Intel: LTU Talks Up Next Generation Processors
September 8, 2008
Update: September 8, 2008, 8 12 am Eastern
More about the Intel quad push is at http://www.yourdesktopinnovation.com/
Original Post
Another item about Intel and search. LTU offers an image processing system that law enforcement professionals find useful in certain matters. But LTU’s technology needs processing horsepower. The company had a deal to embed its image classification technology in a consumer video device, but that was slow out of the gates. The reason, according to my sources, was performance.
At the Intel Developer Forum, LTU showed its image processing system running on Intel’s zippy i7 processor. I can’t keep the names straight anymore, but this processor features more cores on die and more cache plus speed ups for computational intensive applications such as image and content processing.
The crowd loved the demonstration, which should make Intel happy. Search vendors need a way to crank up the performance of their systems. Throwing hardware at search bottlenecks may not be the really smart way to solve problems, but it is one that does not require the search vendors to tackle harder problems such as input output and clunky code in their search systems.
I think Endeca will follow in LTU’s foot steps. Intel is poking around the periphery of search, and the company is going to have to take positive action if it wants to do more than sell chips. My hunch is that smart devices with search and content processing functions on board might be an avenue Intel might investigate.
LTU, in case you are not familiar with the company, is French. The company was Founded in 1999 by software and engineering wizards. LTU Technologies provides multimedia content control solutions. Its patented technology is use by the French Gendarmerie Nationale, and the Italian state police; agencies investigating traffic in cultural goods and stolen objects (OCBC of the French National Police); as well as commercial media organizations such as Corbis and Meredith Corporation. You can get more information about the company at http://www.ltutech.com.
Stephen Arnold, September 8, 2008
Open Text Closes on Spicer Slice
July 11, 2008
Open Text acquired privately-held Spicer Corp. You can read Christian Daems’s “Open Text Acquires Division of Spicer Corporation” here.
You may be wondering, “What’s an Open Text?” The company is a player in enterprise search. Among its search properties are an SGML database and search system, the Fulcrum search and retrieval system, BRS Search (a variant of IBM’s original STAIRS mainframe search system), and the Information Dimension’s BASIS data management and search system. In the first edition of Enterprise Search Report, I provided some background information on these systems, and I don’t know if the 4th edition, which I did not write, retained my original baseline on the company.
Open Text was a category leader in collaboration. I recall seeing a demonstration of the system in Washington, DC, many years ago. LiveLink is a content management, collaboration, and search platform. The company hopped on the email search and management bandwagon as soon as news of corporate fraud gained momentum.
What’s a Spicer? According to the Open Text news release, which you can read in its entirety here, tasty part that Open Text bought is the
division that specializes in file format viewer solutions for desktop applications, integrated business process management (BPM) systems, and reprographics.
Spicer provides file viewing software. Instead of launching a third-party application to view a file in a results list, the Spicer technology displays the file without recourse to the native application. Advantages include speed because the native application like Adobe Acrobat is a fat little piggy, chomping memory and time. The other advantage is an opportunity to step away from Stellent’s Outside In viewing technology, which is getting more expensive with each license cycle. Spider also has some security functions that Open Text wants. You can read the full Spicer story here.
This acquisition accompanies Open Text’s purchase of Corbis eMotion. This is an electronic media management tool, primarily used to keep track of images. Could Open Text be contemplating a push into enterprise publishing systems to compete with IBM and Hewlett Packard? If so, Open Text may want to buy Nstein and beef up its tagging capability.
What’s the connection with enterprise search? Not much in my opinion.
Open Text has become a mini-IBM, offering a range of products, services, and features. My thought is that search technology is not delivering the slices of bacon that Open Text’s management and stakeholders want. Furthermore, the competition in email and litigation support is increasing. The core content management system customers are pushing back because CMS is a mess for many customers and vendors. Upstarts like Brainware and ZyLAB are pushing into accounts once viewed as captive to Open Text’s unit managers. The collection of search technologies is difficult to explain, expensive to maintain, and confusing to some of the new Open Text hires whom I have encountered at trade shows this year.
Open Text, after Research in Motion and Coveo, is a darling of the Canadian high-tech sector. The Canadian government doesn’t want another Delphes-like misfire to tarnish the reputation Industry Canada and provincial governments work hard to communicate.
In my opinion, the Open Text buying spree delivers these benefits in my opinion:
- Customers which can be given an opportunity to buy more Open Text products and services
- Media buzz which translates to investor communications
- Filling in gaps in order to make repositioning easier and more credible if the CMS push back becomes more aggressive.
I am probably an addled goose, but I find Open Text’s messaging about search muddled. Click this link to see a search for “search retrieval” on Open Text’s Web site with its own search system. I hope you find the results crystal clear. I don’t.
My working hypothesis is that when companies buy a number of search technologies, I think the cost of explaining each system is high, maybe as expensive as maintaining, supporting, and enhancing the menagerie search systems.
Yahoo fell into this swamp. IBM is in the swamp’ as well. Microsoft has just waded in with a pack of search technologies. If my research is on target, the more search technologies a company collects, the less effective the company becomes in search, content processing, and text processing.
I think companies need to manage the search brands, messaging, and costs; otherwise, cost control becomes very difficult. Even worse, no customer knows what to buy when for which particular search problem. In my own experience, the engineers who have to keep these complex search systems working and in fighting trim are not given the time, resources, and freedom to make Rube Goldberg devices hum like a Toyota computer controlled welding machine. Customers want search to work, and my research suggests for most users search is a source of dissatisfaction.
With enterprise search getting close to a commodity function, some drastic MBA-type positioning is needed right after a finance type with a sharp pencil tallies the cost of search roll ups.
Agree? Disagree? Help me learn.
Stephen Arnold, July 11, 2008
Video Search Bragging Rights: Blinkx Says It Is Bigger Than Google Video
May 16, 2008
For those stuck in northbound traffic on the slow moving river of traffic that is Highway 101, a quite large billboard that told me that Blinkx is the world’s largest video search engine.” In mid-May 2008, a rumor swirled across the Internet that News Corp. was kicking Blinkx’s tires. Was an acquisition in the wind? Was this billboard part of an acquisition campaign? Was it a reminder to Silicon Valley that Google’s span of control did not include video search?
I was sensitive to digitized video for two reasons. The Auto Channel told me that it has thousands of hours of automotive-related video. One interesting aspect of this is that when a video gets “hot”, it gets a great deal of traffic. What’s mystifying, if I understood what The Auto Channel told me, is that it’s very hard to predict what will strike the user’s fancy.
The other reason is that I spoke with a programmer who once did a bit of work for a couple of the large European video services. I can’t reveal the name of the project this person worked on, but it rhymes with “goosed”. The point was that video is flooding the Internet, and it is difficult to generate enough revenue to keep up with the research, development, programming, and bandwidth charges. Video on a metered line is important to many users, but, if I understood his comments, those users don’t pay. Advertisers want “tight” demographics, and the usage data aren’t compelling enough to allow some video sites to generate enough cash to stay alive at this time.
I am not sure how much video Blinkx has indexed. I heard from one of my sources that Google receives more than 1.2 million video uploads per month. I recall reading that the GOOG accounts for more than 60 percent of video search traffic, but since the ComScore traffic flap, it’s tough to know just how much traffic Google has. Could be 70 percent, maybe more. A few days ago, ComScore said Google was the number one Web site on earth. Maybe? Maybe not? Google knows because it does not have to estimate its traffic. My sources tell me that Google just counts traffic, no sampling necessary, to skew the data.
The Blinkx tag line is “Over 26 million hours of video. Search it all.” Their system appears to have a slather of patent documents in place. I tallied more than 100 when I stopped counting. Its conceptual search that includes speech recognition, neural networks, and machine learning to create text transcripts. That text is then searched.
LTU Releases LTU-Finder 3.0
April 28, 2008
One of the leaders in image recognition and analysis is a decade-old company, LTU Technologies. The firm released LTU-Finder v. 3.0, which it described as “a breakthrough tool for image and video recognition in the field of computer forensics”. However, LTU’s system suits a wide range of enterprise image and video applications in eDiscovery, copyright, and security.
Version 3.0 of LTU-Finder includes image and video content recognition technology can increase the speed and scope of forensic and legal investigations as well as e-discovery. The new version includes enhanced image and video recognition capabilities and introduces text data identification tools that further automate large-scale file searches in the legal, e-discovery and law enforcement fields. You can use LTU’s products to find copyright infringement and digital fingerprints of images.
LTU-Finder also incorporates automatic document identification tools that separate relevant scanned documents, like e-faxes, from other content such as personal photos or Web graphics. Automating this process eliminates the need for a subject matter expert to click through image files one by one. The system reduces the amount of data that needs to be processed and stored during the e-discovery process.
You can get more information about the company’s image search and recognition technologies at LTU’s Web site here.
Stephen Arnold, April 29, 2008
-
- Subscribe to Beyond Search
Feature archive
News archive
-
