mdc's shared items

Flooding in Thailand has spurred discussion of whether or not the natural disaster has created a window of opportunity for SSD manufacturers to break into new markets. OCZ has clearly decided that it has. The company has announced that it plans to introduce new SSDs based on triple-level cell (TLC) technology next year and has signed a contract with SMT Manufacturing to fill “any need in excess of current client forecasts.”

TLC NAND has been available for several years but has thus far been mainly deployed in low-cost USB sticks and other forms of removable flash media where performance and endurance are potentially less important than the need for higher capacities. Speaking at the Needham HDD & Memory Conference last week, the CEO of OCZ, Ryan Petersen, announced that the manufacturer would leverage its recent Indilinx purchase to create the new drives. The controller will ship in January, with products positioned “for low-end servers, consumers, laptops, retail; those sub-segments that really can adopt TLC-based solutions.”

As the name implies, TLC NAND, when it isn’t chasing waterfalls, stores three bits of data per cell compared to two bits of data for multi-level cell (MLC – the technique used by almost every consumer-grade SSD) and one bit for single-level cell (SLC, usually used by enterprises). The ability to pack 50% more data into the same amount of space will theoretically allow TLC devices to hit price points significantly below that of even the cheapest current MLCs. Unfortunately, the situation is far more complex than simply stuffing in an additional bit of data and calling it a day.

Voltage thresholds for MLC NAND. TLC requires eight distinct peaks

The more bits of information stored per cell, the greater the number of distinct voltage states that must be maintained within each cell. The number of voltage levels increases exponentially relative to the number of bits — SLC required just two voltage thresholds, MLC uses four, and TLC needs eight. Manufacturers have experimented with packing four bits of data per cell, but have not been able to achieve the 16 separate voltages necessary to effectively commercialize the technology.

The cost advantage of TLC is further attenuated by the need for additional and more complex ECC capability within the controller. Manufacturers will also need to provide a significantly higher amount of overprovisioned space to prevent long-term data loss and performance degradation. For those of you who aren’t aware, overprovisioning refers to the practice of including “invisible” amounts of flash on a drive that’s used when exhausted blocks begin to fail. Finally, there’s a performance and power tradeoff. Storing more data per block means reads and writes both take longer and require more power relative to MLC drives. All of these issues could leave TLC drives comparing poorly against higher-end HDD products.

Of all the issues in play, we suspect reliability is going to be the toughest nut for OCZ to crack. The company will have to tread carefully to find a sweet spot between offering SSDs at lower price points, without sacrificing the performance that make them attractive or harming long-term data retention.

Yes, it is hard to believe, but email turned 40 last month. No one can question that email has transformed the way business is conducted. According to Radicati group, the average worker processes more than a hundred daily emails, and business email accounts make up only about a quarter of the total of more than 3 billion. Sometime in the next few years, Radicati predicts that the number of IM accounts will exceed email for the first time.

Email and my own working life have been closely intertwined as well. I started using email in 1983 and over the years I have used more than three dozen different systems and sent thousands of messages. So I thought I would put together some important milestones of my own usage and show you how email has changed from those early days. In my working life, I published my first book on corporate email use (seen at left), published a weekly email newsletter and wrote many articles about various email products for dozens of publications, including this one.

Sponsor

And, yes, I still try to have less than ten messages at the end of each day in my inbox.

• 1983: Started using both MCIMail, one of the first global systems that was available to the public (the Internet was not yet available to the average worker) and a conferencing system called EIES. One job I had back then was to write automated scripts for processing messages between the two at a small software firm.
• 1984: At an insurance company, I used an IBM mainframe email product called DISOSS for internal communications.
• 1986: Used 3Com's 3+Mail for internal communications at PC Week. This was one of the early LAN-based email programs. We thought we were hot stuff because we could hook up our remote offices around the country to it, something now taken for granted.
• 1987: Wrote my first column for PC Week about hotels, modems, and email. Today the problem still remains, just replace Wifi for the modems.
• 1989: Covered the launch of Lotus Notes, one of the first collaborative software tools, and lobbied Ziff Davis, where I worked, to start using it in place of 3+. They eventually did a few years' later.
• 1990: First Internet email addresses used by writers in their bylines. We used Network Courier LAN-based email, which was the precursor to Microsoft Exchange and Outlook.
• 1991: Began to chart ways to send emails between two formerly disparate systems, using various gateways. The rise of Soft-Switch, which at its height could connect more than 50 different systems. They were eventually acquired by Lotus. Again, something taken for granted now.
• 1992: I was one of the first wireless email users of a product called RadioMail, which eventually became the BlackBerry. It worked with a one-pound radio and a one pound HP palmtop, shown at right.
• 1993: Obtained my first Internet domain name, strom.com, for free from Network Solutions by requesting it from them via email. Before then, private businesses couldn't really become masters of their own domains easily.
• 1994: Groupware was the big deal back then, and Novell's Groupwise was one of the best. Too bad.
• 1995: Began the first of a series of weekly email newsletters using a collection of Unix scripts. Still writing them, using a Linux server and Mailman.
• 1996: Experimented with Intermind's push technology for notifications instead of sending emails for my newsletter. Didn't last very long. Push pooped out quickly.
• 1997: Gave up my laptop and used borrowed computers when traveling. That didn't last very long either. Did have the very early smartphone from AT&T that used broadband (well, it wasn't all that broad) cellular data called CDPD, the precursor to what we all use today on our phones.
• 1998: Co-wrote my email book with Marshall Rose, the inventor of the POP protocols. The book covered the more popular email programs at the time, which included Lotus cc:Mail (extinct), Netscape Messenger (extinct but replaced by Thunderbird you could say), Eudora Pro (still very much alive, although no longer under the thumb of a phone handset maker thankfully), Compuserve (not extinct but should be), AOL (ditto and back then it was on v3), and Microsoft's Outlook Express (v4 that came with IE v4, and replaced with the Mail app in Vista and now Windows 7). Penn Jilette, of Penn and Teller fame and an early email user, wrote our forward to our book. Out of that research is this Web page that I haven't touched since then that shows the state of email encryption interoperability. Luckily, it has gotten better.
• 2001: Was a regular user of Lotus Notes, which by then had been purchased by IBM, while working back at CMP.
• 2002: Wrote about Michael Dell's bandwidth separation anxiety here, probably one of the first of many popular instances of cutting off email.
• 2004: At the annual VIP economic forum love fest gathering in Davos, Bill Gates proclaimed: "Two years from now, spam will be solved." Right. Not even close on that one Bill.
• 2005: Began using Mozilla's ThunderBird as my regular email client. Here is a story about the trials then.
• 2006: Switched hosting my various email domains over to Google Apps. For free. Began using Gmail as my regular email client, it wouldn't talk IMAP for another year.
• 2008: Reminisced about ten years after my email book in my post here. Vint Cerf wrote this then too about ten years of using the Internet.
• 2009: First of many "email is dead" articles in WSJ and elsewhere analyzed here.
• 2011: The latest in a series of days without email proposed to make some obscure point. We are still using it though, warts and all.

Please share you favorite email stories in our comments, especially ones from the early days. Discuss

Andrew Pollack / New York Times:
John R. Opel, Who Made I.B.M. a Colossus, Dies at 86  —  John R. Opel, who presided over I.B.M. in its final period of dominance in the information-processing industry and oversaw the company's move into personal computers, died on Thursday in Fort Myers, Fla. He was 86.

The International Atomic Energy Agency issued a critical report Tuesday saying that it has "serious concerns" that Iran's nuclear program is geared toward developing nuclear weapons.

Each day since Dennis Ritchie passed away my 1978 copy of “K&R” — The C Programming Language, by Brian W. Kernighan & Dennis M. Ritchie – has glared balefully at me asking why I haven’t written something about his effect on my professional life. Indeed, the book, which replaced the stapled, hand-copied notes that went before it, has followed me across the country and to and from many companies. Its voyage parallels my own, and that of much of the computer industry.

First, in the 1970s, when C was avant garde – a research tool used by Bell Labs and a few Universities to help provide a serious computing environment on small machines – I used the book to help me get my Computer Science degree, writing a compiler that could run the code in the book. Then in the early 1980s, the book was my reference manual for developing real C compilers for monster mainframes, which then ruled the computing world.

Next, the book went with me, off to workstation maker Sun Microsystems, where Unix had finally reached the big time. Sun and dozens of other companies had found religion in the straightforward and easily-licensed C language and Unix operating system, ushering in the dawn of open systems and multi-vendor collaboration on system software. C was indeed king there.

The book got a long rest after that, C having been superseded for me by its later big brothers, C++, C#, and its spiritual descendant, Java. The simple sample programs in the original 228-page book, starting with the timeless classic “hello, world,” having been outpaced by monstrously large tomes on specialized programming for window systems, frameworks, and using esoterica like patterns. I missed the book’s straightforward and almost folksy writing style, which made even the most timid reader feel like a superhero at the keyboard.

But the book wasn’t finished with me. When our daughter dove into the world of robotics, I found that the C language, in close to its original form, was alive, well, and thriving. The book came off the shelf in service of teaching another generation a simple, elegant way to program that allows the developer to be directly in touch with the innards of the computer. The lowly integer variable — int — has grown in size over the years as computers have grown, but the C language and its sparse, clean, coding style live on. For that we all owe a lot to Dennis Ritchie.

Dennis Ritchie will be remembered for many accomplishments, starting with being the co-creator of the Bell Labs Unix operating system, forerunner of not just today’s Unix, but FreeBSD, Linux, Android, and Mac OS — for which he and Thompson won the Turing award and the National Medal of Technology. But for me I’ll always remember best the excitement I felt at being able to explore a programming language on my own when I first got my copy of The C Programming Language.

Walter S. Mossberg / AllThingsD:
Dialing Up 20 Years of Gadget Reviews  —  I began writing these Personal Technology columns 20 years ago, in October 1991, with the aim of reviewing computers and other digital products for average, mainstream users.  The first line of my first column was: “Personal computers are just too hard to use, and it's not your fault.”

Arik Hesseldahl / AllThingsD:
HP's Project Moonshot Aims to Recreate Servers, Again  —  In the late 1990s, there was a shift in thinking around how servers could be made and how several of them are designed to share space.  The idea was to pack several server computers in the space that had previously been required …

Zeljka Zorz / Help Net Security:
Nearly 50 chemical, defense companies hit with cyber espionage attacks  —  Nearly 50 (and quite possibly more) companies in the chemical, defense, and other sectors have been hit with a spear phishing campaign carrying a backdoor Trojan with the ultimate goal of exfiltrating R&D and manufacturing information …

How does Apple’s Siri really work? A SmartPlanet article lays out how voice recognition on a smartphone really works, step by step.