Old school MS-Apple drama: Microsoft bans iPads and Macs

A recent email leak seems to suggest that Microsoft will soon ban employees from expensing Apple products. While they are not prohibited from buying Macs and iPads for their own personal use, the company is banning departments from purchasing Apple products for employees. Furthermore, Microsoft will be phasing out any existing iPads and Macintosh computers.

While Microsoft is refusing to comment on the story or email published by Tom’s Hardware, the message is allegedly from Alain Crozier and is part of Microsoft’s Sales, Marketing, and Services Group. According to his Microsoft profile, there is indeed an Alain Crozier listed as being the Corporate VP and CFO of said group. So, even if this is fake, at least the pranksters did their homework.

Assuming that the internal communication is legitimate, a total ban of any Apple devices seems drastic. I can understand that Microsoft likely wants its employees to use home-grown products (or at least not products from their main competitor). Conversely, I can empathize with the few users that have grown accustomed to performing their job duties using the Apple devices, services, and interfaces. If they are primarily Mac users, for example, the adjustment to working in Windows while growing up with with another OS may be tough.

From the email leak alone, it is not clear exactly who received the communique. After some consideration of the illogical nature of hampering internal developers and Apple-related product teams, it is unlikely that the Apple ban would be enforced throughout the entire company. After all, Microsoft has also started developing iOS applications, including Kinectimals, SkyDrive, and OneNote, there are teams working on Office for Mac, and so on. Rather, the email would make the most sense when applied only to customer- and public-facing employees. Marketing teams, PR personnel, and the sales force are likely candidates to have received the Apple product ban email. By ensuring that the public sees only Microsoft equipment being used by employees, it helps to foster an image of Microsoft standing by its own products. In areas of the company not under public scrutiny, Microsoft could continue to allow developers and engineers to use Apple products. The result of that arrangement would give Microsoft the best of both worlds.

One likely cause for the ban is Microsoft’s increased presence in the mobile market with Windows Phone 7, and the impending launch of the mobile device and traditional computing interface unification that is Windows 8. The company is still the underdog in the mobile space, and it’s likely feeling the pressure from Apple as it tries to grab a larger slice of the market. Essentially preventing employees from using the competitor’s products may be one way to keep them focused on Microsoft’s own ecosystem of products. If employees have to use MS day in and day out, they might be more likely to have that experience trickle into and affect their personal lives, including what products and device are recommended to others.

Preventing employees from using Apple products is not necessarily a bad thing for Microsoft to do, especially since it’s footing the bill for employee equipment. Being able to exert more control and deliver a focused, unified fleet of machines makes the IT department happy and has the potential to create a tighter knit employee environment.

Of course there is something to be said about knowing the competition. Such hands-on and real-world usage information can then provide Microsoft with feedback on what the company should do for its own products. The scenario where employees would be allowed to use Apple products to work for Microsoft sounds like a perfect pool of market and development research data to me.

Read more at Tom’s Hardware

[Image credit]

Source

Nvidia deeply unhappy with TSMC, claims 20nm essentially worthless

One of the unspoken rules of customer-foundry relations is that you virtually never see the former speak poorly of the latter. Only when things have seriously hit the fan do partners like AMD or Nvidia admit to manufacturing problems, and typically only after postponed launches and poor availability have made protestations that everything is fine unsustainable.

That’s why we were surprised — and our source testified to being stunned — that Nvidia gave the following presentation at the International Trade Partner Conference (ITPC) forum last November. Many of the company’s complaints regarding its current partnership with TSMC are exactly what you’d expect given the manufacturing problems the entire industry is facing. What’s surprising are Nvidia’s remarks concerning TSMC’s current cost curves and manufacturing ramps. This is normally the sort of information discussed quietly between a foundry and its customers or by the press with help from various anonymous sources. Discussing the problems publicly is a sign of just how frustrated the company has become.

Watch the underlines for emphasis

TSMC builds hardware for a huge number of companies, but those customers have very different needs and use a wide range of process technologies. Historically, Nvidia (and ATI/AMD) have been regular early adopters. The nature of graphics is that it can easily soak up new processes and the higher transistor counts they enable.

Kepler broke the exponential rise in transistors per GPU

The flip side of that situation is that companies like AMD and Nvidia have also been responsible for assuming the risks associated with “risk production” and footing a hefty bill for the privilege. As those risks mount and costs skyrocket, Nvidia is increasingly unhappy with being asked to shoulder the burden. Nvidia’s slides talk about the need for “real” understanding, compromises on “rough justice,” and a closer relationship that looks more like that of an IDM (Integrated Device Manufacturer). For those of you who don’t know the term, Intel is an IDM — it handles both manufacturing and design. AMD used to be.

When AMD spun GlobalFoundries off, one of the things GF promised to provide that would distinguish it from TSMC was high levels of IDM-style integration. At TSMC, the customization work that is available is highly monetized; specialized work is expensive and time-consuming. In reality, GF’s ability to provide the amount of IDM-like flexibility that it wanted to offer has been sharply constrained by the problems associated with Llano and Bulldozer; our sources tell us that the foundry devoted enormous resources to bringing AMD’s 32nm APU back on track.

According to Nvidia, the current model is unsustainable. Here’s the company’s projected analysis for transistor costs at current and new nodes.

As the process nodes shrink, it takes longer and longer for the cost-per-transistor to fall below the previous generation. At 20nm, the gains all-but vanish. Want to know why Nvidia rearchitected Fermi with a new emphasis on efficiency and performance/watt? You’re looking at the reason. If per-transistor costs remain constant, the only way to improve your cost structure is to make better use of the transistors you’ve got.

As for wafer costs, they’ve become part of the problem.

What this slide states — we can’t even call it a suggestion — is that smaller processes no longer improve yields by leading to a greater number of chips per wafer. Instead, the complexities and difficulties of manufacturing at the new process create a cost structure that provides precious little incentive to manufacture at the new process.

If openly criticizing a foundry partner is unusual, showing data that suggests that your foundry partner can’t provide a cost-effective strategy for building hardware at next-generation process nodes is… a few steps past that point. The recent launch of the GTX 680, and that card’s trifecta of price/performance/power-efficiency actually strengthens the impact of this data. NV would’ve had a good idea how the GK104 was shaping up when it spoke at ITPC in November; this isn’t a case where a company is angry about the performance of a particular part and looking for someone to blame.

Again, follow the underlines.

The GK104 is great, but it doesn’t change the nature or severity of the underlying problems. As for whether Nvidia’s unhappiness with TSMC heralds a potential alliance with GlobalFoundries, we’re dubious. Not only has GF only recently ironed out its own 28nm issues, the nature of the foundry business doesn’t allow for quick shifts. Indeed, part of the reason that manufacturers like TSMC have historically exercised such control over their partners’ PR releases is because once you’ve committed to a foundry, you’re locked in for a substantial period of time. The fact that there’s now two foundries available with cutting-edge technology doesn’t change that, and the Common Platform Alliance favored by IBM, Samsung, and GloFo only mitigates some of the problems with moving a design from foundry to foundry, it doesn’t remove them.

The real question, at least for TSMC’s other customers, is whether the graphs and charts Nvidia has shown are specific to the company’s own products or reflect universal trends. There’s good reason to suspect the latter; Nvidia may have had more trouble than some of TSMC’s other customers, but our analysis of semiconductor industry roadmaps revealed a great deal of uncertainty about the road forward. Nvidia opted to aggressively optimize GK104 precisely because the old strategy of bolting on more cores and ratcheting up transistor counts isn’t sustainable.

Further evidence for the accuracy of NV’s presentation comes, ironically, from the company’s primary GPU competitor. At AMD’s Financial Analyst Day, CEO Rory Read made a point of saying that the company no longer intends to aggressively transition to new process nodes given the diminishing marginal returns from doing so.

Change the color scheme, and Nvidia’s graphs could’ve dropped right into AMD’s presentations in early February.

Nvidia’s willingness to stand up and talk about these problems is an “Emperor’s new clothes” sort of moment. The long-term repercussions, if any, are still unclear.

Source

GlobalFoundries hits 32nm milestone, but AMD’s still refusing to commit to 28nm production

GlobalFoundries announced today that it’s shipped its 250,000th 32nm high-k metal gate (HKMG) wafer, alongside the usual claims that the 32nm ramp (despite the tremendous problems the company ran into with Llano) really is faster/better/more successful than the 45nm ramp before it. We feel obliged to mention that AMD’s 45nm ramp took place in 2008 – 2009, in the teeth of the recent recession. At the time, AMD noted that its ramp had been slower than it originally intended due to weak demand and high inventories of older Phenom parts. Toss in the fact that GlobalFoundries is now building parts for an entire range of customers rather than simply AMD, and one would very much hope that the company was able to push product out the door with more alacrity.

According to GF’s CEO, Ajit Manocha, relatively recent changes to the company’s 32/28nm production have yielded significant improvements. AMD’s CEO, Rory Read, weighed in with the following: “AMD and GLOBALFOUNDRIES worked in close partnership during 2011… In just one quarter, we were able to see more than a doubling of yields on 32nm, allowing us to exit 2011 having exceeded our 32nm product shipment requirements. Based on this successful ramp of 32nm HKMG, we are committed to moving ahead on 28nm with GLOBALFOUNDRIES.” (emphasis added.)

That last sentence is getting a lot of attention. Readers who jump on what they see as misleading headlines would appear to have a bone to pick with me. Rather than play coy, I want to address it directly. Read’s statement about “moving ahead” sounds good, but it doesn’t actually refer to any particular product. That’s not an accident. AMD has been remarkably close-mouthed about where it was building its upcoming 28nm APU and CPUs, but Read’s comment left me wondering if the company had left a trail of breadcrumbs.

Follow the wafers

If you want to know what a company is planning, its SEC filings are a great place to start. The rules and regulations that govern business disclosures often require more specific information than what makes it into press releases, including forward-looking statements that aren’t subject to stealth editing when internal roadmaps aren’t met. In this case, AMD’s various statements reveal that it doesn’t just use GlobalFoundries out of the goodness of its heart. The company’s 10-K Annual Report, filed in late February, states: “Pursuant to the WSA, [Wafer Supply Agreement] the Company is required to purchase all of its microprocessor unit and APU product requirements from GF with limited exceptions.” (emphasis added).

That limited exception was Brazos. The amended agreement from a few weeks back is described as granting AMD “certain rights to contract with another wafer foundry supplier with respect to specified products for a specified period.”

Looking back over the SEC filings, its clear just how badly GF failed to meet its original agreements. Before it amended the WSA last year and switched to paying for only good die, AMD had exprected to pay GF about $1.5B. When it filed that amendment in April 2011, it stated that it expected to pay GlobalFoundries between $1.1-$1.5 billion. Actual amount paid to GF in 2011? $941M. Given that AMD was only paying for good parts, the difference between the expected and actual payouts is a rough approximation of just how bad the 32nm yield situation at GlobalFoundries really was.

Follow the crumbs back further and the story grows more interesting. When AMD spun GF off in 2009, there was a great deal of speculation that the company would move to GF when it transitioned to 32nm GPUs. As it happens, that shift was mandated in the original WSA, provided that GF met certain manufacturing targets. AMD was originally prohibited from seeking any second source for future CPUs/APUs once certain targets were hit, and was required to transition the MPUs of any products that it purchased over to GlobalFoundries within two years.

Why don’t I believe Read’s statements about “moving ahead on 28nm with GlobalFoundries?” Because every move AMD has made since GF spun off has been aimed at distancing itself from its erstwhile foundry, shedding its share of the company, and re-securing the right to manufacture products at TSMC. We won’t see the full text of the new agreement until AMD publishes its next 10-Q at the end of the quarter, but you don’t pay $700 million to regain manufacturing flexibility at the beginning of the month only to swap your plans around three weeks later.

Piledriver — AMD’s next-gen 32nm Bulldozer part — is being built at GlobalFoundries. AMD is building at least one next-generation APU at TSMC. Read may still be holding out an olive branch to GloFo, but his throwaway remark today is anything but a ringing endorsement.

Source

Hands on with Nvidia’s GeForce GTX 680

I’ve gotten to the point in my life where I’m pretty cynical about a lot of things: entertainment, politics, women… But technology is either near or at the top of the list. When I was in IT, we would always semi-joke that the easiest way to hate computers is to be around them all day — and now that I’ve moved into the tech journalism world, I’ve found that’s still basically true. Oh, I don’t hate computers, of course — if I did, I couldn’t go to work without going insane. (Or, to placate my colleagues, insaner.) Still, spending day in and day out working with computers, thinking about them, and ripping out and replacing their innards is a sure way to stop automatically believing anything anyone tells you about them — and become convinced of the unlikelihood you’ll ever see anything truly new.

So when I had a meeting with some folks at Nvidia recently about the company’s latest video card, the GeForce GTX 680 (which debuts today), I naturally couldn’t help but be a little skeptical of some of their claims. Particularly what they were telling me about the card’s power usage. Look, I have nothing against Nvidia, and a lot of its recent top-end cards have been very strong performers. But for years now, AMD has proven better, time and time again, at putting out cards that don’t require powering your PC with a nuclear fission reactor. I assumed that this would be more empty (or, at most, half-filled) promises, and that I would never actually see the promised drops in power draw.

But my skepticism, and maybe — just maybe — the outer layer of ice around my heart melted away once I actually got my hands on a GTX 680 and installed it in my usual testing PC. Using an Extech Datalogger, which measures power draw over two-second intervals, Futuremark’s 3DMark 11 DirectX 11 benchmark software in Custom mode (with all the details cranked all the way up), and a 30-inch Dell monitor, I pushed the GTX 680 and a few other competing cards to their limits — and I’ll be darned if it didn’t come out ahead. Way ahead.

Idling, the system used almost exactly 99 watts; under full 3DMark 11 load, it drew 298. If that sounds like a lot, it sort of is — big video cards are only ever going to be so power efficient, after all. But it was downright frugal compared with the Radeon HD 7970, AMD’s top-of-the-line single GPU card: It also used 99 watts when idle, but drew 342 watts under load. And the 7970 is a card that came in under Nvidia’s previous-generation flagship, the GTX 580 — in my tests this time around, the GTX 580 idled at 116 watts and drew 347 under load. The GTX 680 was even coming in only 20 watts above AMD’s second in command, the Radeon HD 7950 (101 watts when idle, 277 under load) and 30 over Nvidia’s previous next-best, the GTX 570 (111 watts when idle, 267 under load).

And all of this, I should note, with a video card that, in most cases, surpasses the 7970 in terms of performance. I admit it, I was amazed. As I said, it’s not every day that I see a product that not only lives up to its company’s hype, but exceeds it.

So how did Nvidia accomplish this? A couple of ways. First, the Kepler architecture introduced on the GTX 680 uses what Nvidia calls “next-generation Streaming Multiprocessor,” or SMX, units, which build on the previous Fermi architecture’s parallel processing technology but use even less power along the way. As each GTX 680 has eight SMX units, and thus a total of 1536 CUDA cores (three times as many as the GTX 580), there’s an enormous potential for power savings along with improved performance.

Second, power is delivered to the GPU by way of a four-phase power design, and two extra phases are dedicated to the card’s 2GB of GDDR5 memory. This is called a 4+2 phase design, and Nvidia claims it’s more efficient, even when the card is overclocked — and because of a feature called GPU Boost, which dynamically increases the clock rate above the base 1,006MHz (much like Intel’s Turbo Boost), it more or less always is.

Additional side benefits come with all this, too. The GTX 580 had a TDP of 244 watts, and Nvidia recommended your computer be armed with at least a 600-watt power supply; now, you can get away with a 550-watt power supply, and the GTX 680’s TDP has dropped to 195 watts. Even more astonishing from my standpoint is that, although the GTX 680 requires two auxiliary PCI Express (PCIe) power connectors, they’re both six-pin; most high-end single-GPU cards (including the 7970) require one six- and one eight-pin, which has traditionally been a bit less convenient. With the GTX 680, it really seems as if Nvidia has covered almost all the bases as far as power goes. (There are plenty of other interesting aspects to the GTX 680, too, but those will have to wait.)

As I see it, the best part of this Nvidia release isn’t so much what it accomplishes, but what it promises. Yes, it restructures the debate, proving that power usage can go down even as performance goes up — and, in fact, power usage can go down more. But beyond that, it raises the bar, and the expectations, for everyone. For a while now, AMD has been trading on its cards’ demure energy draw; if the company can no longer use that as a selling point, it will have to find something else. Maybe it will improve that aspect of its own cards beyond what Nvidia has done. Or maybe it will keep it the same, but compensate with drastically better performance. Or (most tantalizing, and least likely, of all), maybe it will lower prices. Whatever happens, the consumer wins, which puts the onus back on Nvidia — and on and on. More than with almost anything else in the personal computer field, video cards demonstrate how competition enhances product and end-user experiences alike.

Now that Nvidia is questionably back in the game on the power front, it might be time to rethink some of my cynicism about technology. Apparently surprises are still possible. Now if there would just be as encouraging a surprise along the same lines with entertainment or politics or women…

Read ExtremeTech’s further GeForce GTX 680 coverage

Source

Surviving the cashless cataclysm

In Sweden, just 3% of the economy is powered by coins and paper money. Public buses don’t accept cash, churches have installed card readers to take donations, and there are even some bank branches that refuse to take your money, opting instead to deal with electronic transfers only.

The European average is 9%, and in the US, the credit card motherland, the percentage is still more than twice that of Sweden: 7%. If you stop and think about it, though, none of these figures are particularly surprising. With the rise of credit cards and store cards (and card readers everywhere), PayPal, online shopping, iTunes, Netflix, and app stores, cash is feeling more outmoded by the day. When was the last time you used an ATM, anyway?

The trend is very clear: Cash is on its way out. It might take another 10 or 20 years for Sweden to get there, and longer for the US and other economies, but eventually so few businesses will accept cash that it will be relegated to jangling jars and cruddy sofa crevices. Governments will have no choice but to halt the fresh minting of coins and printing of bills, and eventually non-electronic money will dry up all together.

Change

This will mark one of the most significant paradigm shifts in modern history. At some point in the future, probably within our life times, you will wake up and your cash will be worthless. After centuries of use, we will no longer have a fungible currency. Gone are the gold sovereigns and silver dollars and nickel-alloy pennies — it’ll all be encrypted, digital bytes, stored in programs like Google Wallet. Instead of handing the shop attendant some cash, you will swipe your phone across a card reader; if you need to lend your friend some money, you’ll just key in a number, bump your smartphones together, and NFC will take care of the technicalities.

I think this will all happen naturally and fairly painlessly. Just 50 years ago, cash (and gold bullion and shares and bonds) would have represented the entire economy — and now Sweden’s on the cusp of killing cash off entirely. It’s just a matter of convenience — and if consumerism has taught us anything it’s that convenience is king. When we’re fast moving towards a world where your smartphone replaces just about everything in your wallet — from cash to tickets to ID — how can clunky coins and grubby banknotes really compete?

Sentimentalists will mourn the demise of bills and coins with national figureheads emblazoned on them, I’m sure, and there will also be an issue with older people, who in general haven’t made the leap to electronic money and would be lost at sea if cash was no longer accepted anywhere. Beyond that, though, there’s a far more pressing question that governments (and societies) will have to answer: Will we let private corporations track, collate, and cross-reference every single one of our transactions?

You see, in a cashless society every single payment is digital, which means that every transaction must be confirmed by the bank or institution that governs your money. In turn this means that every move you make will be recorded in a huge database. Your bank will know where you get coffee in the morning, the route you take to work, and if there’s a vending machine at your office it might even know where you work. Likewise, your bank will know that you like to buy things on Amazon while you’re at work, that you enjoy watching X-rated movies when you’re on the road, and that you always leave it until the last moment to buy your wife a birthday present.

At this point it’s commonplace for self-respecting libertarians to leap up and decry the awful, privacy annihilation that I’ve just described. How could you live in a world where the Rockefellers can track your every move?! they cry. Well, get this, every credit card company, bank, and sizable corporationalready tracks your transactions.

Have you ever had a call from your bank, asking if a purchase you just made was fraudulent or not? Banks employ incredibly complex software (on beefy computers) to analyze billions of transactions — ostensibly to detect fraud (which costs banks millions of dollars a year), but of course other patterns can be detected as well. Just as one example, BillShrink has worked with 2,000 banks to analyze your buying habits, and then to provide targeted coupons on your monthly statement. Obviously it’s rather cool to automatically receive a $5 voucher for McDonald’s if your bank detects that you spend $100 under the greasy shadow of the Golden Arches every month, but it’s a little bit creepy too. It is, after all, exactly the same as Google or Facebook’s targeted advertising — but possibly even more accurate.

The fact is, cash is anonymous — and humans love anonymity. It’s not that we require it all the time, but a free society demands that the option is there. Whether you’re paying for something embarrassing, like a prostitute, or privately funneling money to a charity, it’s nice to have theoption of paying with a currency that cannot be directly tracked back to you.

On the flip side, though, corruption, tax avoidance, and many other crimes are all linked to the fact that cash is very hard to track. Taking Sweden as an example again, in 2008 there were 110 bank robberies; in 2011, there was just 16, the lowest since records began 30 years ago. In Italy, where cash is alive and kicking and graft and tax evasion are rife, the Prime Minister recently announced that he plans to limit cash payments to 1000 euros ($1300), down from 2500 euros. It is not a coincidence that Italy and Greece, two of the most volatile economies in Europe, both have strong cash cultures.

It would seem that you can’t have your cake and eat it too. Money is either accounted for by a central authority (electronic transactions), or it isn’t (cash). You are either protected (and snooped on) by the auspices of governments and mega corporations, or you’re not. There is another option, however.

Saved by technology, again

The important thing to remember is that untraceable, anonymous cash isn’t dead yet. Bills and coins might be on the way out, but that doesn’t mean that they can’t be replaced by an electronic equivalent. If Bitcoin has proved anything it’s that you don’t need a central authority to manage electronic trades — and just look at iTunes gift cards and pre-paid Visa and MasterCard credit cards; they’re not quite as anonymous as cash, but they’re close.

Bitcoin has failed because it doesn’t have the backing of a central bank, but there’s nothing to prevent national governments from issuing irreversible, cryptographic cash that could then be stored on “credit chips” (i.e. your smartphone, or even a coin-sized piece of solid-state memory). This digital cash would act just like its physical counterpart: You could deposit it in your bank, pay for groceries, or even lose it down the back of your sofa. You could pay people by physically giving them a credit chip, or by transferring them via smartphone. Digital cash would still have the same issues as normal cash — graft, tax evasion, robbery — but that’s the price we pay for freedom.

The benefits of a purely digital economy are massive. If you think about it, dealing with physical cash is incredibly difficult. You need cash registers (which aren’t cheap); you have to constantly monitor employees for theft; you need a safe to keep the money in between collections, and collections themselves are expensive too. A wholeindustry revolves around the safe handling, transportation, and counting of coins and bills. With digital currency, all of those difficulties disappear. With PayPal, you can set up an online business in minutes. With Square, your iPhone or Android smartphone becomes a point of sale.

And then there’s the analytics! If you think you have problems managing your money, imagine if you could plug your credit chip into your computer at the end of the month and find out exactly how you spent (and wasted) money. You wouldn’t have to give up this info to a third party — but if you did, there would be even more advantages. Imagine a group buying website (like Groupon) that analyzes your credit chip and matches you to 10 other local Coke drinkers; instead of each buying a 12-pack from the supermarket, you could club together and buy a case from the wholesaler.

Remain vigilant

Anonymous, digital cash isn’t a given, though. We will probably have to fight for it. Conspiracy theorists are exceptionally talented at conflating fact and fiction, but they’re not wrong when they say that totalitarian governments and megacorps really want monitor our every move. The demise of physical cash is the perfect opportunity for governments, under the sway of powerful lobby groups, to transition to trackable currencies that are governed by a central computer. This will probably be one of the biggest hurdles that we will face in the twenty-first century.

Swedes, when the referendum to abolish paper money finally rolls around, be sure to vote for the creation of a digital, anonymous equivalent. The rest of the world will watch and take its cues from you.

Source

$1.5 billion: The cost of cutting London-Toyko latency by 60ms

Starting this summer, a convoy of ice breakers and specially-adapted polar ice-rated cable laying ships will begin to lay the first ever trans-Arctic Ocean submarine fiber optic cables. Two of these cables, called Artic Fibre and Arctic Link, will cross the Northwest Passage which runs through the Canadian Arctic Archipelago. A third cable, the Russian Optical Trans-Arctic Submarine Cable System (ROTACS), will skirt the north coast of Scandinavia and Russia. All three cables will connect the United Kingdom to Japan, with a smattering of branches that will provide high-speed internet access to a handful of Arctic Circle communities. The completed cables are estimated to cost between $600 million and $1.5 billion each.

All three cables are being laid for the same reasons: Redundancy and speed. As it stands, it takes roughly 230 milliseconds for a packet to go from London to Tokyo; the new cables will reduce this by 30% to 170ms. This speed-up will be gained by virtue of a much shorter run: Currently, packets from the UK to Japan either have to traverse Europe, the Middle East, and the Indian Ocean, or the Atlantic, US, and Pacific, both routes racking up around 15,000 miles in the process. It’s only 10,000 miles (16,000km) across the Arctic Ocean, and you don’t have to mess around with any land crossings, either.

The massive drop in latency is expected to supercharge algorithmic stock market trading, where a difference of a few milliseconds can gain (or lose) millions of dollars. It is for this reason that a new cable is currently being laid between the UK and US — it will cost $300 million and shave “just” six milliseconds off the fastest link currently available. The lower latency will also be a boon to other technologies that hinge heavily on the internet, such as telemedicine (and teleconferencing) and education. Telephone calls and live news coverage would also enjoy the significantly lower latency. Each of the fiber optic cables will have a capacity in the terabits-per-second range, which will probably come in handy too.

Beyond the stock markets, though, the main advantage of the three new cables is added redundancy. Currently, almost every cable that lands in Asia goes through a choke point in the Middle East or the Luzon Strait between the Philippine and South China seas. If a ship were to drag an anchor across the wrong patch of seabed, billions of people could wake up to find themselves either completely disconnected from the internet or surfing with dial-up-like speeds. The three new cables will all come down from the north of Japan, through the relatively-empty Bering Sea — and the Arctic Ocean, where each of the cables will run for more than 5,000 miles, is one of the least-trafficked parts of the world. That said, the cables will still have to be laid hundreds of meters below the surface to avoid the tails of roving icebergs.

Each cable will be laid by a pair of ships: an ice breaker that leads the way, and a cable ship. Until now it has been impossible to lay cables in the Arctic Ocean, but the retreat of the Arctic sea ice means that the Northwest Passage is now generally ice-free from August to October; a big enough window that cable can be laid fairly safely. Existing cable ships (and there aren’t many of them) are all outfitted for balmier climes, so all three cables will require the use of a polar ice-rated ship that has been retrofitted to carry cable-laying gear.

Read more about the secret world of submarine cables.

For more information on the Russian Optical Trans-Arctic Submarine Cable System (ROTACS), check out the Polarnet Project (machine translated).

The Arctic Fibre and Arctic Link websites have information on the North American cables.

[Image credit: New Scientist]

Source

Does Daisey’s lie change what we know about Apple’s manufacturing?

Almost two months ago, we published a story detailing the results of several investigations into Apple’s manufacturing partner, Foxconn. The allegations of shoddy inspections and poor worker conditions came just days after Apple announced record earnings and profits. Apple hotly denied all of the claims the New York Times’ articles leveled against it and partnered with the Fair Labor Association to verify that Foxconn’s factories were in tip-top shape.

The back-and-forth ignited a great deal of general interest. Mike Daisey, the man behind the one-man show The Agony and Ecstasy of Steve Jobs, went on the radio program This American Life to discuss his visits to Foxconn, interviews with hundreds of workers, and his experience and observations of the company. The resulting TAL episode broke download records and fanned the flames of debate — until follow-up analysis proved that Daiseymade up most of his story. This raises the question of just how far Daisey’s lies percolated into the general body of what’s known and how much of the various analysis done by the media holds up.

Here’s the good news: Daisey’s fabrications were confined to TAL. The NYT stories we based our own reporting on and the investigation Apple launched as a result, regardless of whether or not you agree with the tone of the former or the need for the latter, are solid. This American Life host Ira Glass interviewed Daisey as part of its retraction of the original story; Glass’s normally unflappable demeanor ripples with anger at having been taken in in such a fashion.

Glass states:

Mike’s monologue in reality is a mix of things that actually happened when he visited China and things that he just heard about or researched, which he then pretends that he witnessed firsthand. He pretends that he just stumbled upon an array of workers who typify all kinds of harsh things somebody might face in a factory that makes iPhones and iPads. And the most powerful and memorable moments in the story all seem to be fabricated.

The bits about child workers? Fabricated. Meeting workers poisoned by n-hexane? Fabricated. Showing an iPad maker who lost the use of a hand while building them for Foxconn? Made that bit up, too. Some of these lies were woven into his one-man show, some were created especially for TAL. In both cases, Daisey defends his statements, claiming that “if I untied these things, that the work, that I know is really good, and tells a story, that does these really great things for 15 making people care, that it would come apart in a way where, where it would ruin everything.”

The “big picture” facts about Foxconn, its relationship with Apple, and the question of whether or not Apple bears a unique responsibility to push for improved factory conditions as part of its agreements with suppliers are all still valid. That’s where this post might have ended, if Daisey had managed to keep his mouth shut. Over on his website, the actor has posted a whinging, self-pitying rebuttal that criticizes the recent TAL retraction as “pulled out of context.” He then writes:

Given the tenor of the condemnation, you would think I had concocted an elaborate, fanciful universe filled with furnaces in which babies are burned to make iPhone components, or that I never went to China, never stood outside the gates of Foxconn, never pretended to be a businessman to get inside of factories, never spoke to any workers.

Especially galling is how many are gleefully eager to dance on my grave expressly so they can return to ignoring everything about the circumstances under which their devices are made. Given the tone, you would think I had fabulated an elaborate hoax, filled with astonishing horrors that no one had ever seen before…

But understand that if you felt something that connected you with where your devices come from — that is not a lie. That is art. That is human empathy, and it is real, and even if you curse my name I hope you’ll recognize that and continue reading, caring, and thinking.” (emphasis added)

He’s wrong. It’s a lie. More specifically, it’s deliberate emotional manipulation. As a journalist, I get the allure of this particular slippery slope. The n-hexane poisonings? Real. Could they have been prevented? Absolutely. Are the survivors of that event stuck dealing with crippling side effects for the rest of their lives? Yes. Did he go to China to investigate the labor situation? Yes he did. Put those three facts together, and the temptation to go one step further and fabricate the meeting of an affected individual is enormous. It’s the linchpin that transforms an article or production from a simple discussion of Someone Else’s Problem to a personal intersection of two living, breathing humans.

Daisey’s actions are understandable; his post makes it clear that he believes he was serving a greater moral imperative. The fatal flaw in his reasoning is that the situation didn’t require him to lie. It never does. The original New York Times stories, which have withstood fact checking, put a very human face on the problems Foxconn workers face. The image that accompanied our original story (pictured above) was, to the best of our knowledge, taken at the first Foxconn aluminum explosion in May of 2011. It’s uncomfortable to look at precisely because it makes the viewer a part of the pain and fear that accompanied the disaster.

Movies like Life is Beautiful, Schindler’s List and Saving Private Ryan put human faces on World War II by meshing historical fact with narrative fiction. They, and literally hundreds of other films on various topics, are movies that contain truths, even when they aren’t factually accurate to the last event. Daisey has conflated the two and fell into the trap of using his personal credibility to humanize and tell the story of what goes on to build the tablets and smartphones we gobble up like candy.

The worst part of Daisey’s lying is that he’s damaged the credibility of the cause he sought to champion. Thanks to him, people who followed the situation peripherally will conclude that Apple got accused of a bunch of bad things in China but the guy who did it was “lying all along.” The public mind rarely pauses to parse nuance. In reality, Daisey’s story was only a small piece of the investigative puzzle; his contribution was modest at best. His lie, however, is disastrous. No amount of self-congratulatory blogging on his part can hide that.

Read our review of Daisey’s play, The Agony and Ecstasy of Steve Jobs

Source

Seagate hits 1 terabit per square inch, 60TB hard drives on their way

Seagate has demonstrated the first terabit-per-square-inch hard drive, almost doubling the areal density found in modern hard drives. Initially this will result in 6TB 3.5-inch desktop drives and 2TB 2.5-inch laptop drives, but eventually Seagate is promising up to 60TB and 20TB respectively.

To achieve such a huge leap in density, Seagate had to use a technology called heat-assisted magnetic recording (HAMR). Basically, the main issue that governs hard drive density is the size of each magnetic “bit.” These can only be made so small until the magnetism of nearby bits affects them. With HAMR, “high density” magnetic compounds that can withstand further miniaturization are used. The only problem is that these materials, such as iron platinum alloy or a sprinkling of table salt (really), are more stubborn when it comes to changing their magnetism (i.e. writing data) — but if you heat it first, that problem goes away.

HAMR, which was originally demonstrated by Fujitsu in 2006, adds a laser to the hard drive head. The head seeks as normal, but whenever it wants to write data the laser turns on (pictured below). Reading data is done in the conventional way. Just so you understand how small the magnetic bits are in a HAMR drive, one terabit per square inch equates to two million bits per linear inch; in other words, each site is just 12.7 nanometers long — or about a dozen atoms.

In theory, HAMR should allow for areal densities up to 10 terabits per square inch (magnetic bits just 1nm long!), and thus desktop hard drives in the 60TB range. Meanwhile, conventional perpendicular recording is expected to hit one terabit in the next few years, but the roadmap to greater densities isn’t very clear. There is no word on the cost of HAMR drives, or whether the addition of a laser will significantly impact power consumption or I/O performance.

The biggest winner from larger hard drives, of course, is cloud storage and computing — but then again, the other angle is that you’ll have so much local storage that the cloud seems a bit pointless,especially when we all have 100Mbps internet connections. But then again, with the unstoppable surge of smartphones and tablets and flash memory, do mechanical hard drives really have a future in consumer electronics?

Read more at Seagate

Source

ISPs to become copyright cops this summer

The internet rose up with great vengeance and furious anger earlier this year when it seemed all but sure that the Stop Online Piracy Act (SOPA) would pass the US House of Representatives. A well-orchestrated online campaign killed that bill, but this coming summer the private sector could be the next big threat to online freedom. According to the RIAA’s chief lobbyist, this July is when the so-called “graduated response” anti-piracy measures will be going into effect at the nation’s largest ISPs.

A global trend

The entertainment industry has been working to get harsh three-strikes laws installed in many nations with some success. France instituted a system called HADOPI last year, and the European recording lobby, IFPI says that it’s working. In this three-strikes system, if a user gets three infringement notices from their ISP, they are cut off from the internet and cannot re-subscribe for at least two months, but as long as one year, depending on the case.

In France, 763,000 subscribers have gotten at least one notice, 62,000 a second, and 165 a third. The actual decline in P2P use is heavily debated, but the most likely decline is 26% in just a year. However, rates of piracy in France appear to be on the upswing again.

Other countries like New Zealand and the UK have passed similar measures. The three-strikes system in the UK, known as the Digital Economy Act, has been opposed by various ISPs concerned about the cost of policing users and maintaining a blacklist of infringers. The High Court in the UK has just dismissed those claims, which could clear the law for implementation, and could mean users accused of infringement will be disconnected.

The coming storm

While the entertainment lobby was unable to get a law passed in the US, major ISPs like Comcast, Cablevision, Verizon, Time Warner Cable, and more have signed onto a similar plan of their own volition. Certainly Comcast, which owns NBC-Universal, was in favor of protecting its own content. Possible legal action got the others in line.

Unlike some of the laws overseas, there is no firm procedure for what to do with a (suspected) American pirate. The ISPs involved will send out one or two educational emails to politely remind the account holder not to run illegal torrents. After that, things get more serious with a stern warning that consequences are coming. ISPs can choose what punishments to implement to stop the user from further torrenting. These so-called “mitigation measures” could include throttling of the connection speed, or temporary disconnection. The one step that no US ISP has been bold enough to put forward is a permanent disconnection and blacklist of problem P2P users.

The deal was worked out last year, but according to the RIAA and MPAA it has taken months to get the systems into place to track and notify users. Considering there will be no judicial oversight, this is essential to reduce (we hope) false positives. It was this agreement last year that gave the entertainment lobby the courage to try and push bills like SOPA through, and we saw how that turned out. There are some serious, endemic problems with ISPs policing copyright, however.

The inherent problems

Any notices sent out will go to the account holder, even if another individual is using the connection for infringement. This has the potential to catch many less tech-savvy users totally off guard. If someone’s teenager is running torrents of movies and music with reckless abandon, it’s the parents that are going to be dealing with the ISP. Likewise, anyone with an unsecured wireless network (which does still happen) could be taking the heat for the whole neighborhood.

If your ISP is among those participating in the initial rollout of the graduated response, there’s not much you can do. Cable operators in the US, which most large ISPs are, have the benefit of a natural monopoly. Natural monopolies have been granted to businesses throughout history in order to get infrastructure built out quickly. Your cable company actually owns the line to your home, so no one can compete with it.

Considering the lock-in, many are asking if the service providers aren’t going about this the wrong way. The ISP should stand with the user they have been allowed exclusive access to, not with an entertainment industry lawyer. It should be unthinkable that your ISP, the gatekeeper of the internet and de facto public utility, is proactively spying on you for a third-party. If some ISPs choose to engage in such activity, maybe it’s time to allow more cable operators into the market and deregulate cable connections like phone lines are.

When IP addresses are treated as people, it is always possible for the wrong individual to get the letters and consequences for another person’s actions. We can only hope that the initial educational notices come with instructions on securing one’s network. It seems only fair that the ISP should take an interest in protecting customers from unfair penalties resulting from a lack of technical know-how.

It’s a good thing that ISPs have decided against blacklisting P2P users for now, but that could change if the system rollout goes smoothly. Still, the prospect of having your connection throttled is not too exciting. A speed slow enough to make torrents unrewarding would also cut you off from much of the interactive web, such as YouTube, Netflix, or Hulu streaming. Additionally, a system based around torrents could just push P2P users to other sources of content like direct download, or that old favorite Usenet.

Are ISPs going to be hardline or accommodating when the new anti-piracy system comes out? It’s hard to say, but the online community at large will most likely react negatively when the serious infringement letters start going out and the first user finds their connection slowed to an unusable speed. This summer is definitely going to be interesting.

The Pirate Bay plans fleet of low orbit server drones to evade the authorities

In what reads like an April 1 missive, The Pirate Bay has announced that it plans to launch “Low Orbit Server Stations” that will act as a front line defense against future take-down attacks and censorship attempts on the world’s largest torrent site.

“With the development of GPS controlled drones, far-reaching cheap radio equipment and tiny new computers like the Raspberry Pi, we’re going to experiment with sending out some small drones that will float some kilometers up in the air,” The Pirate Bay says. This would, of course, be the first true example of cloud computing — and a whole new spin on the meaning of “pirate radio.”

Details are sketchy, but it sounds like these low orbit servers will act as proxies that redirect users to geographically-hidden servers that host the actual torrent indexes and trackers. At the moment, TPB uses a similar setup: public front-end servers (mostly in Sweden), and secret back-end servers that are (apparently) hidden away in a mountain bunker. Orbiting servers would replace some or all of these front-end servers, with the ostensible purpose of making it very hard for the authorities to shut them down.

The whole idea is utterly crazy. For a start, TPB wants to use flying drones — something like a quadcopter or the NASA Pathfinder UAV — but these aren’t “low orbit” at all. Low earth orbit is 100 miles up and requires a launch vehicle capable of achieving speeds of 17,000 miles an hour. At “some kilometers up in the air,” these drones would require significant power to stay afloat, and that’s even before the power required to transmit megabits per second over a wireless connection. Realistically, these drones would have refuel regularly — and the coming and going of a fleet of drones would probably annul many of the benefits of such a system.

Then there’s airspace issues. TPB tells TorrentFreak that the drones will fly above international waters, but at a few miles up there’s a serious risk of being hit by airplanes and helicopters. Most countries allow for low-flying aircraft (think model airplanes) without a license, but again the Pirate Bay drones would run into issues when coming in to refuel.

In our opinion, a far more realistic approach would be a helium balloon (pictured above), which can stay airborne for days and carry a fairly sizable load. A balloon might even have enough surface area to provide solar power for the on-board computer and antenna. You could tether the balloon to a buoy in international waters. It might just work.

Source