Humidity rather than heat is the number one enemy of the hard disk

Humidity rather than heat is the number one enemy of the hard disk

Heat isn’t the biggest enemy for the humble hard disk, rather humidity is what causes the most failures, a new piece of research has observed.

The study, carried out by Rutgers University and entitled ‘Environmental Conditions and Disk Reliability in Free-cool Data Centres’, found that the most negative effects on drive controllers and adapters were felt when humidity levels increased.

As Network World reports, the testing took place in Microsoft data centres and encompassed over a million hard drives over a period of several years, and unsurprisingly found that the vast majority of hardware failures in the data centres – 89% of them – were disk failures.

Clear difference

And as the humidity level rises, hard disk rises, failures increase to such an extent that the study authors noted you could easily tell which data centres had humidity controls, as those which didn’t showed up clearly when they looked at the annualised failure rate of controllers.

Humidity is such a danger that researchers found that positioning drives in the “hot region at the back of the server” actually improved the reliability level of the drives, because the heat kept humidity at bay – and the heat is clearly the lesser of two evils.

Whether the cost of advanced humidity controls for a data centre is worth it compared to what you’d fork out replacing the extra failed disks is another matter – although that also depends on how long-term you’re looking.

Last month, we saw some research from Google on the reliability of the hard disk’s big rival, the SSD. That also turned up an interesting finding, namely that it wasn’t the amount of usage the SSD had seen which correlated with high error rates, but simply the age of the drive.

In other words, heavy usage isn’t the big demon which it used to be, and today’s SSDs cope far better with heftier workloads.

Intel’s blisteringly fast Optane SSD tech will be compatible with MacBooks

3D Xpoint memory chips.

Intel’s been quiet about its super-fast Optane memory and SSD products, but a few emerging details may hint at how they could be used in products like Apple’s MacBooks.

Optane is a brand name for a new type of memory and SSD based on 3D Xpoint, a technology jointly developed by Intel and Micron that can be 10 times denser than DRAM, and 1,000 times faster and more durable than flash storage.

The soon-to-be-released 3D Xpoint technology will be compatible with NVMe, a super-fast storage protocol that could help Optane achieve its blistering speeds. Some MacBooks already have NVMe-based SSDs, and Optane could provide a further speed boost.

Apple is among the first vendors to implement the latest laptop technologies and may jump at the chance at putting Optane in its MacBooks. Apple was the first to implement Thunderbolt and processor technologies from Intel.

The first Optane products will likely be SSDs and reach enthusiasts’ PCs next year, then spread to other desktops and mobile products. Optane memory DIMMs, which can be plugged into existing memory slots, are also coming.

Some Windows laptops also have NVMe storage, but most still rely on the older and slower SATA interface. Enthusiast desktop users like gamers are early adopters of new technology, and many will likely move over to Optane.

Optane products will be initially based on Intel’s Skylake architecture. If Intel ships memory DIMMs, they will need to be compatible with the DDR3/4 DRAM bus that is in most PCs today.

Optane’s compatibility with NVMe was revealed during a session focused on Lightning, an open-source storage design, at the OCP U.S. Summit in San Jose, California. Lightning, developed by Intel and Facebook, provides flexible storage in which capacity can be cranked up depending on processing needs.

Optane could drive changes in server designs and in application processing. Servers could have more capacity for in-memory applications like SAP HANA or 3D Xpoint memory, and SSDs could be decoupled into separate boxes.

Separating storage, memory and processing resources could help reduce data center costs. Intel has a data-center architecture called Rack Scale that focuses on decoupling those resources.

With 3D Xpoint technology, memory cells sit in a three-dimensional mesh. The structure allows data to be written in small sizes and at much faster speeds.

Western Digital ships $31, 314GB PiDrive for the Raspberry Pi 3

WD's PiDrive 314GB drive is designed for Raspberry Pi 3.

If you’re looking to turn your Raspberry Pi 3 into a usable PC, you’ll probably want an external hard drive to provide additional storage. Western Digital’s newest PiDrive might be the answer.

The company already offers a PiDrive with a 1TB storage capacity, but at $79.99 it’s more than twice the cost of the Pi 3. But Western Digital is now offering a 314GB versionfor $31.

The Raspberry Pi 3 is the first version of the board that has the processing power to be a full-fledged PC. It has no internal storage, so you need an SD card or external hard drives to store files.

Not all external drives work with Raspberry Pi boards, but the PiDrives are compatible. Western Digital says it tweaked the drive’s magnetic recording and electrical set-points to reduce the power draw compared to a standard hard drive.

The new PiDrive drive has a USB 3.0 interface, which can connect to the Raspberry Pi’s USB 2.0 port.

A BerryBoot installer, available for download from Western Digital’s website, can load multiple programs or operating systems onto the drive at boot.

Razer’s refreshed Blade gaming laptop adds more power for far less cash

Razer Blade (2016)

If you saw our review of the Razer Blade Stealth earlier this week and thought you’d abdicated to an alternate dimension where the black-and-green gaming mavens at Razer no longer made gaming hardware, fear not: The 14-inch Razer Blade (no surname) still exists, and it’s getting an update for 2016.

For the most part it’s the usual incremental hardware advances. Last year’s i7-4720HQ is now an i7-6700HQ. SATA M.2 drives have given way to speedier PCIe drives. The old Intel wireless card has been replaced with a Killer Wireless card. The Nvidia GeForce 970M graphics chip is…still a 970M, though now with 6GB of onboard VRAM. And Razer’s sticking with the same 16GB of RAM configuration introduced last year, but now it’s using DDR4 memory.

Razer Blade (2016)

The other changes are pretty predictable, too. The new and improved Razer Blade sports per-key Chroma (RGB) lighting, and if you didn’t see that one coming then you’re probably too blind to appreciate those 16.8 million colors anyway. As always, the usual disclaimer: Is RGB lighting necessary? No. Do people love it? Some do! Do people buy it? Yes.

The Blade will also charge by way of USB-C, the same as the Blade Stealth. And following logically from that, it will be compatible with Razer’s new Thunderbolt 3 graphics amplifier, the Razer Core, whenever it’s released. Pair the two and you could run apowerful desktop graphics card at home for enhancing gaming performance and still have the decently robust 970M as a fallback on the road.

But this is all secondary to what is—without doubt—the most important change to the Blade in 2016: Price. The new Blade will start at $1,999 with a 256GB SSD, or $2,199 for a model with a 512GB drive. Last year’s models listed for $2,399 and $2,699, respectively. Shaving off $400 doesn’t exactly make the Blade affordable, but it’s definitely more in line with what I’d expect for a 970M/4K screen/aluminum chassis.

In other words, the Blade is still a damn luxurious gaming laptop. Slightly better specs and a cheaper price? Sounds like 2016’s update is an all-around improvement. Orders start today and ship in April. We’ll hopefully have a corresponding review soon. Until then, stay tuned to PCWorld for more news out of this week’s Game Developers Conference.

Intel adds Vulkan 1.0 support to Windows PC chips for gaming

Vulkan 1.0 is coming to Intel's Skylake chips.

Windows games have mostly been defined by DirectX 12 tools, but a competitive API is coming to PCs running on Intel chips.

Intel is releasing graphics drivers that support the Vulkan 1.0 API for chips running Windows 7, 8 and 10 PCs. The drivers, specifically, add “new beta support for the Vulkan 1.0 API for 6th Generation Intel Core and related processors,” the Intel driver page says.

Most games for Windows are written using Microsoft’s DirectX 12 programming tools. The Vulkan 1.0 API, however, provides an alternative set of game development tools.

Vulkan 1.0 was introduced last month by industry consortium Khronos Group and replaces the aging OpenGL, which was first introduced in 1991 by Silicon Graphics. It improves graphics on PCs, mobile devices, VR headsets, robots and other devices.

Vulkan is tuned to exploit the latest features on modern hardware, like powerful GPUs and multicore CPUs, so games have more life-like images and higher frame rates.

With the new drivers, developers will be able to exploit features on Intel GPUs, like the Iris Pro, that are integrated in chips alongside CPUs. Intel’s rival AMD has already released Vulkan drivers for Radeon graphics processors.

Much like DirectX 12, Vulkan 1.0 provides close access to hardware, which reduces the processing and power overhead in drawing up images. Developers can define, with better access to specific hardware features, how they want graphics rendered. That’s an improvement over OpenGL, which had abstraction layers that made hardware virtually invisible.

Graphics quality would also degrade when trying to port games from Windows to the OpenGL standard — unlike DirectX, OpenGL runs on Linux machines. But with Vulkan, quality of games remains mostly intact when porting from DirectX, said Jason Ekstrand, a developer at Intel, during a talk.

Vulkan 1.0 APIs will also work with Linux-based PCs like Steam Machines. Intel has made available open-source Vulkan drivers for Linux PCs running on chips code-named Broadwell and Skylake.

Samsung’s Flow bridges the gap between Android handsets and Windows devices

Galaxy TabPro S starts at $899.

Like Apple, Samsung is trying to link its devices so they work better together. A new effort bridges the gap between the company’s Android handsets and Windows devices.

A new feature called Flow in Samsung’s new Galaxy TabPro S helps the Windows 10 tablet work smoothly with Galaxy smartphones running Android. The super-thin tablet started shipping in the U.S. on Thursday, with a starting price of $899.

Until now, Samsung’s Windows PCs and Android devices didn’t work hand-in-hand. Flow provides some interesting hints on Samsung’s plan to link Windows and Android devices.

Flow will allow Android smartphone users to use their devices to log in to Windows PCs. Users can swipe a finger on a smartphone’s fingerprint reader, logging a user into the TabPro S, which doesn’t have a fingerprint reader. Similarly, users can log in to the TabPro S by pattern authentication — by drawing specific shapes — on smartphones. Those actions require a Flow application on a smartphone, which passes on authentication information to the TabPro S through NFC protocols.

Flow will also allow notifications of phone calls and text messages on Galaxy smartphones to also appear on the TabPro S. This was not possible on earlier Samsung PCs. It’s a feature drawn from Samsung’s Gear smartwatches, which can show smartphone notifications.

Flow also links the TabPro S to a mobile Wi-Fi hotspot on a Galaxy smartphone.

The TabPro S screen can be mirrored on Samsung TVs. Samsung’s smartphones already work well with the company’s TVs.

Samsung is working hard to ensure its devices talk to each other. The company is building smart appliances like refrigerators and washing machines that could be controlled through smartphones.

But trying to make Windows work with Android creates a whole new set of challenges due to differences in codes and protocols.

There are many possibilities with Flow, but Samsung hasn’t determined what additional features will be built into it, a company spokeswoman said at a press event in New York City.

The Galaxy lineup is dominated by Android smartphones and tablets, and the TabPro S is the company’s first Galaxy Windows tablet. The $899 entry-level price is the same as Microsoft’s base Surface Pro 4 model, which has a similar configuration, with an Intel Core M3 chip and 128GB of storage.

The tablet is is 6.35 millimeters thick and weighs 680 grams. A highlight of the tablet is a Super AMOLED display that shows pictures at a resolution of 2160 x 1440 pixels. OLED screens are more vibrant than conventional LCD/LED screens and also more power efficient because they don’t include backlighting.

Asus’s itty-bitty VivoMini PCs embrace Nvidia graphics


Today appears to be mini PC day. We’ve already given you hands-on looks at Intel’s Skull Canyon NUC and Zotac’s VR-ready Magnus EN380, not to mention news about the release of MSI’s cylindrical Vortex powerhouse. That makes today a good day to talk about something that was recently announced: two new additions to Asus’ VivoMini line-up, one of which comes with extra graphical oomph thanks to discrete Nvidia graphics.

The recently announced VivoMini VM65 and VM65N feature your choice of an Intel “Skylake” Core i3 or the 2.3GHz Core i5-6200U, according to Anandtech. The PCs come loaded with two SO-DIMM slots for up to 16GB of RAM, and space for either a 3.5-inch hard drive or an optional Vivo DualBay to fit in two 2.5-inch storage drives.

Ports and other components include 802.11ac Wi-Fi, Bluetooth 4.0, singlular HDMI and dual-mode DisplayPorts, four USB 3.0 ports, and—on the VM65N only—a pair of Type-A USB 3.1 connections.

The VivoMini has a 7.48-inch square footprint with a height of 2.21 inches. That tiny size, which is a little bit bigger than the NUC but smaller than a mini-ITX PC, means you can maximize your desk space by attaching the device to the back of your desktop display via a VESA mount.

Why this matters: We’re well into the era of PCs as appliances and the rise of mini PCs really reinforce that. With little need for disc drives, massive drive bays, or USB thumb sticks, not everyone requires a giant tower under their desks. Tiny PCs like the VivoMini line are well equipped to handle email, productivity apps, and web applications while still giving you that traditional Windows desktop experience without taking up much space.

Discrete (mobile) GPU

The big difference between the two VivoMini models is that the VM65N rocks an Nvidia GeForce GT 930M mobile GPU. That isn’t a super powerful GPU for serious gaming, but Asus told AnandTech that the GPU is really meant to support photo and video editing, and powering up to three displays at once.

Poking around YouTube, however, it looks like this GPU could work for gaming at lower resolutions and graphics settings. I saw examples of laptop gamers getting around 40 frames per second at 1366-by-768 resolution on medium settings, on some relatively modern games. That’s console-level frame rate quality for games like Battlefield 4 andGrand Theft Auto V. It certainly wouldn’t be a premium experience—especially since the VivoMini would use a desktop-sized display—and you can forget about super graphics-intensive games like The Witcher 3, but it’s still a neat capability to have in your back pocket.

Asus hasn’t announced pricing or a release date for the new VivoMini PCs.

MSI’s tiny, cylindrical Vortex gaming PC hits the streets with dual GeForce GTX 980s

msi vortex

Now the Windows world has a trashcan PC to call its own. Except this one looks much, much cooler than Apple’s Mac Pro and it has enough power to satisfy the biggest PC gaming fantasies. The VR-ready Vortex “gaming cylinder” is now available, MSI recently announced.

The Vortex measures 10.5-inches tall, which MSI says is about the size of a sub-woofer. Packed into that tiny package, however, is a whole lot of super-powerful guts. The PC is powered by up to a 4.0GHz Intel Core i7-6700K, 64GB DDR4 of RAM, dual Nvidia GeForce GTX 980 graphics cards, and a Super RAID 4 set-up for onboard storage.

The impact on you at home: The Vortex is available now from various online retailers. You can find a complete list on MSI’s Vortex page. Prices for the stub-sized powerhouse start at $2200. If you’re interested in using it for VR gaming as well as traditional PC gaming, getting it delivered in time for day the Oculus Rift launch later this month depends on when and from where you order it. At this writing, Amazon quoted a ship time within two to five weeks, B&H said the device was coming soon, and Newegg had the PC in stock.


Power! Unlimited power!

All that power translates into some pretty impressive tricks. MSI’s li’l bucket can support up to six simultaneous video outputs. It’s 4K ready of course, and will have no trouble powering that Oculus Rift or HTC Vive. Online gamers will also be pleased with the Vortex’s addition of Killer DoubleShot X3 Pro connectivity. That setup means less lag when you’re shooting baddies on the quarantined streets of New York while simultaneously streaming your glory live on services like Twitch.

There’s also a mobile app for Android and iOS allowing you to adjust your machine’s performance from your phone.


And then there’s why this thing is called the Vortex to begin with—it’s cooling. MSI says it loaded the Vortex with the company’s 360-degree Silent Storm Cooling technology. The system, according to MSI, “creates a vortex-shaped wind path” by taking heat from the bottom of the machine and moving it up and out of the unit.

So yes, the Vortex takes its name from the cooling system. If it makes you feel better, you can just pretend it’s called the Vortex because of the gaming mayhem the system brings into your home.

Razer’s laptop-enhancing Core graphics card dock goes up for preorder at a steep price

razer core

Razer’s adamant that its new Blade Stealth ultrabook is not a gaming machine—and that’s fine! It’s a damned good laptop in its own right, especially for the price. But part of the Stealth’s pitch at CES was that it could be a gaming machine, provided you had a spare graphics card around and were willing to pony up for Razer’s Thunderbolt 3 enabled graphics amplifier, the Razer Core.

…Which made it a bit weird when Razer launched the Blade Stealth all on its lonesome,sans-Core.

But fear not: It’s coming. Razer announced today that the Core is available for preorder, with the first units shipping in April. The price for turning your laptop into a gaming laptop—or your gaming laptop into a better gaming laptop—is steep though: $499, or $399 if you buy (or recently bought) a Razer Blade/Blade Stealth.

Now, about that price: It’s early-adopter tech. The Core is literally the first Thunderbolt 3 graphics amplifier to market, and there’s a good chance that other external GPU docks like Asus’ ROG XG Station 2 will fall in the same price range.

Plus, our own Gordon Ung went hands-on with the Core at CES (watch the video below) and told me it’s “very nicely built” after I regaled him with the cost. The Core is all aluminum, with a sleek slide-locking mechanism and the ability to (at least with supported video cards) seamlessly plug/unplug without necessitating a reboot of the Blade/Blade Stealth. That functionality will be in select AMD cards at launch, thanks to its new Radeon XConnect technology. Razer says Nvidia’s graphics cards, including all GTX 700- and 900-series GPUs, will offer support for the Core at launch as well, though it didn’t say whether GeForce hardware will be hot-unpluggable like Radeon cards. (Which means they probably won’t be.)

Even so, the Core could still be a hard sell in these early stages. That $399/499 price doesn’t include the actual graphics card you’d slot inside, so you’re looking at an additional investment. And you can find Alienware’s proprietary Graphics Amplifier for $200 these days, though the build quality is less impressive and the hardware less future-proof.

I guess the main consideration is whether Thunderbolt 3 graphics amplifiers become standard across the industry. If you’re still using the Core five or ten years down the road, it makes $499 up front seem like less of a big deal—especially if the Core plays nice with your future Alienware laptop or Lenovo laptop or whatever. If we end up with a bunch of semi-proprietary ecosystems though? Or if Thunderbolt 3 is relegated to the sidelines like it has been in the past?

It’s hard to predict. Personally I have my fingers crossed the price comes down. It seems like great technology, but maybe not “Half the cost of the actual Blade Stealth” great.

Intel may turn to AMD for Radeon graphics tech licensing


Intel could dump Nvidia for a licensing deal with AMD as the chip giant tries to prop up its patent portfolio.

Currently, Intel is under a $1.5 billion licensing agreement with Nvidia, which the two companies signed in 2011. At the time, the two companies had spent years fighting each other in courts over patent licensing, and the agreement put all that litigation to rest.

Intel’s Nvidia deal is set to expire on March 17, 2017, and a recent report by Bloomberg claimed that Intel is now looking to cut a deal with AMD instead. (Strangely, the report does not appear on Bloomberg’s website, but was cited by analysts and relayed by Barron’s. It may be worth taking with a grain of salt.)

What does all this mean for consumers? Potentially, not much. Intel’s current licensing deal with Nvidia never translated to any sort of Nvidia-branded technology in Intel chips. At best, the biggest impact for consumers is that the two companies didn’t waste more time and energy trying to sue each other into oblivion.

Having said that, Intel and AMD have been playing nice together lately. Both companies were part of a collaboration with Razer on the Razer Core GPU dock, which provides gaming muscle to laptops or low-powered desktops through a Thunderbolt 3 connection. During a press event, Intel even used an AMD graphics card to demonstrate how the dock works with its upcoming Skull Canyon NUC mini-PC. As ExtremeTech notes, Intel has also indicated that it will support AMD’s FreeSync standard for variable refresh monitors, rather than licensing Nvidia’s rival G-Sync technology.

It’s also worth noting that AMD recently spun off its GPU business into a separate group that covers software, hardware, and intellectual property. Just to speculate, this could pave the way for AMD to more directly integrate its graphics innovations into other companies’ chips—including Intel’s.

Why this matters: Given AMD’s experience with integrating CPUs and GPUs on the same die, it’s intriguing to think of what a licensing deal with Intel might bring, especially with Intel leading the mini-PC invasion. But this is just a rumor for now, and even if it comes to fruition, the result could be nothing other than more favorable licensing terms than what Intel gets through Nvidia now. At most, it’s a development worth watching.