| ActiveMac | Anonymous | Create a User | Reviews | News | Forums | Advertise |
|
|
|
User Controls |
|
New User |
|
Login |
|
Edit/View My Profile |
|
|
|
|
|
|
|
ActiveMac |
|
Articles |
|
Forums |
|
Links |
|
News |
|
News Search |
|
Reviews |
|
|
|
|
|
|
|
News Centers |
|
Windows/Microsoft |
|
DVD |
|
ActiveHardware |
|
Xbox |
|
MaINTosh |
|
News Search |
|
|
|
|
|
|
|
ANet Chats |
|
The Lobby |
|
Special Events Room |
|
Developer's Lounge |
|
XBox Chat |
|
|
|
|
|
|
|
FAQ's |
|
Windows 98/98 SE |
|
Windows 2000 |
|
Windows Me |
|
Windows "Whistler" XP |
|
Windows CE |
|
Internet Explorer 6 |
|
Internet Explorer 5 |
|
Xbox |
|
DirectX |
|
DVD's |
|
|
|
|
|
|
|
TopTechTips |
|
Registry Tips |
|
Windows 95/98 |
|
Windows 2000 |
|
Internet Explorer 4 |
|
Internet Explorer 5 |
|
Windows NT Tips |
|
Program Tips |
|
Easter Eggs |
|
Hardware |
|
DVD |
|
|
|
|
|
|
|
Latest Reviews |
|
Applications |
|
Microsoft Windows XP Professional |
|
Norton SystemWorks 2002 |
|
|
|
Hardware |
|
Intel Personal Audio Player 3000 |
|
Microsoft Wireless IntelliMouse Explorer |
|
|
|
|
|
|
|
Site News/Info |
|
About This Site |
|
Affiliates |
|
ANet Forums |
|
Contact Us |
|
Default Home Page |
|
Link To Us |
|
Links |
|
Member Pages |
|
Site Search |
|
Awards |
|
|
|
|
|
|
|
Credits |
 |
Big Mac world's 3rd fastest computer | |
|
||
| Time: 12:07 EST/17:07 GMT | News Source: Macworld UK | Posted By: Byron Hinson | ||
|
The Big Mac supercomputer – which combines 1,100 dual-2GHz Power Mac G5s – is officially ranked the world's third fastest computer. The Top 500 ranking ranks machines on the maximum number of floating operations per second (flops) achieved during the test. Virginia Polytechnic Institute and State University's Big Mac super computer came third, with a 10.3 teraflops. Each of the clsuter's G5s is connected using Infiniband interconnect technology and Gigabit Ethernet switches. | ||
|
Write Comment
Return to News |
Displaying 1 through 25 of 310 Last | Next |
The time now is
2:23:13 PM
ET.
Any comment problems? E-mail us |
| #1 By 3339 (64.160.58.135) at 11/18/2003 12:03:41 PM |
|
parker, why don't you look for the answer before spouting off your idiocy? Go to the VT site. They have developed memeory checking software that is more efficient than ECC RAM. Algorithms and programs do not need to be run twice, never mind three times. And error-checking is not as simple a process as going: one result, two results, do they match, three results, what does this result match, you fool. The full efficiency that is being reported in the top500 includes error checking since this software is running throughout the benchmarks. This post was edited by sodajerk on Tuesday, November 18, 2003 at 15:13. |
| #2 By 3339 (64.160.58.135) at 11/18/2003 8:32:37 PM |
|
You don't understand memory correction and scientific computing do you, parker? A flipped bit in a complex scientific calculation can produce an incorrect result. A flipped bit in a memory correction algorithm isn't going to make the memroy checking algorithm wrong, fool.
And, besides, did you just walk away from the idea that it's actually 2x or 3x slower? I do believe you did. I'm getting tired of trying to keep you informed. Can't you learn how to use the web? |
| #3 By 3339 (64.160.58.135) at 11/19/2003 1:07:42 PM |
|
Just because you can't find the information doesn't point to it not being available... maybe it points to your literacy problem, but the fact remains that the Terascale Cluster is not being used to render the next Pixar flick but to do research in Nanoscale Electronics, Quantum Chemistry, Computational Chemistry/Biochemistry, Aerodynamics, Cell Cycle Modeling, Molecular Statics, and Computational Acoustics, and VT is working with agencies like the NSF, NIH, DOE, DARPA, DoD, AFOSR, ONR, National Labs, NASA, PSC, ISR, etc., and no one is concerned about memory errors. This post was edited by sodajerk on Wednesday, November 19, 2003 at 13:09. |
| #4 By 3339 (64.160.58.135) at 11/21/2003 2:53:33 PM |
|
The question is parker: are people NOT using Big Mac? Do the people who are running algorithms understand its limitations and strengths? Does it still allow them to perform their work? At the least cost possible?
You sit around all day attempting to act as if your words alone will kill Apple, that they can't succeed at everything? The fact is they and VT clearly have succeeded at a very impressive project. Hell, even the fact that they will move to ECC means that months after the first Apple supercomputer has been built (instantly jumping to #3) it will support ECC RAM. As for your claim that every thing needs to be run twice, I don;t see that at all. I see "may" be run twice. I don;t know if this is for test purposes initially to determine the rate of error. I don;t know if this is only for extremely complex math. I do know that VT staff have said that multiple runs will not be necessary for most of the projects being run. Jesus, dude, you quote an author of an article and he tells you: "you are misinterpreting me, and this is how it really is." And you sit there going, "No, no, no, waaaah, waaaah, waaah, I don't want to hear it." Wake up, dude and join reality. Oh and by the way, where are your posts about the Jaguar security update? This post was edited by sodajerk on Friday, November 21, 2003 at 14:55. |
| #5 By 3339 (64.160.58.135) at 11/21/2003 5:44:26 PM |
|
What crack are you smoking now? I didn't respond to your "supposedly" comment. tbridge did. tbridge is Tom Bridge. I haven't responded to your pathetic-ness in that regard because Tom already has. I've been busy the last couple of days so I can't go hunting for links. But I will and I'll post them if in the meantime you post on the topic of Jaguar security updates.
As for your sources, you've been shot down by the author of one of them. And the second is a patch work of a discussion thread quoting Dr. V under who knows what circumstances by people not involved in the project. You are notorious for using poor sources which are clearly not accurate. To reiterate: 1. Terascale is engaging many projects; these projects are satisifed with the level of accuracy for their needs. 2. it is still the cheapest performance to cost ratio of any supercomputer thus far, with BUTTLOADS of performance at that. 3. VT and Apple are apparently working on an ECC solution 4. Apple provided the best, cheapest, quickest solution to attaining #3 status 5. In the meantime they are doing just fine. 6. They will continue to improve Terascale/Big Mac over time, including incorporating ECC 7. What's your beef? And again, where are your comments on Jaguar security updates? And why the fck do you think I am Tom Bridge? Have you flipped your lid? |
| #6 By 3339 (64.160.58.135) at 11/21/2003 5:50:07 PM |
| Uh, parker, the flaws were patched in less than a month, right? That's faster than you can ever expect a patch to be fixed for Windows now. |
| #7 By 3339 (64.160.58.135) at 11/21/2003 7:30:45 PM |
|
You are a complete idiot, parker. All of what you state is dependent on sources which have proven themselves suspect and inaccurate. And...
"No. The flaws were fixed for Panther in x number of days. Then Apple told @Stake to go and announce the flaws." Isn't that Microsoft's policy? How is that any different or worse? "Then Apple was shamed into announcing it would fix the flaws." No, it wasn't. When a reliable source asked, they said: "We fix all secuirty issues for all supported operating systems when possible." Which is borne out by policy and history. "Then Apple dragged its feet for a month leaving all the Jaguar users hanging in the wind." Feet dragging? Are you suggesting they had the fix and didn't release it? Any proof? Again, how is this any different from Windows. I can point to MANY, MANY flaws which remain unpatched. Now you have to wait a MONTH no matter what, even if MS has created a patch. Hanging in the wind? Two of the flaws required either local admin access (not very hard to protect against--just don't let any unauthorized person login as admin) and local file access (again either required the person be logged in or have remote access--very easy to stop) respectively. The third flaw required /core to be activated which I have never seen done under any circumstances and cannot fathom why anyone would ever do so. In comparison to MS flaws, these were minor, and most users were always safe. Apple didn't screw anyone. @Stake gave bad advice. Other security firms actually published fixes which were easy to implement. All you had to do was download updated software binaries, disable a feature or two, or run a script. @Stake tried to sell some copies of Panther, not Apple. And by the way, the flaws were found when? I believe Oct 28, but if you can point to an earlier date, I'd be happy to correct it. So to be clear, we are talking about TWENTY-TWO days. And had this been an MS issue, and they had the patches available on the 19th, you would have waited through Thanksgiving weekend, and then the first week of December, and then gotten them on the 11th of December. Consider that. This post was edited by sodajerk on Friday, November 21, 2003 at 19:58. |
|
Write Comment
Return to News |
Displaying 1 through 25 of 310 Last | Next |
The time now is
2:23:13 PM
ET.
Any comment problems? E-mail us |
