- apple-list - University of Manitoba Mailing Lists

MACTEL terms: Centrino
by billing 08 Jul '05

08 Jul '05

While no mention of "Centrino" has been made in Apple's announcements in its movement toward the Intel line of chips, the following link does give some insight into Pentium M vs Celeron M (as well as background on Intel's "Centrino" branding for laptop configurations) and talks about some considerations of the two for laptops. ...from: http://blogs.zdnet.com/BTL/?p=1583&tag=nl.e539 Intel chimes in on ZDNet slam of Centrino brand -Posted by David Berlind @ 3:02 pm, July 8, 2005 [...] [Intel spokesperson Barbara] Grimes wrote: Our research has consistently told us that mobile PC users care most about four things: 1) performance, 2) battery life, 3) thinner & lighter form factors, and 4) built-in wireless capability. We conceived of Intel Centrino mobile technology to deliver on those four requirements. The idea of the [Centrino] brand is to give customers a simple, straightforward way to identify laptops that deliver the four things they care about most in buying a mobile PC. My understanding of your objection is that you feel that the wireless component does not contribute to Centrino's great battery life, so it shouldn't be part of the brand. I see a couple flaws in this thinking. First, the wireless component DOES contribute to the platform battery life (certainly not as significantly as the processor or chipset, but it does nonetheless). But just how much does it contribute? In a separate e-mail, Grimes wrote: In an average laptop configuration, the processor and chipset together make up more than 30% of the platform's power consumption, and the wireless LAN makes up 1%. [...] Another interesting revelation that I think challenges Centrino's value proposition is the existence of Intel's Mobile Celeron processor — the Celeron M. In my ongoing debate with Grimes, I've routinely been reminded that buyers of Centrino-branded notebooks have greater assurance than buyers of non-Centrino notebooks that their systems will successfully get a connection in a Centrino- branded Wi-Fi hotspot. (In other words, pairing a Centrino notebook with a Centrino hotspot promises a better shot at a connection then does compliance with good old fashioned wired and wireless networking standards.) [...] Grimes reminded me: The [Centrino] brand is designed to enable consumers or buyers to easily identify systems that deliver the four capabilities or features that they care about most in a laptop: great performance, great battery life, thinner and lighter designs, and built-in wireless capability. After reminding me that the brand stands for all four attributes in combination, she said the Celeron M systems don't qualify because they don't do as well on battery life and performance as the Pentium M systems do. And Pentium M systems with non-Intel radios don't qualify because they're not as wirelessly capable. What's the bottom line? I agree with aforementioned message as long as there's no official implication that you can't have those four things without a Centrino system. You can. System performance and battery life, by the way, are often dependent on other things that the Centrino "requirement" says nothing about (amount of RAM and size of battery for starters). If only HP would give me some idea of how its new Turion-based notebook fares in performance and battery life against the company's most comparably equipped Intel notebook, we might learn that they're equally capable of all four attributes plus one more: the 64-bit capabilities of the Turion's AMD64 architecture. If only it wasn't for that stupid code of silence. So far, Intel hasn't shipped any mobile processors with the 64-bit capability (beyond its servers, Intel has however finally shipped an AMD64-compliant desktop Celeron). [...]

1 0

INTEREST: 40 million account breach: followup
by billing 07 Jul '05

07 Jul '05

--State Attorneys General Demand Information About CardSystems Security Breach (28 June 2005) Attorneys General from 44 states have written a letter to CardSystems demanding that the payment processor inform people whose accounts were compromised in a massive breach. The letter asks that the company be forthcoming with information regarding the breach. 40 million accounts may have been exposed in the breach; at least 200,000 are known to have been stolen. The letter also said that CardSystem's action was unacceptable. The letter gives the company until July 25 to provide detailed state-by-state information about the breach, remediation efforts, and steps to avoid further compromises. http://news.findlaw.com/scripts/printer_friendly.pl?page=/ap/o/ 51/06-29-2005/d583001a47792379.html

1 0

INTEREST: iTunes store 500,000,000 counter...
by billing 05 Jul '05

05 Jul '05

Apple, for a bit of fun, has a counter that is showing the rate of iTunes purchases on the way to half a billion. It's at the Apple home page: http://www.apple.com

1 0

INTEREST: Duke-iPod update, Class Action results of security breach, Spammers in court, more
by billing 30 Jun '05

30 Jun '05

...reported on "Silicon.com" at http://hardware.silicon.com/storage/ 0,39024649,39131285,00.htm: Free iPods for students: The verdict June 20 2005 by Jo Best The university that gave all its first-year students an iPod has evaluated the success of the pilot scheme - and come up with mixed results. US institution Duke University gave all its 1,650 freshmen the Apple music player, as well as a voice recorder, in autumn 2004 to "encourage creative uses of technology". One year on, the university has found that over 600 students in both the autumn and spring terms made use of the device. Humanities students, particularly those studying music and foreign languages, were the most forthcoming in their use of the iPods, although the whole first class of engineering students had to use the device in a project for their Computational Methods class. Among the faculties that took part in the experiment were Spanish, where students recorded themselves speaking the language for evaluation on their music players, and electrical and computing engineering, where students used the iPods to record pulse rates. Mostly, however, the students used the devices to record or play back lectures and tutorials. Duke's report into the project said: "Initial planning for academic iPod use focused on audio playback; however, digital recording capabilities ultimately generated the highest level of student and faculty interest." The evaluation found that 60 per cent of the undergraduates involved in the scheme used their iPod to record academic material, while 28 per cent used the device for storage. [...] --Class Action Suit Filed Against CardSystems, Visa and MasterCard for Data Security Breach (27 June 2005) A California attorney has filed a class-action lawsuit against CardSystems Solutions, Visa and MasterCard on behalf of California residents and merchants who accept credit cards. The lawsuit alleges the companies did not appropriately secure their systems and did not inform people in a timely manner about a security breach that exposed data on 40 million accounts. Apparently information on approximately 200,000 accounts was verifiably transferred out the payment processor's computer system, but the credit card companies said they do not intend to notify people of the problem unless the accounts are used fraudulently. The lawsuit asks that the companies notify all people whose data were exposed and provide special notification to the 200,000 whose data were stolen. The suit also asks that chargeback fees and penalties to merchants be waived on transactions where the information was used fraudulently. http://news.com.com/2102-7350_3-5765383.html?tag=st.util.print http://www.techfirm.com/cardsystems.pdf --Alleged Spammer First to be Tried Under Australia's Spam Law (24/23 June 2005) The Australian Communications Authority will try Wayne Mansfield, managing director of Clarity 1, for allegedly sending at least 56 million spam messages since Australia's Spam Act took effect in April 2004. The ACA is also seeking an interim injunction against Clarity 1 to prevent it from sending out more unsolicited email until the court hearing, which is scheduled for July 20. Mansfield maintains he has not broken the law and is looking forward to proving his innocence in court. http://www.theregister.co.uk/2005/06/24/ australia_sues_alleged_spammer/print.html http://internet.aca.gov.au/ACAINTER.65636%3ASTANDARD%3A%3Apc=PC_3000 http://www.techworld.com/security/news/index.cfm?NewsID=3912 http://australianit.news.com.au/articles/0,7204,15714808%5E15331%5E% 5Enbv%5E15306%2D15318,00.html --Microsoft Suing German Company for Sending Spam (22 June 2005) Microsoft is suing an unnamed German company for allegedly inundating Hotmail users with spam. The German company allegedly used organizations in the US and the Ukraine to send the unsolicited commercial email. The company's director denies any personal wrongdoing; he says his partners are responsible for the spam and that they are "out of control." Because Germany does not presently have any laws against distributing spam, Microsoft is seeking an injunction in an attempt to shut down the North Rhine Westphalia-based company under German fair trade law. http://www.scmagazine.com/news/index.cfm? fuseaction=newsDetails&newsUID=71539896-5ee5-46b5-b9fe- b72d83d68d83&newsType=News http://www.computerworld.com/printthis/2005/0,4814,102670,00.html --Indian Police Need Formal Complaint Before They Can Arrest Alleged Data Thief (24/23 June 2005) A man who reportedly sold data on 1,000 bank accounts of UK citizens to an undercover reporter in India is still at large. The information was reportedly stolen from call centers. Delhi police cannot arrest the man, Karan Bahree, until they receive a formal complaint either from the call centers where the data were stolen or from banks or individuals in the UK affected by the theft. In a BBC interview, Mr. Bahree said he gave a CD to the undercover reporter at the behest of another person and did not know what the disk contained. UK organizations that outsource customer services are liable for security problems that arise; the banks affected in this case could be found to be in violation of the UK's Data Protection Act. http://www.computerworld.com/printthis/2005/0,4814,102755,00.html http://www.msnbc.msn.com/id/8345423/ http://www.silicon.com/research/specialreports/offshoring/print.htm? TYPE=story&AT=39131387-3800003026t-39000857cs

1 0

INTEREST: the "tech geek" view of G5 (64 bit, Velocity Engine) & P4 (EM64T, SSE2)
by billing 22 Jun '05

22 Jun '05

Let me say right off that if gory innards, techgeek terms like SPECfp and SPECint, and looking at charts showing how fast a Photoshop tranform happens are not of interest to you, stop reading now. ::-) For those of you brave enough to venture forward, here's one attempt at comparing what difference may exist between the current Apple Macintosh hardware heart and the upcoming Apple Macintosh hardware heart (from that techie/geeky viewpoint ::-). Mostly it'll be just a bunch of existing numbers and graphs but I've also tried to include some of the sources from the manufacturers that explain the numbers and graphs in words as well..... so that you have the numbers.... in.... words...... as well..... (Disclaimers: Due to the nature of the tables, it'll be easier to read if you change to a monospaced font. My apologies for any errors found - please let me know. The following raw numbers do not define the entire picture. How everything gets put together can have a dramatic effect; the work done during a G4 or G5 clock cycle differs from that done during a P4 cycle, etc. Nor does this comment on OS concerns. Source links for this information are given at the end of the message. All Intel CPUs listed below, except for the Celeron, Celeron-M, and Pentium-M can support 64 bit addressing. ***Intel's SSE2 is not available in the Celeron processor.) Very rough guidelines based upon passed usage: Desktops: G4, G5, PXE, PD, P4, Celeron Laptops: G4, Pentium-M (couples the execution core of the Pentium III with a Pentium 4 compatible bus interface - code name "Yonah"), Celeron-M Servers: G5, Itanium 2, Xeon *FS Bus: Front side bus Processor Name/Number | Clock Speed | L2 cache | FS Bus* | Dual Core | | (GHz) | (KByte) | (GHz) | | ------------------------------------------------------------------------ -- IBM/Motorola G4 | 1.25, 1.42, | 512 | .333 | | | 1.33, 1.5, | | | | | 1.67 | | | | ------------------------------------------------------------------------ -- IBM/Motorola G5 | 1.8, 2.0, | 512 |.600, 1, | | | 2.3, 2.7 | | 1.15,1.35| | ------------------------------------------------------------------------ -- Intel PentiumXE 840 | 3.2 | 2x1000 | .8 | yes | ------------------------------------------------------------------------ -- Intel Pentium D 840, 830, | 2.8 - 3.2 | 2x1000 | .8 | yes | 820 | | | | | ------------------------------------------------------------------------ -- Intel P4 571, 570J°, 561, | 2.8 - 3.8 | 1000 | .8 | | 560J, 560, 551, 550J, 550,| | | | | 541, 540J, 540, 531, 530J,| | | | | 530, 521, 520J, 520 | | | | | ------------------------------------------------------------------------ -- Intel P4 670, 660, 650, | 3.0 - 3.8 | 1000 | .8 | | 640, 630 | | | | | ------------------------------------------------------------------------ -- Intel Celeron D 346, 341, | 2.5 - 3.06 | 256 | .533 | | 336, 331, 326 | | | | | ------------------------------------------------------------------------ -- Intel Celeron*** | .95 - 2.8 | 128, 256 | up to . 4| | ------------------------------------------------------------------------ -- Intel Pentium M 770, 765, | 1.5 - 2.13 |1000 - 2000| .4, .533| Yonah | 760, 758, 755, 753, 750, |Low voltage: | | | coming | 745, 740, 738, 735, 733, | 1.3-1.5 | | | 2006? | 730, 725, 723, 718, 715, |Ultra low | | | | 713 |voltage:1-1.2| | | | ------------------------------------------------------------------------ -- Intel Celeron M 383, 373, | 1.2 - 1.5 |512 - 1000 | .4 | | 370, 360, 353, 350, 340, | Ultra low | | | | 333, 330, 320, 310 | volt:.9-1.0 | | | | ------------------------------------------------------------------------ -- Intel Itanium 2 | 1.0 - 1.6 |256 (1500- | bandwidth| | | | 9000 L3) |6.4GB/ sec| | ------------------------------------------------------------------------ -- Intel Xeon | 2.8 - 3.66 |1000 - 2000| bandwidth| | | | |6.4-14GB/ s| | ------------------------------------------------------------------------ -- Floating Point and Integer Arithmetic: -------------------------------------- Apple/IBM/Motorola AltiVec (Velocity Engine): 128-bit vector registers that can represent sixteen 8-bit signed or unsigned chars, eight 16-bit signed or unsigned shorts, four 32-bit ints or four 32- bit floating point variables. supports a special RGB "pixel" data type Intel SSE2: replacement for MMX and SSE. 128-bit vector registers that can represent sixteen 8-bit signed or unsigned chars, eight 16- bit signed or unsigned shorts, four 32-bit ints or four 32-bit floating point variables. Performance: ------------ SPECfp 2000 (Bigger is better): - Dual 3.66 GHz Xeon EM64T: 28.6 (IBM xSeries 336: 2M L2 cache, 8 GHz FS Bus) - Dual 2GHz G5: 15.7 - Dual 3.06 GHz Xeon: 11.1 - Single 3 GHz P4: 8.07 SPECint 2000 (Bigger is better): - Dual 2GHz G5: 17.2 - Dual 3.06 GHz Xeon: 16.7 - Single 3 GHz P4: 10.3 Adobe Photoshop 7.01 600 MB file: 45-filter function, 2 GB memory (performance of a single P4 3GHz system is "1") - Dual 2GHz G5: 2.2 - Single 1.8 GHz G5: 1.8 - Single 1.5 GHz G5: 1.5 - Dual 3.06 GHz Xeon: 1.1 BLAST DNA sequence matching (in millions of nucleotides per second - higher is better): WORD LENGTH 10 20 30 40 ----------------------------------------------------- Dual 2 GHz G5 | .2 2 2.9 3.1 Single 1.8 GHz G5 | .2 1.5 2.5 2.5 Single 1.6 GHz G5 | .2 1.2 2.1 2.3 Dual 3.06 GHz Xeon| .2 0.5 0.6 0.6 Single 3 GHz P4 | .2 0.5 0.5 0.4 HMMer genome matching sequencing (performance of a single P4 3GHz system is "1") - Dual 2GHz G5: 6.9 - Single 1.8 GHz G5: 4.8 - Single 1.5 GHz G5: 4.2 - Dual 3.06 GHz Xeon: 1.7 Links: ------ G4/G5: http://www.apple.com/ibook/specs.html http://www.apple.com/powerbook/specs.html http://www.apple.com/macmini/specs.html http://www.apple.com/imac/specs.html http://www.apple.com/powermac/specs.html http://www.apple.com/xserve/specs.html http://www.apple.com/lae/powermac/performance/ Intel: http://www.intel.com/products/processor/pentiumXE/index.htm http://www.intel.com/products/processor/pentium_D/index.htm http://www.intel.com/products/processor/pentium4/index.htm http://www.intel.com/products/processor/celeron_D/index.htm http://www.intel.com/products/processor/celeron/index.htm http://www.intel.com/products/processor/pentiumm/index.htm http://developer.intel.com/design/Pentium4/prodbref/ http://www.cpuid.org/PentiumM/index.php http://techreport.com/reviews/2003q3/pentiumm-1.4ghz/index.x?pg=1 http://en.wikipedia.org/wiki/Pentium_M http://www.amd.com/us-en/Processors/ProductInformation/ 0,,30_118_11600_11613,00.html AltiVec: http://www.absoluteastronomy.com/encyclopedia/A/Al/AltiVec1.htm http://en.wikipedia.org/wiki/AltiVec http://www.tommesani.com/AltiVec.html SSE2: http://en.wikipedia.org/wiki/SSE2 http://www.tommesani.com/SSE2Intro.html http://www.absoluteastronomy.com/encyclopedia/S/SS/SSE22.htm EM64T (Intel's 64 bit addressing technology): http://www.answers.com/topic/em64t http://www.best.de/download/marketing/AMD-Sun-Opteron_I.pd Wayne

1 0

Re: Spyware Floods In Through BitTorrent
by Will Christie 20 Jun '05

20 Jun '05

Just to clarify the article, the title is a bit misleading, and really should read "Spyware Floods In Through Any-Program-You-Use-To-Connect-To-The-Internet-With." It really underscores that you shouldn't be downloading or executing anything off the Internet unless you know where it came from (or are prepared to deal with the consequences). BitTorrent works by creating a torrent file that describes how a file(s) is constructed. Users can download the torrent file, and when run with a torrent program, it will contact other torrent programs on the Internet, looking for the pieces. The torrent program will also share existing pieces that you've collected, but only for the particular torrent file (it does NOT share out your hard drive). Torrents work well in distributing information quickly as multiple sources can give you several pieces simultaneously. You don't need to obtain the file from a single source one piece at a time if you were downloading it via HTTP or FTP. In this case Direct Revenue is creating torrents of files that users want that include their spyware program IN ADDITION TO the original file. They are not co-opting the original file or infecting torrents that are not their own. In fact there is some legality on their repackaging and distributing these torrents. If you download a torrent distribution file for a Linux boot disk from Knoppix.org then there is no way for Direct Revenue to add their spyware into that torrent. If you download a torrent distribution file for the same Linux boot disk from Direct Revenue then there is NO guarantee that you won't have an unwelcome hitch-hiker added to your download. Even then, once their spyware is downloaded to your machine you will need to execute it for the spyware to sink its hooks into your system. Frankly, anyone who obtains (for example) boot disk.ISO and an executable file they don't why it's there, and decides to run the executable, deserves to be thoroughly chastised with the same tone of voice reserved for people who cross the street without looking both ways. Slashdot has a good discussion about this here: http://yro.slashdot.org/article.pl?sid=05/06/16/1558229&tid=158&tid=172&tid… -- Will_Christie(a)UManitoba.CA "You probably think Canada's leader Instructional Software Coordinator is called the, uh... Grand Poobah!" Academic Computing & Networking "No, it's the Governor-General." University of Manitoba "WRONG! It's the Prime Minister." 622 Engineering "Technically Hank, it is Ph: 204-474-9475 Fx: 204-474-7920 the Governor-General." --Corner Gas

2 1

INTEREST: Spyware Floods In Through BitTorrent
by billing 20 Jun '05

20 Jun '05

...from: http://www.eweek.com/article2/0,1759,1828633,00.asp Spyware Floods In Through BitTorrent By Ryan Naraine June 15, 2005 BitTorrent, the beloved file-sharing client and protocol that provides a way around bandwidth bottlenecks, has become the newest distribution vehicle for adware/spyware bundles. [...] "This is the marketing campaign to end all marketing campaigns," said Boyd, the Microsoft Security MVP (most valuable professional) known throughout the security industry by the "Paperghost" moniker. In an e-mail interview with Ziff Davis Internet News, Boyd said rogue files have popped up occasionally in BitTorrent land but those were usually just random executables. "This is the first time I've seen a definite money-making campaign with affiliates, distributors and some pretty heavy-duty adware names," he added. Boyd, widely known for chronicling spyware, hacking and malware exploits, has published details of the BitTorrent distributions and identified Direct Revenue and Marketing Metrix Group as the companies responsible for the rigged files. Boyd said he got the first inkling that BitTorrent was a major adware distribution vehicle while searching for the source of Direct Revenue's Aurora, an adware program that includes the prevalent "nail.exe" component. Sifting through mountains of HijackThis logs posted on security forums, Boyd said the answer was staring him in the face. (HijackThis is a popular freeware spyware removal tool that keeps detailed logs of Windows PC scans). In the logs, he found that "nail.exe" and "aurora.exe" were always listed alongside "btdownloadgui.exe," the user interface that downloads/uploads when using BitTorrent. "I checked hundreds of those logs, and more often than not, [btdownloadgui.exe] was chugging away in the background. No wonder none of the victims (or spyware experts) seemed to know what site Aurora was coming from—there was no site. It would have never occurred to the end users that it could have crept in by another means altogether," he said. Because BitTorrent strips digital files into tiny shreds and reassembles them locally once a user completes a download, it has emerged as the perfect place to bundle adware programs among the bits, without the end user ever knowing. A BitTorrent user downloading a movie clip only becomes aware of the associated adware after the files are reassembled. At that stage, when the user attempts to load the reassembled file, he or she is greeted by an installation notice for an adware bundle distributed by MMG (Marketing Metrix Group), a Canadian company that specializes in P2P network marketing. [...] Direct Revenue admitted to using MMG to push Aurora distributions via BitTorrent, but insisted that the actual adware installation was done with adequate and up-front disclosure. In an interview, Direct Revenue chief technology officer Daniel Doman said MMG is "one of many affiliates" used to distribute Aurora. "They [MMG] specialize in doing content distribution on peer-to-peer channels, and we think they provide an easy mechanism for people like us who want to monetize software or content." Doman, a former director of engineering at DoubleClick Inc., said the increased visibility of Aurora and the "nail.exe" component was not the result of new installations, pointing out that Direct Revenue is auto-updating its file-naming convention to address criticisms that the adware program was hidden on purpose. "We just recently launched a full awareness and campaign to the entire user base, and the fact that those files are showing up in logs shows that we're having success," he said. The campaign, announced on May 17, includes the placement of an uninstall facility within the add/remove panel on Windows for PCs that points users to the previously hard-to-find MyPCTuneUp Web site for adware program removal. "We've taken pains to brand all of our windows so that the source and prominence of the advertising we serve are extremely clear," Doman said. So far, about 90 percent of Direct Revenue's user base has received the branding updates, which happen without any user action. Direct Revenue has been heavily criticized for forcing users to visit the MyPCTuneUp site to complete the program removal, but Doman defended that strategy, insisting the Web-based uninstall utility is the most efficient way to make sure the removal is properly done. He said the company was seeing increased traffic to the site since the launch of the campaign, adding that the daily uninstall count was "in the thousands." Even so, he said, the thousands of daily uninstalls represent only a fraction of a percentage of the entire user count and are not materially affecting Direct Revenue's business. Doman described Boyd's posts on VitalSecurity.org as "misleading" and pointed out that the screenshots provided by the researcher "clearly show full disclosure" before the Aurora program is installed. He acknowledged that a "grey area" exists in the timing of the disclosure, but insisted that it was done in full compliance with existing laws. "We require all our distributors to fully inform end users about what is being installed. It's a clear opt-in procedure," he said. "The user is downloading something through BitTorrent that is ad- supported and [Boyd's screenshot] shows the disclosure that is provided. The idea that somehow the download is surreptitious is wrong. It's very apparent that if the BitTorrent user goes through with the MMG download, they agree to install the ad-supported software." Doman added: "The notion that the user has accidentally found all this software on his machine is false. [MMG] is using a 'pull' technology. Nothing is being snuck in the back door."

1 0

INTEREST: will Apple use Intel's 100Mb Centrino laptop technology in 2006?
by billing 20 Jun '05

20 Jun '05

...from: http://www.extremetech.com/article2/0,1558,1829128,00.asp Intel Research Merges Centrino With 802.11n By Mark Hachman June 17, 2005 Intel will present a paper Friday that will demonstrate its progress toward integrating the current Wi-Fi technologies with the next- generation 802.11n protocol. At the 2005 VLSI Symposium on Circuits in Kyoto, Japan, Intel executives will present two papers, one on integrating a 2.4/5-GHz WLAN as well as a second on a 90-nm filter chain. Intel speakers will show off photographs of a prototype chip. While the papers don't describe a landmark breakthrough, they do provide proof that Intel can take an incremental next step and integrate the 802.11n component into its existing Centrino chipset, according to Manny Vara, a technology strategist with Intel. [...] ...from: http://www.reed-electronics.com/electronicnews/article/CA445702 802.11n: The Next WLAN Frontier By Suzanne Deffree -- Electronic News, 8/19/2004 The wireless industry is raring to go with its next connectivity technology, 802.11n. But before the 100Mbit/sec. minimum throughput wireless LAN technology can see a standard, things could get ugly. The successor to 802.11a/g aimed at consumer applications that require very high throughput, like HDTV and streaming video, 802.11n passed the proposal posting phase at the IEEE last week with an expected 60-plus groups and individual companies laying out what they hope will win the wireless organization's eventual approval. The odds are heavily laid in the favor of two equally matched groups, TGn Sync and WWiSE -- each of which recently sent out news releases cheering themselves on and purposefully pitched their plans to the media. TGn Sync, which stands for task group n synchronization, holds Intel, Agere, Atheros, Sony and Philips among its ranks. While WWiSE, standing for worldwide spectrum efficiency, includes such companies as Broadcom, Conexant, Texas Instruments, Airgo and STMicroelectronics. What both consortiums agree on is the use of MIMO, multiple input multiple output antenna technology, to take 802.11n to its promised throughput. Where they mostly differ is on channel bands. TGn looks to use 40MHz channels in the 5GHz spectrum, the same one used by 802.11a, while WWiSE prefers 20MHz channels in the 2.4GHz consistently used 802.11b/g spectrum, explains Phil Solis, a senior analyst at ABI Research. "It's not like one side is heavy against the other. The biggest difference is that the WWiSE group wants to stay with 20MHz bands so they don't tie up too many channels," he said. "Both groups will use some form of MIMO technology, it's just that the WWiSE group wants to be more conservative with the spectrum and use other methods to bring the speeds up. The concern of WWiSE in using a 20MHz band is a reasonable one, if you are going to be in the 2.4MHz space where there are only three nonoverlapping channels. But in the 5MHz band in the United States, there are 24 channels and none overlap with b and g." Agere, for one, argues that use of TGn's channel plan could increase 802.11n's minimum 100Mbits actual throughput speed fivefold. The company's personal end target: 500Mbits/sec. "If I widen the channel to 40MHz, I can get 125Mbits of data through, just with the wider channel. If I add MIMO to that, another transmitter, then I go to 250Mbits," remarked Mary Cramer, strategic marketing manager and Agere's business representative for the IEEE task force n group. "I can realistically go up to four transmitters and get 500Mbits. In this proposal we are asking for a mandatory two transmitters -- because it's more realistic for power and cost -- and to be capable of 40MHz channels, and that would make up 250Mbits for an 802.11n device." Conexant's rebuttal on behalf of WWiSE is that it's not just about speed, but also about future spectrum availability and backward Wi-Fi compatibility to 802.11b/g. "20MHz channel width is consistent with the frequency plan used in all our legacy 802.11 equipment sold to date," Jim Zyren, executive director, wireless and residential gateway access products at Conexant, said, adding that the use of 20MHz channels helps conserve unlicensed spectrum. "We've all heard anecdotal evidence of crowded spectrum and interference in the 2.4GHz unlicensed domain. As a group we feel we have just scratched the surface in terms of global Wi-Fi adoption. Now is the time to begin implementing spectrum conservation measures. By making 20MHz channels a mandatory element of our proposal, we'll ensure maximum utilization of the existing channels before resorting to the use of wider channels." The Curse of Ultra Wideband Beyond MIMO and channel use, both groups agree that no one wants to end up like ultra wideband technology. UWB, which is still awaiting final ratification, has been tossed and twirled about IEEE for more than a year now as the proposal parties can't seem to agree on common ground. "One of the other main group's technology is some what similar to ours and in some ways it isn't," Agere's Cramer said, alluding to WWiSE and its use of MIMO. "I don't think this is going to be like UWB and the situation that has happened there, where the whole standard is just locked up because people can't agree on it." ABI Research noted in a recent report that had UWB come to market quicker its superior speeds to wireless LAN may have made it more of a contender for applications in the home. Having 802.11b/g already staked out in the home gives it a significant advantage over UWB, Solis said. "Wi-Fi will continue to work its way into home entertainment networking, and will become entrenched," the analyst said in the report. "By the time UWB comes out -- or just a little later -- 802.11n solutions will start to appear." Both sides are hoping that UWB's IEEE curse will be avoided and compromises will be made … just not anytime soon. With the next meeting to discuss the proposals set for September, TGn and WWiSE aren't expecting a first draft specification to come down the pike until at mid-to-late 2005 and are eyeing a final spec in 2006. That hasn't stopped anyone from starting to plan out their products, though, with proper marketing. While both Agere and Atheros have released consumer-oriented HDTV and video products based on ultra- fast data transmission using principles of their group's proposal, they have been careful not to promote them as "pre-n" technology. "I think anyone who claims their product is pre-n, given that first proposals haven't been made, would be really stepping out wide," Atheros Product Line Manger Sheung Li said. "What makes pre-n product so premature now is that if you launch such a thing, then you can only talk to such products from the same manufacturer and you don't get any performance gains when talking to a, b or g equipment." But product appearance on shelves before ratification has a strong history in WLAN. 802.11n's predecessor 802.11g went down that path, with technology out about six months before the IEEE had give its green light to the 54Mbit specification. "The market's perception has been shaped by the experience that we went through collectively in the 802.11g space. The take-away everyone got from 802.11g was that we are going to see products in the market before the standard is finally ratified," Zyren said. ...other links of interest: - updates following the 802.11n standard: http://wifinetnews.com/ archives/cat_80211n.html - 2MHz-3G cellphone-UWB crosstalk: http://www.ultrawidebandplanet.com/ technology/article.php/2173501 - http://www.ofcom.org.uk/media/mofaq/rcomms/uwbfaqs/#content - Fat pipes from thin air?: http://www.networkmagazine.com/shared/ article/showArticle.jhtml?articleId=8703372&classroom= - Find you way around the RF spectrum: http://www.techworld.com/ features/index.cfm?featureID=212&printerfriendly=1 - The A to Z of wireless terms: http://www.techworld.com/features/ index.cfm?featureID=212&printerfriendly=1

1 0

SciTech WEBCASTs: June 16th and 23rd (fwd)
by Doug Hamilton 15 Jun '05

15 Jun '05

---------- Forwarded message ---------- Date: Wed, 15 Jun 2005 09:48:16 -0400 Subject: SciTech WEBCASTs: June 16th and 23rd =================================================================== There are two exciting SciTech webcasts coming up, the latter with Canadian content... - Tiger/Tiger Server for HPC - Storage for Research Computing: includes Dr. Jackie Vogel from McGill talking about her Xsan implementation. For more details, visit www.apple.com/science/webcasts.html =================================================================== Mac OS X and Mac OS X Server for High Performance Computing Thursday, June 16, 2005 10 a.m. PT, 75 minutes Presenters: Matt Macinnis Market Manager, Science and Technology Markets Apple Computer Ernest Prabhakar, Ph.D. Product Manager, UNIX and Xgrid Apple Computer Mac OS X v10.4 �Tiger� and Mac OS X Server v10.4 �Tiger� provide the ultimate platform for high performance cluster and grid computing. As true 64-bit operating systems, Tiger and Tiger Server allow a single process to directly address up to 16 exabytes of virtual memory. Both client and server versions of the operating system include a distributed resource management service, called Xgrid, that allows you to rapidly assemble grids or clusters without complex configuration steps. And with a refined kernel, an improved UNIX environment and updated Linux compatibility, the latest versions of Mac OS X and Mac OS X Server are unrivaled for stability, performance, compatibility and ease-of-use in cluster environments. In this webcast, you�ll learn about the technical underpinnings of Apple�s latest operating system and how new features and updates will benefit users of Mac OS X-based clusters and grids. More details: http://seminars.apple.com/seminarsonline/tigerforhpc/ apple/ =================================================================== Cost-Effective Storage Deployments for Research Computing Thursday, June 23, 2005 10 a.m. PT, 75 minutes Guest speakers: Alex Grossman Senior Director, Server and Storage Hardware Apple Computer Jackie Vogel, Ph.D. Assistant Professor of Biology and CIHR New Investigator McGill University M. Michael Barmada, Ph.D. Associate Professor of Human Genetics University of Pittsburgh Ryan Klein SAN Architect QLogic Corporation In today�s research environment, IT organizations are faced with finding cost-effective, scalable and flexible storage solutions to accommodate the relentless growth and complexity of scientific data. At the same time, scientists want faster and better access to their data as well as ways to collaborate easily with peers across academia, industry and government. In this webcast, you�ll learn how to achieve better data accessibility and faster storage retrieval in a more cost-effective way with industry-leading solutions from Apple. You�ll also hear examples of successful storage deployments in research environments and why Mac technology was chosen. More details: http://seminars.apple.com/seminarsonline/ researchstorage/apple/ ###

1 0

INTEREST: What chips might appear in Macs next year?
by billing 14 Jun '05

14 Jun '05

...from: http://www.macworld.com/2005/06/features/intelchips/index.php/? lsrc=mwweek-0613 June 13, 2005 Macworld Inside Intel’s offerings What chips might appear in Macs next year? By Jon L. Jacobi [It will be a year before Intel-built chips start appearing in Apple hardware. But it's not too early to start getting familiar with the processors currently available from Intel.] When it comes to sipping daintily from the AC and running cooler, Intel chips are unmatched by anything IBM has to offer. And that’s a handy characteristic to have for a CPU that’s wedged within the narrow confines of a notebook. But before we get into the nitty-gritty of what the Santa Clara, Calif.-based chip giant’s products will do for you, what they won’t offer is a giant leap in performance on the desktop. Apple’s current G5 is a powerful chip—despite running at slower clock speeds than the processors powering Wintel machines. Slower clocked x86-compatible Athlon 64s from über Intel-rival AMD’s regularly benchmark as fast or faster than Intel’s offerings. (Notebook performance is another matter as you’ll gather when you read about the Pentium M below.) The Wintel world is cluttered with far more products than the Mac universe and Intel offers a bewilderingly large array of CPUs that vary in cache size, speed, and features. A lot can change in a year, but with a recent flurry of new technology releases from Intel, the first MacIntels should feature one of the following processors—or something very close to them. Pentium 4 Designed for: Desktops and desktop replacement laptops The best known of Intel’s processors, the P4 was designed to achieve high clock speeds, but it generally does less per clock cycle than competing CPUs. The latest P4s run as fast as 3.8GHz with 1MB or 2MB of cache and an 800MHz front-side bus. They’re also 64-bit courtesy of Intel’s EM64T—a version of the x86-64 64-bit instructions developed by AMD for its Intel-compatible Athlon 64 CPUs. EM64T P4’s perform both 32-bit and 64-bit instructions with equal facility, and, if Apple systems employing them are anything like the PC-side of things, they should support at least 64GB of directly addressable memory. Some P4 models also feature Hyper-Threading, which creates two virtual CPUs to improve performance when running simultaneous tasks— like applying a Photoshop filter while you’re browsing the Web. Pentium D Designed for: Desktops and performance laptops The Pentium D evolved out of the P4 and, like the G5 and most Intel CPUs, is manufactured using a state-of-the-art 90-nanometer process. The “D” stands for Dual-core which means two execution cores—the heart of a CPU that actually processes the instructions—on the same chip. It’s essentially the same concept as a dual-processor G5, except that the CPUs and support logic are found on a single piece of silicon. With two cores, Pentium D CPUs are especially adept at multi- tasking and all three models (2.8GHz, 3.0GHz, and 3.2GHz) feature the EM64T instruction set, 1MB of cache per core, and an 800MHz front- side bus. Pentium 4 Extreme Edition Designed for: Gaming and performance desktops The newest version of the P4 EE is actually a Pentium D with Hyper- Threading enabled, meaning two physical processor cores split via Hyper-Threading for a total of four virtual CPUs. This latest flagship EE shouldn’t be confused with older EE models, which are single-core P4’s with HT and either a 2MB cache (3.73GHz, 1066MHz front-side bus, 90-nanometer model) or dual 512KB L2 and 2MB L3 caches (3.2Ghz/3.46GHz, 800MHz/1066MHz front-side bus, 130-nanometer models). Pentium M Designed for: Laptops Some think that the 32-bit Pentium M is what Apple was really after from Intel in the short term. The Pentium M isn’t based on the P4 core; it’s a separate animal that does a lot more work per clock cycle. Available in clock speeds from 1.5- to 2.13GHz, it easily outperforms the G4 found in Apple’s fastest notebooks and offers wonderfully parsimonious power consumption. Pentium M notebooks regularly last more than five hours on a single battery charge— something iBook and Powerbook users can currently only dream of. Celeron Designed for: budget desktops, laptops Celerons are the lower-priced, 32-bit-only cousins of the Pentium that generally suffer a slower front-side bus, less cache or some other performance-inhibiting characteristic. They cost only a little less than low-end Pentium models but you never know what might make an appearance in a budget MacIntel.

1 0