delayed. VR-Zone is now claiming that the release will take place alongside the Ivy Bridge processors and the 7-series chipsets. That could mean an April release, though there are also rumors of a delayed IVB launch. Either way, here's a quick rundown of the upcoming 313 SSDs.

The 313 Series is the successor of the 311 Series (or Larson Creek if you prefer codenames), which is a 20GB SLC NAND SSD meant for caching with Intel's Z68 chipset with Smart Response Technology (SRT). SRT will be even more useful with the Ivy Bridge platform because there will be widespread support among the 7-series chipsets: two of the three consumer desktop chipsets (Z77 and H77) will feature SRT, along with one of the business chipsets (Q77). We will also finally see mobile chipsets with support for SRT.

The major change with the 313 Series is that it will switch to 25nm SLC NAND and offer a 24GB model, but otherwise we are looking at a product very similar to 311 Series. The controller is Intel's own, but that's all we know for certain. Most likely the controller is the same as in 310, 311, and 320 Series, i.e. Intel PC29AS21BA0, because Intel's roadmap shows no plans for any other SATA 3Gb/s SSDs. It wouldn't make much sense to make a new SATA 3Gb/s controller just for one product, or to create a new SATA 3Gb/s controller in general at this point. Unfortunately we don't have any performance figures but given that the controller should be the same, the performance should be on par with 311 Series—the 24GB model should have slightly higher write speeds as it uses six NAND channels while the 20GB model uses only five.

As for pricing, VR-Zone is claiming a suggested retail price of $99 (20GB) and $119 (24GB). For comparison 20GB 311 Series had an MRSP of $110 and retails for ~$120 now, so a ~$10 price drop sounds plausible. A few German retailers have already listed the 24GB model and it's selling for around €104 without tax, which translates to ~$138, but Euro prices tend to run higher than USD. Moreover, one of the sellers is listing availability as 2-3 weeks so 313 Series may hit the retail channel sooner than April. However, some of the retailers are listing the SSDs as OEM models, which explains the early availability and possibly high pricing as well.

In August of last year, Samsung laid out it's naming strategy for their Galaxy line of phones, with a grand total 5 device categories. Well, turns out they left two off. The just announced Galaxy Ace 2 and mini 2 are sequels to their respective cost conscious predecessors, and while the specifications aren't mind boggling, there may be more to them than meets the eye. 

The Ace and mini were popular siblings to the Galaxy S line, in Europe; and squarely targeted those not willing to pay the high premiums of, say, the Galaxy S II. The Ace, introduced last year, looked strikingly similar to the S II, though its internals were more 2009 than 2011. With a single 800 MHz ARM11 core, and Adreno 200 graphics, this phone wasn't going to be posting high Linpack or glBenchmark scores. Its successor, the Ace 2, seems to be packing a bit more oomph, keeping the clock speed the same while doubling the core count. But what is it?

It's possible, though somewhat unlikely, that the processor is some variant of the Exynos line, say a lower binned sample selected for lower speed operation. Take a look at that chart above, though. The Classic ARM line ends with the Ace's ARM11 processor; and sliding over to the Cortex line, you find the Cortex-A5. First announced in 2009, we haven't actually seen a Cortex-A5 processor in production. The design is very power efficient and aimed at low cost devices. It has about half the performance of the Cortex-A15 line at similar clock speeds, and supports core counts of 1, 2 and 4. It would make a great fit for a dual-core smartphone targeted at the sub-$300 price point, as the Ace 2 probably will be. 

The Ace 2 improves on the Ace, further, by moving from 3.5" HVGA (320x480) to 3.8" WVGA (480x800), though the panel remains a vanilla LCD. Storage climbs to 4GB, with up to 32GB additional through a microSD slot. The SoC is paired up with 768MB of RAM, and powered by a 5.5Whr battery. The mini halves the core count, and apes the original Ace in most other specs, though the screen is a touch smaller at 3.27" and retains the Ace 2's 4GB of NAND. The phones will launch with Android 2.3 (Gingerbread) running Samsung's TouchWiz skin and myriad and assorted Hubs. UK buyers will get their hands on the Ace 2 in April, while French readers will be able to try out the mini 2 this March; global roll-outs will be sometime after. Pricing was not announced, though don't expect them to land far off their predecessors (Ace 2 ~$300, mini 2 ~$200). We're digging to find out more about the SoC's inside and will update as we learn more, in the interim, follow the source link for the (announced) specs.

NVIDIA has just posted GeForce graphics driver 295.73, its latest WHQL-certified driver package for desktops and laptops running 32-bit and 64-bit flavors of Windows Vista and Windows 7. The drivers, NVIDIA's first non-beta driver package since last October, improve performance in a number of high-profile games, add a few new features, and fix a number of bugs.

NVIDIA's benchmarks for the new drivers showcase measureable improvements in performance for Skyrim for users of GTX 500-series cards (though some of these improvements may well trickle down to owners of older GPUs), as well as the addition of Ambient Occlusion support for that game, the Diablo III beta, and Call of Duty: Modern Warfare 3. NVIDIA also highlighted performance improvements for SLI users running the Mass Effect 3 demo.

The new driver also adds SLI and 3D Vision profiles for a number of games, updates the PhysX driver to 9.12.0209, fixes some graphical bugs in Battlefield 3, and enables WHQL support for NVIDIA Surround on Intel X79 motherboards certified for SLI. The drivers support all GeForce 6000-series and newer cards on desktops, and most GeForce 8000-series and newer chips and DirectX 10 and 11-capable Quadro chips on laptops. Links to the driver downloads and to the NVIDIA release notes have been posted below.

32-bit desktop

32-bit laptop

64-bit desktop

64-bit laptop

Source: NVIDIA

BlackBerry PlayBook OS version 2.0 has been released to the public, according to a Research In Motion press release that was posted today.  The new OS version introduces features meant to close the feature gap between it and competing tablets, and to address some of the criticisms voiced in reviews of the device (including ours).

First and foremost, PlayBook OS 2.0 introduces an email client to the platform, addressing a key issue with the platform as it was at release - the client supports a unified inbox, as well as integration with Facebook, Twitter, and LinkedIn messages. The calendar and contacts apps can also make use of information pulled from these social networks. The BlackBerry Bridge app has also been updated, and alongside generic promises that it makes things “quicker and easier,” it adds a feature that allows the keyboard on a BlackBerry phone to be used to control the PlayBook.

Other, smaller features include a Print To Go app, an updated touch keyboard with auto correct and predictive typing, an improved browser, and a new BlackBerry Video Store for US users. RIM also announced the availability of a program called BlackBerry Mobile Fusion, which will allow for management of BlackBerry phones and tablets in an enterprise environment. The version released today is a preview, with availability of the final version scheduled for late March.

The BlackBerry PlayBook had a starting price of $499.99 when it launched, but that has fallen by about half at many retailers. If you’re interested in trying some of these new features, getting into a PlayBook is considerably less expensive than it was even a few months ago.

Source: RIM

Microsoft has talked a bit about SkyDrive before in its plans for Windows 8, but Microsoft's Mike Torres and Omar Shahine have authored a Building Windows 8 blog post that have made the company's plans for it a bit clearer, and outline the degree to which SkyDrive support will be built in to the new operating system.

The Metro app, pictured above, has been built from scratch for Windows 8 and Windows on ARM and will eventually become the preferred method of working with SkyDrive data (though the service's web interface will still be available). Any Metro application in Windows 8 that can save or open files will be able to do so to and from SkyDrive manually and automatically without any extra coding on the part of third-party developers.

The SkyDrive desktop app is less tightly integrated with the OS - when downloaded and installed, it works more like Dropbox, keeping your SkyDrive files in one folder (which is stored locally for offline access) and synchronizing changes to the server as files are created, deleted, and updated - the interface (as pictured above) is very Dropbox-like, right down to the green checkmarks next to synced items. The SkyDrive client will be able to upload files as large as 2GB to your storage space, and the client will also be available to Windows Vista and Windows 7 clients when it' s launched - sorry, Windows XP users, but the writing has been on the wall for awhile, and Microsoft would very much like it if you would upgrade your operating system.

Microsoft's description of its cloud services - "a single drive that’s available across all of their devices, tailored to the experiences they’re using, providing instant, secure, and private access to their files" - mirrors to a large degree Apple's efforts with iCloud (though iCloud's storage is all used directly by applications via APIs, rather than actually allowing users to access a filesystem. iDisk offers more traditional online storage functionality for Macs, but its discontinuation is imminent) - think one part iCloud, one part Dropbox. Expect Microsoft's strategy for SkyDrive to become clearer as the Windows 8 launch date gets closer.

Early versions of the SkyDrive Metro and desktop apps will be made available with the Windows 8 Consumer Preview later this month. See the link below for additional details.

Source: Building Windows 8 Blog

We got off pretty easily with our 7750 launch recap earlier today - with only three cards, it was pretty simple to compare and contrast them and be done with it. The 7770 represents the higher end of the mainstream, though, and as such AMD's partners appear to be pushing it into the channel a bit more aggressively. We've got ten cards to look at this evening, details about which you can find in the tables below.

Our 7750 and 7770 series review noted that these cards provided excellent performance per watt, each of which competes with (7750) and outperforms (7770) the old 5700 series while using significantly less power (remember, the 6700 series was just a rebrand of the 5700 cards). The downside comes in gaming performance for the price - the 7770 is generally outdone in performance by the 6850, which can currently either match it or beat it in price after rebates. That said, if you want decent performance and GCN features without shelling out for a pricier 7900-series card, the 7770 may be the card for you.

As I noted in the 7750 roundup, manufacturers haven't standardized on a unit of measurement for their cards - some measure in inches and some in metric. I've done the necessary conversions and presented all measurements in both inches and millimeters, but manufacturers play a bit loose with these measurements and the actual physical dimensions may not exactly match the dimensions given on the spec sheet.

Common to all of these cards is 1GB of GDDR5 on a 128-bit bus, Eyefinity support, and all of the features inherent to GCN (including DirectX 11.1 and OpenGL 4.2 support and a host of others). All cards also require a single 6-pin power plug, making their power supply requirements just a bit more demanding than the 7750.

ASUS (Product page)

This ASUS card is the first of several on our list to come in at $159.99, the cheapest price point in our recap and the one to beat if you want a 7770 for the least amount possible. This card has a very slight 20 MHz core overclock and 25 MHz memory overclock - barely worthy of the name - but acquits itself better with a solid 3-year warranty and a modest bundle of accessories (it is one of only two cards in this recap to include a CrossFire bridge).

Gigabyte (Product page)

This card's oddly-shaped cooler is the most distinctive thing about it, but it's certainly not a bad deal despite its stock clocks. It costs $159.99 and comes with a 3-year warranty, and according to the manufacturer-supplied measurements it's one of the shorter cards on the list as well.

*: Gigabyte also lists an OC edition of this card (GV-R777OC-1GD) on its web site, but it isn't avalable through Newegg as of this writing. The only difference appears to be its core clock, which is 1100 MHz instead of the 1000 MHz default.

**: This card may come with additional accessories, but none are specified on the card's product page or shown in any available product pictures.

MSI (Product page)

The MSI card, another $159.99 offering, runs stock memory clocks and offers just a 20 MHz core overclock, but its cooler is more impressive than the single-fan offerings - it uses a slightly larger version of the dual-fan cooler included on its 7750 offering - the fans on the 7770 are larger and the cooler is extended to cover the longer PCB, but the styling is otherwise identical. The large fan may make for a better overclock, if that sort of thing appeals to you. The 3-year parts/2-year labor warranty isn't as good as some on the list, but is a bit better than the two-year warranty on many of the other cards.

*: MSI's product pages for all 7000-series cards use the same accessories list, and note that "accessories will probably be different according to the different selling areas of product models." The accessories list includes a DVI to VGA adapter, a CrossFire bridge, a Mini DisplayPort to DisplayPort adapter, and a 4-pin to 6-pin power adapter. The card may actually include fewer accessories - no available product images show the card's bundled accessories.

PowerColor (Product page)

PowerColor's contribution to the list is also unremarkable - $159.99 gets you a card with stock clocks, a 2-year warranty (tied for the shortest on the list), and a rather basic-looking single-fan cooler. It does include a Mini DisplayPort to DisplayPort adapter if you need it, but otherwise this one doesn't stand out from the crowd.

Sapphire GHz Edition (Product page)

Sapphire's got three cards on our recap list, and they all use the same single-fan cooler and offer the same 2-year warranty. This one is the cheapest of their offerings at $159.99 and just runs at stock clocks, but it offers a slightly more robust set of accessories. If you're an overclocker, the fact that Sapphire uses the same fan in both this card and its substantially overclocked OC Edition means that you stand a pretty decent chance of getting a good overclock without running into cooler-imposed limitations, but (as with all overclocking) your mileage may vary.

Sapphire OC Edition (Product page)

First, some strangeness - Sapphire's web page is only listing two SKUs for the 7770, one the stock-clocked GHz Edition card we just looked at, and one the overclocked OC Edition that we see here. However, Newegg is listing two separate part numbers for this card, one of which sells for $169.99 and $179.99. The difference appears to be in accessories - the more expensive SKU is packaged with an HDMI cable and a CrossFire bridge along with the other accessories. Because the cards are otherwise identical, I'll lump them together here, but if you need either of those two accessories you should consider the more expensive card.

Strange SKU issues aside, both of these cards offer the best overclock in our list, with a 15% (150 MHz) core overclock and 10% (125MHz) memory overclock - compared to most of the other stock overclocks in this list, this one should actually give you a measureable increase in framerates. 

XFX Core Edition (Product page)

As was also the case with the 7950 launch, XFX is actually offering four cards using two different coolers, and each cooler ships on one card with stock clocks and one card with an overclock. Newegg is only selling three of these SKUs at present, though, so those are the ones we'll focus on.

Like so many cards on our list, this guy comes in at $159.99 and features only a 2-year warranty (compared to the lifetime warranty on the other XFX SKUs). However, I have to ding XFX for including only one Mini DisplayPort on all three of its cards - every other card in our lineup includes two, and this limits the XFX cards a bit when it comes to EyeFinity setups - every other card in the lineup can drive six displays, but the XFX cards can only handle four. Beware, if that's why you're looking at these cards.

XFX Double D Edition (Product page)

This Double Dissipation Edition card could be described as a mix of the Core edition card above and the Black Edition card below - it runs at stock clocks, but includes the larger heatsink and fan (meaning that overclockers stand a decent chance of getting Black Edition clocks for $10 less). It's tied for second-most expensive at $169.99, but features a lifetime warranty that the Sapphire card can't match - if you don't need two Mini DisplayPorts or a lot of accessories, this card's warranty makes it appealing.

XFX Double D Black Edition (Product page)

This is the same card as the Double D edition above, right down to the lifetime warranty and included accessories, but it sports roughly 9.5% overclocks on both the core (1095 MHz) and memory (1245 MHz) along with a $10 increase in price - at $179.99, it's tied with the higher-clocked Sapphire OC Edition card as the most expensive in this recap. As with the Double D Edition, warranty good, one Mini DisplayPort bad.

In May of last year NVIDIA was on a roll. Hot off the release of the first Tegra 2 based phones, NVIDIA announced the intent to acquire baseband manufacturer Icera for $367M. Four months later, NVIDIA announced that it would have a new addition to the Tegra family, codenamed Grey, in 2013 with an integrated Icera modem. The timing makes sense (albeit a bit aggressive) as it does take time to integrate new technology into an SoC, followed by all of the testing and validation that goes along with the integration. Between now and 2013 however there was always the chance that NVIDIA would deliver a discrete Icera based solution to a customer. Today we have the first customer that has done just that.

The ZTE Mimosa X is a Tegra 2 based Android phone running Ice Cream Sandwich (4.0.x). It features a 4.3-inch qHD (960 x 540) display, 5MP rear camera and 0.3MP front facing camera and 4GB of NAND (expandable up to 32GB). The big news with the Mimosa X is that it integrates an NVIDIA Icera 450 modem (ICE8065 baseband + ICE9225 transceiver). The 450 gives ZTE's Mimosa X support for up to 21Mbps HSPA+ (category 14). 

This is a huge milestone for NVIDIA as it marks the first apps processor + baseband design win for the company. A single datapoint isn't enough to declare NVIDIA's Icera acquisition a success but it's a start. Icera's software baseband is supposed to offer greater flexibility to device manufacturers, especially in bringing products to market with varying network requirements. Icera also promises smaller die sizes as a result of its software based architecture, which should translate into lower cost offerings (or more profit for NVIDIA). The Tegra 2 SoC guarantees that the Mimosa X will be focused squarely on the value/mainstream smartphone market. The Mimosa X will be shipping in Q2 of this year. 

When Samsung introduced the Galaxy Note at IFA, we were excited, but a bit perplexed. The Galaxy Note's 1280x800 HD Super AMOLED display could be a big winner, but the phone's size seemed too ungainly to make an effective phone. Later PR seemed to confound expectations further by referring to it as a 'phablet' rather than a phone or tablet. After strong European and Asian sales, though, the phone has been released on AT&T in the States, and interest is quite high. So it should be no surprise to see other entrants into this oversized phone space.

LG's Optimus Vu has been teased and leaked numerous times over the last few weeks, and so its announcement this weekend didn't come as a huge surprise. Unique to the Vu, versus the Note, is a 4:3 aspect ratio, which gives its 5" display a more squat appearance than the elongated 16:10 slates we're used to. The IPS display touts a 1024x768 resolution, so it's pixel density isn't quite as high as the Note's, though if it is an RGB screen its subpixel density should be impressively high and its color reproduction should be a bit more accurate than Samsung's SAMOLED display. 

Most recent Optimus devices have put Qualcomm's S3 processors to use, with the MSM8660 finding its place in the Optimus LTE recently, alongside the MDM9600 for LTE connectivity. Though Krait SoC's are just around the corner, we expect based on development times that the Vu will feature a similar MSM8660/MDM9600 combination. We have AT&T's Optimus LTE variant in house and so we know what kind of performance we can expect from the device. From a graphics standpoint, the Mali-400 GPU in the Note's Exynos SoC clearly outpaces the Adreno 220 in Qualcomm's S3. Compute performance should be comparable, but with the GPU being leveraged to a greater degree in Ice Cream Sandwich, the user experience delta might grow when both are updated later this year. 

When the 4:3 iPad was chased by 16:10 Honeycomb tablets, part of the discussion focused on the ergonomic differences of a slate that's so wide while in landscape mode. So while viewing widescreen content on a widescreen device might be a bit more satisfying, actually using the wider device can be a bit more of a hassle. LG is obviously hewing to the user data that says that larger devices benefit from the 4:3 form factor, and it'll be interesting to see whether the Note or the Vu feel better in the hand while being used as a tablet. 

But these are phones, and as such, we're concerned about the feasability of putting such a large device up to our heads. Anand spent some time with the Dell Streak as his only phone, and found the experiencing satisfying and the size a non-issue. What's notable is that the Streak, though longer than the Note and Vu, was also somewhat narrower. The Note and Vu are within a centimeter of each other in both height and width, but the Vu comes in nearly a half inch wider than the Streak. Portability could be a concern, too. With each device pushing six inches of length and over three inches in width, the ability to slip these devices into a pants pocket could be challenged. 

So, this begs the question? Are you interested in a phone this big? Is this meant to somewhat canibalize the small tablet market? Could this be your next phone? Let us know in the comments, we'll have more on the Vu next week, at MWC.

Unlike some of our launch recaps (including the 7770 recap article, which covers a whopping ten cards and should come later today), we have only a handful of cards to cover in the Radeon HD 7750 lineup—while our review noted that neither it nor the 7770 are a slam dunk from a price/performance perspective, at present the 7750 has a few things that make it interesting: (1) it's the cheapest way to get in on AMD's GCN architecture, coming in right above $100, and (2) it's one of the fastest cards you can get that doesn't require external power, making it a slightly better candidate as a drop-in graphics card for an OEM PC with a lackluster power supply.

We've only got three cards, and they'll go by quickly. Let's get started.

First, indulge me in a rant about card dimensions—the manufacturers for these three cards provided their dimensions in three different units (inches, centimeters, and millimeters)—for consistency's sake I've converted all units as needed and listed both inches and millimeters above. However, it'd be nice if everyone could unify around one unit of measurement to make these apples-to-apples comparisons easier and to cut down on inconsistencies between manufacturers—the card dimensions above all vary quite a bit for cards that, at least judging from pictures, appear to be pretty close to the same size (note the height and the length of each card, in particular)—this was especially true in our 7970 launch recap, where sizes varied wildly among cards that all used pretty much identical PCBs and coolers. My point being, these figures are provided above, but without physical access to each card and a ruler, it's difficult to ensure accuracy. There, I've said my piece.

Carrying on: Common to all of these cards are the basic characteristics of the 7750 series—1GB of GDDR5 memory on a 128-bit bus, Eyefinity support, DVI-I, HDMI, and DisplayPort outputs, and all of the features inherent to GCN (including DirectX 11.1 and OpenGL 4.2 support and a host of others). All of the cards are pretty light on accessories, which is par for the course once you get down into this price range—multiple adapters and bundled HDMI cables are for more expensive cards.

ASUS (Product page)

ASUS' card sports a very mild overclock on both the core and the memory, and when I say "very mild" I mean about 2.5% and 2.2%, respectively. Still, if you prefer not to overclock, this card is the only one in the lineup with any sort of memory overclock, and it includes the best warranty in the lineup (three years, which covers both parts and labor). This card also uses a single-fan cooler that, in pictures, appears to take up about a slot and a half of space in a case—this card, along with the HIS card, should be a good option for those with cramped cases.

Overall, given this product's price and clocks, I'd say it's the best value here by a narrow margin from a performance perspective, though the MSI card's larger fan may be better for those hoping for slightly better overclocks.

HIS (Product page)

This card is the only one in our lineup running at stock clocks out of the box, though intrepid overclockers should be able to match (or best) the other cards' wimpy stock overclocks without much trouble. The card's two year warranty is also the shortest of the three, though it costs the same as the ASUS card.

HIS calls the single-fan cooler on this card an "iCooler," and its claim to fame is silence—HIS claims it generates less than 28 dB of noise at idle, and a nice analysis over at TechPowerUp seems to confirm this (at both idle and load, the fan is still within a few dB of quieter cards like the 5670 and louder ones like the 7770). Quiet computing enthusiasts may find something to like here.

*: Note that as of this writing, this card's Newegg product page lists a memory clock of 1375 MHz (5500 MHz effective), while the official HIS product page cites a speed of 1125 MHz (4500 MHz effective). Since 1125 MHz is the correct stock speed for this card, I believe that the values from HIS are correct—don't order this card thinking you're getting a massive memory overclock.

MSI (Product page)

The MSI card's main differentiating factor is its large dual-fan, dual-slot cooler, which also includes a large heatsink—overclockers may want to look at this one over the other two, though the fancier cooler drives this card's price up $10 compared to the competition. From the factory, the MSI card's 830 MHz core clock edges out the ASUS card's 820 MHz clock, but it sticks with the stock memory clocks (and that extra 10 MHz overclock is a rounding error, in any case). The warranty here falls right in the middle—it has a three year warranty on parts, but only a two year warranty on labor.

*: MSI's product pages for all 7000-series cards use the same accessories list, and notes that "accessories will probably be different according to the different selling areas of product models." The accessories list includes a DVI to VGA adapter, a CrossFire bridge, a Mini DisplayPort to DisplayPort adapter, and a 4-pin to 6-pin power adapter—since the 7750 can't use the last three accessories in that list, it's safe to assume that a DVI to VGA adapter is all you'll get with this card.

At IDF last year Intel's Justin Rattner demonstrated a 32nm test chip based on Intel's original Pentium architecture that could operate near its threshold voltage. The power consumption of the test chip was so low that the demo was powered by a small solar panel. A transistor's threshold voltage is the minimum voltage applied to the gate for current to flow. The logical on state is typically mapped to a voltage much higher than the threshold voltage to ensure reliable and predictable operation. The non-linear relationship between power and voltage makes operating at lower voltages, especially those near the threshold very interesting.

At this year's ISSCC Intel is presenting details of a number of NTV (near threshold voltage) research projects. For starters, Intel is sharing more details on Claremont - the 32nm NTV Pentium processor demonstrated at IDF. At 3MHz Claremont can operate at 280mV and scale up to 1.2V at 915MHz. Minimum power for Claremont is a meager 2mW.

Intel is also sharing details of a 22nm NTV SIMD engine for use in processor graphics. Given Intel's new focus on improving processor graphics performance, the fact that we're seeing more Intel driven research around GPU technologies isn't surprising. It's also important to point out that Intel needs the experience in building NTV circuits for both CPUs and GPUs if this technology is ever to make it into an actual product. NTV operation grants much better power efficiency where possible, making eventual productization very desirable.

This week is the annual International Solid-State Circuits Conference (ISSCC) where chip companies from all walks of life present papers documenting everything from shipping architectures to future research projects. Intel has always had a large presence at the conference and this year is no different. I'm still trying to get my hands on some of the actual papers being presented but Intel invited some press to a pre-brief on the high level announcements from the conference. 

One such announcement is a test SoC called Rosepoint. It's a 32nm dual-core Atom SoC with an integrated WiFi transceiver. Despite the high levels of integration we see in smartphone SoCs, WiFi is typically serviced by an external combo chip that integrates WiFi and Bluetooth among other radio technologies. Rosepoint brings the WiFi functionality on-die. The name of the game in the mobile SoC space is integration, making Rosepoint a research project with significant real world implications.

Integration is nothing new of course. AMD, Intel, NVIDIA, Qualcomm, TI and all of the others playing in the SoC space have been slowly integrating more functionality on-die over the past decade. Intel claims the difficulty in bringing WiFi on-die is mitigating interference between the RF transceiver and the rest of the SoC. The details of Rosepoint's architecture and how Intel was able to reliably integrate the two are likely described in the ISSCC paper. If I can get my hands on it I'll see about updating this post.

Rosepoint is important because of Intel's dillema as it enters the smartphone SoC space. Most high-end smartphone SoCs sell in the $14 - $25 range, a significant reduction compared to the $50 - $1000 Intel is used to getting for CPUs. Even if you look at smartphone-sized x86 CPUs, Intel can typically get somewhere between $50 - $100 for the CPU. Then add another $20 - $30 for the chipset and margins start looking very nice. Intel can't guarantee > 60% margins selling ~$20 smartphone SoCs. At the same time, Intel based smartphones wouldn't sell very well if they were significantly more expensive than the competition. This puts Intel in a difficult position: settle for lower margins (and upset wallstreet) or figure out a way to offer more value by integrating other parts of the bill of materials. 

Offering and integrating radios where possible is clearly one step, although we'll likely see integrated cellular baseband before we see on-die WiFi. Intel's recent restructuring left the new mobile & comms group with a mandate to deliver an ultra low-power WiFi solution that could work in a smartphone. The first Intel based WiFi in smartphones will begin as a discrete chip but it's clear that integration is on Intel's mind.  The other options for Intel to bring some of that precious BOM in house is to offer reference platforms and/or use software as a differentiator.

In the early days Intel would just sell a CPU and rely on third parties for the rest of the chips on the motherboard. Then came Centrino and the new platform-centric Intel. Expect to see a similar effort in smartphones.

True to their word, Acer is announcing the availability of the Android 4.0 Ice Cream Sandwich update for their recently launched A200 tablet. We posted the PR blurb about the A200 in early January, and we were able to get some hands on time with the A200 at CES 2012. The quick summary is that Acer has trimmed a few features from the A500, like the rear-facing camera and HDMI port, but they also changed up the industrial design. While the plastic exterior may not sound as impressive as the aluminum shell of the A500, the reality is that in hand the textured plastic actually feels better (at least in our opinion).

As far as the Ice Cream Sandwich update, while Acer’s press release states that they’re updating all of the Iconia line to Android 4.0 (including the A100, A200, and A500), the reality of the updates isn’t quite so impressive. The update is currently being pushed to Iconia A200 customers (it’s not clear if you can get it faster by checking for a system update or not) and should be completed over the coming week. A100 and A500 users eager to get the ICS update on the other hand will have to wait a while longer—April to be precise.

Besides all the Android 4.0 goodness, there are a few features that are unique to Acer. Chief of these is their Acer Ring, allowing you to navigate directly to apps from the lock screen. While the idea is nice, in practice we haven’t found the Acer Ring to be necessary and at least the one time I looked at it, it felt a bit chunky and sluggish to launch and navigate (though I could say the same for Android 3.x in general on Tegra 2 hardware). The only other extra feature Acer mentions is an improved weather widget.

You can read more about ICS in our Galaxy Nexus review, though obviously there will be some differences when running on a Tegra 2 tablet. It should provide a better UI experience on Tegra 2 tablets, but unfortunately unless you’re one of the users that purchased an A200, you’ll have to wait a bit longer. The A200 8GB is currently selling for $330 while the 16GB model goes for $350. However, at this late stage in the life of Tegra 2, we wouldn’t recommend buying a new tablet running Tegra 2 hardware unless the price is under $200, especially considering Tegra 3 tablets are expected to hit sub-$300 prices in the very near future.

Using several media outlets, Apple has just announced major details about Mac OS X 10.8, the next version of the company's desktop operatng system. The new release, codenamed "Mountain Lion," will be available to people with Mac developer accounts soon in the form of a preview, and a release to the public is expected late this summer. This short development cycle, unheard of since the early days of Mac OS X over a decade ago, reflects a desire at Apple to mirror the roughly yearly release cycle of iOS.

Despite the name, which suggests a version relatively light in feature changes over the previous version (like the transition from Leopard to Snow Leopard), Mountain Lion is intended to be a major new feature release that continues the work of bringing iOS features to the Mac: many of its major features are iOS transplants, including the Notification Center (which will bring unified notifications to OS X, replacing third-party apps like Growl), Game Center, iMessage support (in the form of an app called Messages, which replaces iChat - there's a free beta available for Lion users now), AirPlay Mirroring, a Notes app, Reminders, Twitter integration, tighter iCloud integration, and others. Frankly, this list of iOS imports actually seems to make more sense for the Mac as a platform than did some of the features (like Launchpad) that were brought over in Lion.

Mountain Lion will also include some new features all its own: Gatekeeper, which is aimed straight at system administrators, will allow admins to lock down the type of apps allowed to run on Macs. You can choose to allow apps only from the Mac App Store, apps from the Mac App store as well as those from developers you approve, or apps from anywhere (which is the default behavior in OS X currently). This can be seen as another step toward disallowing non-Mac App Store programs from running in OS X, but taken at face value it appears to be a solid compromise between the security of iOS-like behavior and the flexibility to install code from anywhere that users have always been accustomed to in OS X.

We don't have any information about system requirements yet, so we don't know whether Mountain Lion will run on any Lion-compatible Mac (which seems technically possible) or whether it will drop support for some older machines (which has historically happened with new OS X releases - see this page of our Lion review for in-depth information on what got dropped from the support list and why). The Apple developer site is currently down, but as soon as it comes back up those with developer accounts should be able to download and play with the next version of OS X. We'll continue to cover the new OS as details are made public.

Update: As we suggested might happen in our Lion review, Mountain Lion's developer preview appears to do away with support for any Mac that cannot boot into OS X's 64-bit kernel. I'll link you to that page of our Lion review again if you'd like deep technical information about what that means, but the short version is that a wide range of Apple's products from 2007 and 2008 are being dropped regardless of whether they include a Core 2 Duo processor. The list of supported Macs includes:

• iMac (mid 2007 or later)
• MacBook (13-inch Aluminum,  2008), (13-inch, Early 2009 or later)
• MacBook Pro (13-inch, Mid-2009 or later), (15-inch, 2.4/2.2 GHz), (17-inch, Late 2007 or later)
• MacBook Air (Late 2008 or later)
• Mac Mini (Early 2009 or later)
• Mac Pro (Early 2008 or later)
• Xserve (Early 2009)

The cutoff happens in different places for different products, but here are some rules of thumb: if your Mac uses the ATI Radeon X1600 graphics chip or the Intel GMA 950/X3100 integrated graphics chips, you're out of luck. If you've got a white iMac or one of the very first Mac Pros, you're out of luck. There are a few easy ways to check whether your Mac can run the 64-bit kernel, and Apple outlines all of them in this support document.

It should be noted that this information comes from the developer preview's release notes and may not be indicative of the final support list, but Lion's dropping of Core Duo Macs (and Snow Leopard's dropping of PPC Macs) were known quantities pretty early in the development of those operating systems - support for these older Macs may be added before the final release, but history suggests otherwise.

Source: The Verge

Microsoft’s Jonathan Garrigues has detailed some aspects of the submissions process for the upcoming Windows Store on the company’s Windows Store for Developers blog. The post describes in detail the various steps involved in the app submission process, and Microsoft’s efforts to address developer complaints about its competitors’ storefronts.

Most of Microsoft’s efforts focus on transparency: they want app developers to be aware of the approval process before they ever submit any code, and they want developers who have submitted apps to be able to track those apps through every step of the submission process. Each step in the process is described, both in terms of what the steps entail and how much time the step usually takes. The submission procedure is mostly automated and focuses on validating code and making sure it runs, but ultimately depends on a human being to launch your app, verify that it works and that it does not violate any of Microsoft’s policies, and verify that it does not contain any malicious code.

Visual Studio will include some features that work toward these ends, including a feature that will allow developers to reserve a name for their app before the app is actually ready for submission to the store. Microsoft will also supply a Windows App Certification Kit that automatically goes through your code to make sure that your app launches and meets a set of basic requirements – the certification kit can be run locally and corresponds to one of the stages of Microsoft’s certification process, meaning that if your app passes this test it stands a better chance of being allowed on the Windows Store.

As is often the case with Microsoft’s latter-day offerings, the Windows Store is trying to find a place in between Apple and Google’s respective app stores for the iOS and Android platforms: it wants to maintain some aspects of Android’s openness and flexibility with the added benefit of iOS’s safety and freedom from malware while at the same time steering clear of Apple’s obscure and sometimes fickle approval process. While Microsoft’s system does result in a “walled garden,” the company hopes that the Windows Store’s relative transparency about the approval process pleases longtime Win32 developers and newcomers alike.

For more on this process, you can read the full post - it is linked below for your convenience.

Source: Windows Store for Developers Blog

LaCie showcased their 2big Thunderbolt drive at CES 2012 and it has now started shipping. 

The LaCie 2big Thunderbolt Series is essentially the LaCie Little Big Disk in 3.5" form factor. It has two swappable 3.5" drives whereas the Little Big Disk has two 2.5" drives. 3.5" drives offer larger capacities and better performance, particularly with hard drive. The 2big is rated at transfer speeds up to 327MB/s while the Little Big Disk tops out at 190MB/s (both results are for the non-SSD versions). LaCie has an 8TB model on their site as well but there is no price listed, which suggests that it's not shipping yet.

There are also two Thunderbolt ports for daisy-chaining, so you can daisy-chain several 2bigs or other Thunderbolt devices. The 2big most likely uses the same SATA controller as the Little Big Disk, which means you are limited to two SATA 3Gb/s ports. For the standard configuration this isn't an issue, but people who are interested in swapping the hard drives for SSDs should be aware that the SATA controller will be a potential bottleneck with the fastest SATA 6Gb/s SSDs.

We have had epic discussions with quite a few readers about the importance of virtualization in our reviews. In our six-core Opteron review we wrote:

"Let there be no misunderstanding: how well a new Server CPU handles virtualization determines whether it is a wallflower or a blockbuster."

Even back in 2008, IDC expected that 52% of the servers would be used for virtualization, but in other reports the numbers were significantly lower. For example, more recently (April 2011) IDC reported that about 20% of all newly purchased servers are used in virtualized environments. No doubt there is some confusion between buying a server for virtualization and the numbers of workloads that find a home in a VM. IDC reports (Dec 2010) that more than 70% of applications are running inside a VM, but there is more.

The 20% virtualized servers number seems low, but you have to drill down a bit in the data. First of all, when we focus on the "mature" markets (US, Europe, Japan, Parts of Asia) the percentage of virtualized servers rises to 30%. And if you then take into account that a few players, such as Google (installed base of 1 million servers), facebook (100k+ servers) and Intel (100k+ servers) are buying massive amounts of non-virtualized servers, you can understand the percentage of virtualized servers is a lot higher among the rest of the server market. In other words, if you do a survey among the server buyers (instead of looking at the server volumes), the percentage of people buying a server for virtualization is much higher. In fact, when we talked to several analysts they indicated that if you ignore the Googles and Facebooks of the earth, the virtualization rate of servers might be as high as 70%.

Not convinced yet? Well, luckily for us Canonical did a survey among 6000 (!) users of Ubuntu Server. Interestingly, 50% of the respondents stated that they use Ubuntu server as a guest OS inside a VM, in other words it runs virtualized. Although this is only a (small) part of the total server market, it is another datapoint that gives us an idea what these Opteron and Xeon boxes are used for.

Interestingly, VMware and not Xen or KVM are the most used hypervisors. To summarize, the percentage of servers bought for virtualization reported by IDC and others are heavily influenced by Google and other "Cloud" buyers. We suspect that a much higher percentage (than the quoted 30%) of the server buyers among our readers consider the virtualized benchmarks as the most important ones.

After last week's broad, sweeping post about Windows on ARM, the Building Windows 8 team is back to discussing the nuts and bolts of the operating system - today, Microsoft's Jennifer Norberg discusses enhancements made to accessibility tools in Windows 8. Chief among the new announcements is a set of baseline accessibility requirements that Windows Store apps must meet before being declared "accessible."

These requirements, which mainly focus on presenting information to users in a clear and simple fashion, can be implemented mostly through standard WinRT-supported languages like HTML5, JavaScript, XAML, and others. Apps created using project templates in Visual Studio Express use code that is already "accessibility ready," and the program's IntelliSense feature can also be used to type (and declare) accessibility information more quickly as you code. Once a developer's app is deemed "accessible," it can be marked as such in the Windows Store to make it more easily searchable.

Microsoft has also made efforts to give developers consistent APIs to use when programming Metro-style apps - the company says that developers (both of standard apps that need to use accessibility features, and of apps designed themselves to enhance accessibility) should be able to depend on on the "UI Automation" accessibility API to communicate with Wndows' accessibility tools more seamlessly than in the past, when developers would sometimes have to code their own methods and workarounds into their programs.

Other enhancements focus on improving on existing technologies: the Windows Narrator, which reads text on the screen back to the user, now performs better, includes additional voices and languages, can be configured to speak more quickly or slowly, can read more content on pages loaded in Internet Explorer, and can respond to user customizable keyboard commands. Narrator can also be invoked during Windows Setup to help with out-of-box setup or an upgrade install, and vision-impaired users on touch-enabled devices can have Narrator read what your finger is touching before activating it.

The Windows Magnifier, another existing tool, has been adapted to work better on touch-enabled computers - it can now be invoked by pressing the Windows logo key and the volume up button, which zooms in and brings up on-screen borders that can be used to change your focus.

The full post, which is linked below for your convenience, covers these topics in greater depth - read on if you're interested.

Source: Building Windows 8 Blog

Intel has finally filled out the Sandy Bridge E lineup by releasing the Core i7-3820. The initial Sandy Bridge E lineup launched back in November 2011 and it consisted of two SKUs, the i7-3960X and i7-3930K. While the i7-3820 wasn't released until this week, we reviewed it over a month ago, so head there for a longer analysis. The table below summarizes the current Sandy Bridge E lineup:

The short summary is that i7-3820 is Sandy Bridge E on a budget. In terms of CPU performance and price, it's equivalent to the i7-2600(K) but provides higher I/O performance due to the quad-channel memory and 40 PCIe 3.0 lanes. Anand summed it up nicely in his review, so we'll just repost here.

There are three reasons why you'd want the Core i7-3820:

1. You need PCIe 3.0 today and/or you need more PCIe lanes than a Core i7-2600K can provide.
2. You need tons of memory bandwidth for a particular application.
3. You want a 2600K but you need a platform that can support more memory (32GB+).

So in general, most users will be better off with a LGA 1155 based platform. While the i7-3820 is actually cheaper than the i7-2600K, the total price of the platform is not. LGA 1155 based motherboards go for as little as ~$50 (e.g. Gigabyte GA-H61M-DS2). If you want more features such as Intel Rapid Storage Technology, you can get a Z68 based motherboard for around $90 (e.g. ASRock Z68M/USB3). In contrast, the cheapest LGA 2011 based motherboard starts at $210. Unless you benefit from the extra features that Sandy Bridge E offers, your money is better spent else (e.g. on an SSD).

I just found out that AMD's Eric Demers (Corporate VP & CTO, Graphics Division) is leaving the company at the end of this week. He's not going to Intel or NVIDIA but I suspect that someone of Eric's talents will remain in the industry. I just had dinner with Eric a couple of weeks ago and he seemed very positive on AMD's roadmap going forward. Given how important the GPU is becoming in this ever expanding industry, someone like Eric is in very high demand. 

We now have an official statement from AMD:

Eric Demers, AMD Corporate Vice President and CTO, Graphics Business Unit, has decided leave AMD to pursue other opportunities. 

AMD Chief Technology Officer Mark Papermaster will assume interim responsibility for the Graphics Business Unit CTO role until a replacement is found. 

AMD remains fully committed to our critical graphics IP development and discrete GPU products.  We have a tremendous depth of talent in our organization, a game plan that is resonating with our customers and our team, and we are continuing to bring graphics-performance-leading products to market.  We will attract the right technology leader for this role.

We thank Eric for his contributions to the business and wish him well in his future endeavors.