• On CHOW: Sexy vampire party
March 7, 2008 4:00 PM PST

Intel set to take leap in solid-state drives

by Brooke Crothers
  • Font size
  • Print
  • Post a comment

Intel doesn't enter markets gently. Its new high-capacity solid-state drives (SSDs) are expected to jolt a market currently dominated by Samsung, Toshiba, and SanDisk.

Intel's current offering, the Z-P140 PATA solid state drive

Intel's current offering: the Z-P140 PATA solid-state drive.

(Credit: Intel Corp.)

At the moment, Intel offers small-capacity chip-level (what are called Thin Small Outline Packages or TSOPs) technology that provides end-product sizes ranging up to 16GB. But this modest line of products will get a big boost in the second quarter when Intel offers 1.8- and 2.5-inch SSDs ranging from 80GB to 160GB in capacity, said Troy Winslow, marketing manager for the NAND Products Group at Intel. Intel's new SSDs will compete with Samsung, for example, which is slated to bring out a 128GB SSD in the third quarter.

With new competition, drive speeds will jump. Currently, the fastest SSDs from companies like Samsung approach 100MB/second for reading data. "What I can tell you is ours is much better than that," Winslow said. Hard drives typically read data at about half this speed.

"We will be supplementing our product line with a SATA offering," he said. Serial ATA, or SATA, is an interface used in high-performance hard disk drives. Intel's products will be based on the SATA II specification that offers speeds of 3 gigabits (Gb) per second. Samsung is now shipping 64GB SSDs to Dell using the same technology.

"When Intel launches its...products, you'll see that not all SSDs are created equal," Winslow said. "The way the SSDs are architected, the way the controller and firmware operates makes a huge difference," he said, referring to the chip (controller) that manages the SSD and software (firmware) that the controller uses.

Intel believes 2008 is the year of the SSD. (See SSD primer below.) "For the first time, flash is going into the compute environment. In the last nine years or so when it experienced all of its growth, this has been in digital cameras and USB keys," Winslow said. But now flash memory, in the form of SSDs, will be used as the main storage device in PCs. "When you're putting all your critical applications and data into notebook or server (SSDs), who knows those markets better than the manufacturer that's supplying the world with CPUs," Winslow added.

While the latter statement seems like typical marketing spin, it's more than just spin in Intel's case. The largest chipmaker in the world is in a competitive position because it already supplies many of a PC's core components including the processor, chipset, communications silicon, and in some cases, the graphics processor. Add the main storage device to the mix, and--with the exception of an optical drive and screen--that's all the core component in a notebook PC.

But to be competitive with hard drives, SSD prices have to come down--a lot. In many cases, upgrading from a hard drive to an SSD in a notebook can mean paying an extra $1,000. Intel, like Samsung and Toshiba, sees steep declines in cost in the next two years. "Price declines are historically 40 percent per year," Winslow said. "And in 2009, a 50 percent reduction, then again in 2010."

Also, like Samsung, Intel sees SSDs playing a role in the server market as a "performance accelerator." Winslow said that Intel recently did a video-on-demand demonstration where it streamed 4,000 videos simultaneously. Just to do the streaming (not to store the video), it took 62 15,000 RPM (very high-performance) hard drives, he said. "We were able to replace those 62 hard drives with 10 SATA (SSD) technology drives," he said.

Finally, Winslow addressed the price collapse in the flash market in general--a topic that generated a lot of press after the Intel analyst meeting on Wednesday. "A majority of flash is being sold in very cyclical consumer electronics devices. Q1 and Q2 are soft quarters," he said. On top of this, suppliers continue to shrink manufacturing process technologies, leading to more capacity at lower cost, he said.

SSD Primer, Part 1: SSDs are based on flash memory chip technology and have no moving parts. Hard-disk drives (HDDs), in contrast, use read-write heads that hover over spinning platters to access and record data. With no moving parts, SSDs avoid both the risk of mechanical failure and the mechanical delays of HDDs. Therefore, SSDs are generally faster and more reliable. The catch is the cost: SSDs are currently much more expensive than HDDs.

SSD Primer, Part 2: Intel will be shipping in the second quarter a Multi-Level Cell or MLC solid-state drive. This is a more sophisticated technology than current Single-Level Cell or SLC. The advantage is larger capacity since MLC uses multiple levels per cell to allow more bits to be stored. The disadvantage is more complexity which can result in lower performance. "Inherently, MLC is slower and inherently fewer write cycling endurance," Winslow said. Intel, however, has technology that will get around these problems, he said.

Intel Flash/SSD capacity: Intel and Micron have a joint venture called IM Flash Technologies. Both companies are currently making flash on a 50-nanometer process with plans to move to 40nm later this year. There are three NAND flash fabrication plants and one more currently being built in Singapore. The Intel-Micron venture provides funding for the development of silicon technology and the capacity to produce that silicon, according to Winslow. But marketing and end-product decisions are "absolutely separate," he said.

Brooke Crothers has served as an editor at large at CNET News, an editor at Dow Jones' Asian Wall Street Journal Weekly, and a senior editor at InfoWorld. His CNET blog covers chip technology and computer systems, and how they define the computing experience. He also contributes to The New York Times' Bits and Technology sections. He is a member of the CNET Blog Network and is not an employee of CNET. Disclosure. Follow Brooke on Twitter @mbrookec.
Recent posts from Nanotech - The Circuits Blog
Will the 'smartbook' be a better Netbook?
Firefox: Heat and the CPU usage problem
AMD upgraded as 'Fusion,' 16-core chip future looms
Dell's 'Mr. A' is a key figure in Intel defense
AMD unveils 'world's fastest' graphics card
Intel an investor in storage firm for Apple users
Chip designer ARM leads Android alliance
Intel unveils supercomputer chip, NEC partnership
by LcdGreg March 7, 2008 6:40 PM PST
I've known since cartridge video game days that no moving parts = faster computing, now more so since the read speeds are actually faster than hard drives, it may take a few years but maybe the capacity will as well exceed hard drives which currently can fit 500 GB in a 2.5" drive, I'm sure that they will milk it for profits a long while though!
Reply to this comment
by the_inventor March 8, 2008 10:27 AM PST
Garbage In = Garbage Out. Flash memory is unreliable over time.

The cost of having PERMATELY FIXED storage is too high.

Read about the real future from someone who predicted solid state storage in 1980 but
has since moved on to Spintronics and Holographics storage.

http://colossalstorage.net
Reply to this comment
by eyetexasbob March 11, 2008 5:55 AM PDT
The life span of the SSD is 50-100 years, depending on who you believe. For most of us, that is plenty of time. The only permanent storage is the mind of God. While colossal storage is cool, in theory, the SSD is now. If you are looking for sometime that lasts eternally, avoid human inventions.
Reply to this comment
by satanmode March 17, 2008 4:32 AM PDT
can anyone explain WHY write cycles wear out a SSDs, all i read on the internet is that there is a limitation to every block written.
is it the semiconductor that wears out after a million electric fluctuations (write cycles) ??
Reply to this comment
by Astonct March 26, 2008 11:08 AM PDT
Solid State Drive Array for the financial industry, especially large volume trading firms. They actually use Intel processors as well. Funny how Intel is sort of etching into the storage arena piece by piece...

A.
Reply to this comment
by sadchild May 27, 2008 8:11 AM PDT
Does SATA II mean SATA 3Gb/s?
No. The term SATA II was the name of the original organization formed to develop the SATA technology specifications. In 2004, the group changed its name to the Serial ATA International Organization, or SATA-IO.

The term SATA II is no longer valid. Unfortunately, some customers have mistaken the SATA II designation as a moniker for the 3Gb/s data transfer rate introduced with SATA-IO?s 2005 release of the SATA Revision 2.5 specification. For an accurate description of SATA capabilities and the official guideline to SATA product naming, please visit http://sata-io.org/namingguidelines.asp.
Reply to this comment
advertisement
Click Here

The 411 on early-termination fees

Verizon Wireless has doubled its early-termination fees for smartphones, but what does it mean for the rest of the industry?

Google has its own plan for Netbooks

No, the search giant isn't saying it will build a Netbook. But it sure knows what it would like one running Chrome OS to resemble, and that's a little different from the Netbook of today.
• Screenshot tour of Chrome OS

advertisement

About Nanotech - The Circuits Blog

Brooke Crothers has served as an editor at large at CNET News, an editor at Dow Jones' Asian Wall Street Journal Weekly, and a senior editor at InfoWorld. His CNET blog covers chip technology and computer systems, and how they define the computing experience. He also contributes to The New York Times' Bits and Technology sections. He is a member of the CNET Blog Network and is not an employee of CNET. Disclosure.

Add this feed to your online news reader

Nanotech - The Circuits Blog topics

advertisement
advertisement

Inside CNET News

Scroll Left Scroll Right