That's a nice theory, but Cray's XD1 "supercomputer" uses the same Mellanox switch chips as some "clusters". Cray is splitting hairs to justify their product.
(BTW, I get 100 us ping time on my GigE network, but you're right that that's still 100x too slow for HPC.)
Good clusters don't use IP; they use Infiniband, Myrinet, or Quadrics, which all have OS bypass and trasport offload features so that the app can talk directly to the NIC. In fact, Cray's XD1 "supercomputer" uses the same Infiniband interconnect as some "clusters"; Cray just has better NICs.
If RPOWs are really worth something, you could set up a currecy exchange to convert them to real money, and RPOW servers could fund themselves with a transaction fee (which would be paid in RPOWs themselves, naturally). It wouldn't work for many reasons, but it's an interesting idea.
That's not the point. It is a technical fact that equipment that exists today can suffer from interference (or "receiver confusion" for the politically correct).
Wow, Wal-Mart is ripping you off. I saw an Apex DVD player at the grocery store for $38 the other day. It was stacked up near the checkout lines with all the other impulse buys. That's when something is really commoditized.
The difference is that a DVD player is a DVD player, while RAM standards change every year or two. In the case of DRAM, manufacturers have reduced production on SDR (using that production capacity for DDR instead) because of lower demand, which causes the price to stay the same. It appears that these Chinese DVD player manufacturers can't easily shift to some other, more profitable product, so they just keep cranking out DVD players at lower and lower prices.
The author is obviously not an RF engineer, so (surprise!) he doesn't use RF engineering terminology. It is clear from the context that "very low frequency" means below 3GHz. And the author is right: even 100 MHz (3%) of that spectrum is enough to provide a lot of last-mile access.
"Applicable Patent Rights" mean:... (i) claims of patents that are now or hereafter acquired, owned by or assigned to You and (ii) that cover subject matter in Your Modifications, taken alone or in combination with Original Code.
So we're only talking about the user's patents which cover DSS, not all of the user's patents.
You hereby grant to any person or entity receiving or distributing Covered Code under this License a non-exclusive, royalty-free, perpetual, irrevocable license, under Your Applicable Patent Rights... to use, reproduce, display, perform, modify, sublicense, distribute and Externally Deploy Your Modifications of the same scope and extent as Apple's licenses
So you only grant back your patents for people to use with your code, not for any purpose. And if you make no modifications, you don't have to grant any patent license to anyone.
I don't think you understand that post you linked to. Live Channel is indeed $1000, but it is software, not hardware. And Live Channel is a switcher/effects system, not a codec. And you only need QT Broadcaster or Live Channel or (free) MPEG4IP if you want to do live streaming.
If you want to do on-demand streaming, all you need is software to create MP4 files, of which there is plenty out there.
VLC and (I think) Mplayer can play MPEG-4 streams from QuickTime/Darwin Streaming Server.
Slashdot Mirror
The XD1 uses Mellanox Infiniband switches.
How would a cluster of commodity hardware behave, if using a low-latency interconnect?
#2 in the world, if you consider Itanium to be commodity. The fastest Xeon cluster is #5 in the world.
A Cray XD1 is quite different from an SGI Altix.
the XD1 gives you a single systems image, which simplifies programming and makes shared memory applications
Is this really true? The XD1 Web site just talks about MPI.
That's a nice theory, but Cray's XD1 "supercomputer" uses the same Mellanox switch chips as some "clusters". Cray is splitting hairs to justify their product.
(BTW, I get 100 us ping time on my GigE network, but you're right that that's still 100x too slow for HPC.)
Only in routed Infiniband networks, which no one uses. The normal Infiniband protocol is very lean and totally different from TCP/IP.
Good clusters don't use IP; they use Infiniband, Myrinet, or Quadrics, which all have OS bypass and trasport offload features so that the app can talk directly to the NIC. In fact, Cray's XD1 "supercomputer" uses the same Infiniband interconnect as some "clusters"; Cray just has better NICs.
With many well-known strong encryption schemes, why was the weak WEP made standard?
So what well-known strong link-level encryption schemes would you suggest they should have used?
RPOW servers are so slow and expensive that any spammer who buys one will go out of business.
RPOWs are worth so little that it's a waste of time to "crack" them. It's like making counterfeit pennies.
If RPOWs are really worth something, you could set up a currecy exchange to convert them to real money, and RPOW servers could fund themselves with a transaction fee (which would be paid in RPOWs themselves, naturally). It wouldn't work for many reasons, but it's an interesting idea.
Generally these schemes involve the token being in some way tied to the specific message. A hash containing the to and from addresses for instance.
Yep, but that doesn't work if you want to make the POWs reusable.
That's not the point. It is a technical fact that equipment that exists today can suffer from interference (or "receiver confusion" for the politically correct).
Wow, Wal-Mart is ripping you off. I saw an Apex DVD player at the grocery store for $38 the other day. It was stacked up near the checkout lines with all the other impulse buys. That's when something is really commoditized.
The difference is that a DVD player is a DVD player, while RAM standards change every year or two. In the case of DRAM, manufacturers have reduced production on SDR (using that production capacity for DDR instead) because of lower demand, which causes the price to stay the same. It appears that these Chinese DVD player manufacturers can't easily shift to some other, more profitable product, so they just keep cranking out DVD players at lower and lower prices.
The author is obviously not an RF engineer, so (surprise!) he doesn't use RF engineering terminology. It is clear from the context that "very low frequency" means below 3GHz. And the author is right: even 100 MHz (3%) of that spectrum is enough to provide a lot of last-mile access.
For reference, here's a link to the APSL.
... (i) claims of patents that are now or hereafter acquired, owned by or assigned to You and (ii) that cover subject matter in Your Modifications, taken alone or in combination with Original Code.
... to use, reproduce, display, perform, modify, sublicense, distribute and Externally Deploy Your Modifications of the same scope and extent as Apple's licenses
"Applicable Patent Rights" mean:
So we're only talking about the user's patents which cover DSS, not all of the user's patents.
You hereby grant to any person or entity receiving or distributing Covered Code under this License a non-exclusive, royalty-free, perpetual, irrevocable license, under Your Applicable Patent Rights
So you only grant back your patents for people to use with your code, not for any purpose. And if you make no modifications, you don't have to grant any patent license to anyone.
Microsoft tried this and it totally sucked. I'm sure Apple could do better; the question is how much better.
XVID implements MPEG-4, which is streamable if you know what you're doing.
I don't think you understand that post you linked to. Live Channel is indeed $1000, but it is software, not hardware. And Live Channel is a switcher/effects system, not a codec. And you only need QT Broadcaster or Live Channel or (free) MPEG4IP if you want to do live streaming.
If you want to do on-demand streaming, all you need is software to create MP4 files, of which there is plenty out there.
VLC and (I think) Mplayer can play MPEG-4 streams from QuickTime/Darwin Streaming Server.
Darwin Streaming Server is free and has existed for years. There is also free software to encode MP4 files.
Speed and security are orthogonal, but of course all recent 802.11 chips support i.
And I hear the limitations and faults in WAP 2 have been addressed in HTML 3.2...
Actually, Fujitsu's SPARC64 V has pretty decent performance.
IBM does not give away AIX any more. A "Linux-ready" pSeries box with no OS costs less than a pSeries with AIX.
That's called a smart card/USB token/iButton.