Trevor Floyd

Scientists at MIT have have used a combination of silicon and gallium nitride, a hard material frequently used in LEDs, to create a hybrid microchip that they say is smaller, faster and more efficient than today's processors. The predicted upgrades have continued since then, though some observers have long predicted that leakage and energy consumption could keep Moore's Law from continuing at some point. Researchers around the world have been working for decades to create such a hybrid microchip that could help chipmakers keep Moore's Law alive . The more than 40-year-old prediction by Gordon Moore holds that the number of transistors on a chip doubles about every two years.

However, if scientists can find new ways to increase efficiency while continuing to make the chips smaller and faster, then the law stands a much better chance of holding true for years ahead. "We won't be able to continue improving silicon by scaling it down for long, so it's crucial to find other approaches," said Tomas Palacios , an assistant professor at MIT, in a statement late last week. Jim McGregor, an analyst at In-Stat, said the new hybrid chip is important because it shows that the industry is moving beyond a singular silicon focus. "We're really in a situation where we're now playing with the entire periodic table and experimenting with different combinations of materials," McGregor said. "The silicon is the basic building block that we put everything on. He added that besides microprocessor chips, the new integrated technology also could lead to more efficient cell phone designs, for instance, by combining the functionality of several different chips onto one. We've been messing around with the silicon and now we're adding something to it ... so we can change the properties and do things with the chip that we couldn't do before." McGregor said he suspects that MIT's hybrid chip design could reduce leakage and thus increase chip performance. "Chipmaking is becoming the ultimate chemistry and physics experiment," McGregor added. "We're using more and more parts of the periodic table and we're down to nanometers and looking at how many electrons can flow from transistor to transistor. Instead, they embedded it into the silicon substrate, which is an underlying layer. It's important for the entire industry, which is focused on this type of research." MIT noted that Palacios, along with student researcher Will Chung, didn't add the gallium nitride as a layer on top of the silicon.

Because the semiconductor industry already uses the same type of silicon substrate, MIT contends that the hybrid chip could be made using today's manufacturing processes, which would be less costly than using different substrates. "We are already discussing with several companies how to commercialize this technology and fabricate more complex circuits," said Palacios, adding that it could take several years before the technology is ready to be commercialized. The nanotubes should someday be used to replace the copper wires that connect the transistors and also may even replace the transistors themselves even further down the road. The move was the latest in a series of recent chipmaking announcements by MIT. Last week, the university announced that researchers there have found a new way to grow carbon nanotubes that could be used by manufacturers to build smaller, faster computer chips. Last year, MIT announced that a research team at the school had created a new chip design that could be 10 times more energy efficient than processors now used in mobile devices. The design is intended for use in portable electronics, like cell phones, PDAs and even implantable medical devices.

Pillar Data Systems Inc. said this week that it's replacing Intel's X25-E solid state disk (SSD) drive as an option for its storage arrays with a drive from STEC Inc. because of firmware problems with the Intel's drive that lead to performance slowdowns. The company also said it is now doubling the available cache on the array to 192GB , shipping boxes with 2TB hard drives - double the capacity of previous drives - and replacing the current dual-core processor controller blade with a quad-core AMD Opteron 2354 chip. The SSD change was one of several upgrades to Pillar's modular storage array, the Axiom. The new hard disk drives push the Axiom's overall usable capacity to more than 1.6 petabytes per system, while cutting overall power consumption by 50% and reducing the space required for the array.

Intel's X25-E, which has had firmware problems in the past, causes operational timeouts, Maness said. The change-over to STEC's SSD also means the company will be able to offer higher capacity 256GB flash drives, since Intel's X25-E tops out at a maximum capacity of 64GB. Bob Maness, Pillar's vice president of worldwide marketing and channel sales, said the company decided to go with STEC's Mach8 SATA SSD because it proved to have better performance in the array. Pillar is continuing to test its arrays with the X25-E and said Intel is working closely with it to solve the issue. Intel did not immediately respond to a request for comment on Pillar's decision. Intel has admitted to firmware problems with the X25-E SSD in the past, but said it resolved them with an upgrade.

Based on commodity hardware, the Pillar Axiom acts as an application-aware storage-area network (SAN) and network-attached storage (NAS) server all controlled by a single management interface. In June, Pillar announced it would begin shipping the Axiom storage system with Intel's X25-E solid state drive as one option. The company's claim to fame is the software used to manage the array by automatically allocating CPU, cache and storage capacity separately to applications as they need more resources. STEC's Mach8 SSD is a consumer-rated drive not normally used in high-performance storage arrays. Maness said Pillar's Axiom achieves the maximum performance capable with the Mach8 SSDs, which use a serial-ATA interface and are less costly than the Zeus SSD. "I'd like to get back to Intel's SSD, but I think STEC has a corner on the market," Maness said. "STEC is just a little further along in terms of their drive." The Pillar Axiom enables users to use any number of drive types, from serial-ATA hard disks to SSD in order to tier storage and utilize the level appropriate to the application being supported.

STEC's Zeus-series SSDs have a Fibre Channel interface normally used for enterprise-class data center operations. Maness said a single tray that houses SSD drives for the Axiom will retail for $49,000, roughly the same cost as six trays of Fibre Channel hard disk drives.

CA Monday announced plans to acquire NetQoS for $200 million, adding application-aware network and systems management products to the software maker's broad enterprise IT management portfolio. The added technology will also boost CA's efforts to manage advanced infrastructures that feature virtual systems and cloud computing environments, the vendor says. "NetQoS technology complements CA's Wily products and will help network and systems engineers better design their infrastructure to ensure application issues don't occur from the start," says Roger Pilc, senior vice president and general manager of CA's infrastructure and automation business unit. "The technologies will help network and systems management be more application aware." The deal, anticipated to close in CA's fiscal third quarter, would augment an already full software lineup grown via previous acquisitions of Wily Technology, Concord Communications and Aprisma. Hottest tech M&A deals of 2009 CA executives say the pending acquisition offers little overlap by way of products and will help CA products diagnose the root cause of application errors within the network and systems infrastructure.

CA executives say NetQoS products, designed for network managers responsible in part for application delivery, will add to the company's Wily products that detect performance problems in the application environment. Customers can visualize the links and relationships between the delivery technologies and the business applications and services with Wily, and understand the real-time application and service activity across those links and relationships with NetQoS traffic flows," says Jasmine Noel, co-founder and principal analyst at Ptak, Noel & Associates. NetQoS tools are able to detect application performance problems using network-centric measures such as traffic flow. "The acquisition is good because NetQoS has a focus on application delivery, so when combined with Wily, it offers a good one-two punch. With some areas of overlap in the former Concord eHealth and Aprisma Spectrum tools, CA's Pilc say the company will work to address issues after the deal closes. NetQoS technology will target network engineers who focus on application delivery where the management of traffic flows is the primary task, rather than the management of thousands of network devices." CA also expects the NetQoS technology to play a bigger role in its virtual and cloud management offerings.

Noel says customers should not expect NetQoS tools to get lost in the shuffle as CA could have targeted plans for each product suite. "In terms of portfolio, CA now has two network performance management solutions, eHealth and NetQoS. But I think CA has specific targets for both solutions," she says. "CA's eHealth technology will target network engineers who spend most of their time managing performance of specialized network infrastructure. With its ability to track flows across virtual and physical elements, NetQoS tools could be coupled with Cassatt assets CA acquired earlier this year, the company says. NetQoS co-founder and CEO Joel Trammel says CA represented the best fit with his company's technology, and customers shouldn't expect any change in products or support as the deal unfolds. With no previous partnerships, the two vendors share some 200 customers and CA's Pilc foresees "very little modification in the NetQoS product set and its approach to customers going forward." That is why NetQoS executives found the deal to be synergistic. NetQoS has more than 1,000 customers worldwide and reported revenue of $56 million in 2008. "We sought out CA because we saw a clear fit with us and the company's success in acquiring Wily, Concord and Aprisma.

Industry watchers expect the deal could benefit both parties going forward if CA sales teams focus on the NetQoS suite. "For a small vendor, being acquired could be good because a larger sales force means a bigger pipeline. We were excited and see the clear fit between tying these acquisitions together," Trammel says. Or it could be bad if it gets lost in the portfolio. Do you Tweet? In the Swainson era, CA has handled its acquisitions fairly well, and with Wily as a tag-team partner I don't see NetQoS getting lost," Noel says.

Follow Denise Dubie on Twitter here.  

A San Francisco Superior Court judge has dismissed three of the four felony charges brought against Terry Childs, a former network administrator who was arrested last year for allegedly sabotaging a crucial city network.

The charges that were thrown out relate to allegations that Childs quietly placed three modems on a San Francisco city network to have backdoor access to the city network. In dismissing the charges, Judge Kevin McCarthy ruled on Friday that there was insufficient evidence to show that Childs had placed the modems on the network with malicious intent.

But the judge left standing a fourth charge, that Childs refused to hand over passwords he had used to lock up the city network for days. If Childs is convicted on that count, he faces up to five years in prison. A hearing is scheduled for today to determine a date for the trial.

A spokeswoman for the San Francisco District Attorney's office today said that prosecutors in the case will appeal Judge McCarthy's ruling. In December 2008, Superior Court Judge Paul Alvarado had held that Childs would need to stand trial on all four of the counts he had been charged with, the spokeswoman said. But, in response to motions filed by Childs' attorney, three of those counts were vacated last Friday, the District Attorney's spokeswoman said.

"We disagree with Judge McCarthy's ruling. We believe that Judge Alvarado's ruling should stand," the spokeswoman said.

Childs, a network administrator working for San Francisco's IT Department of Telecommunications and Information Services (DTIS), was arrested in July 2008 for allegedly locking up access to the city's FiberWAN network by resetting administrative passwords to its switches and routers.

He is also alleged to have planted network devices that enabled illegal remote access to the FiberWAN network, which carries almost 60% of the city government's traffic. He was arrested after first refusing to provide the passwords to city officials and then providing them with wrong passwords. Childs was jailed on a $5 million bond and is currently awaiting trial.

Judge McCarthy's ruling is a significant victory for the defense team, which has been arguing that Childs' actions were far from criminal, and were instead in line with standard network security practices.

Childs' lawyer, Richard Shikman, has argued that the three modems that Childs is supposed to have installed for criminal purposes were instead used for work-related purposes. In court filings the lawyer has maintained that one of the modems was set to dial out on Childs' pager in the event of a network emergency, while another was designed to connect city computers to a disaster recovery site. The third modem had been set up even prior to Childs' arrival at DTIS and was designed to test the citys access to the Internet, his lawyer has said.

In other court filings, Childs had said that he refused to hand over the network passwords to city officials because he feared the passwords would be shared with outside contractors and others who were unauthorized to access the network.

The case has evoked mixed responses so far. Some have argued that Childs may have felt justified in withholding the password information because he was concerned about unauthorized users gaining access to a crucial city network. They have also argued that Childs had been well within the scope of his job description and responsibilities when he installed the modems that prosecutors claimed were meant for malicious purposes.

Others though have highlighted the case as a classic example of the kind of havoc that an insider with privileged access can wreak on a network. They have pointed to the case as an example of why companies need to implement measures for limiting what users with administrative access to critical enterprise networks and systems can do with their access.