You don't watch it live, of course. Your computer/set top box "records" the download (possibly faster than real time) and shows the file to you once the whole thing has been received.
Did you notice that the article says Asia? People with 100Mbps FTTH should have no trouble downloading games. I read an article the other day saying that in Japan "heavy users" are people who download more than 2.5GB per day, so downloading a new PS3 game every week or so should be no big deal.
But in the future you'll have no choice. DDR3 can probably only fit one DIMM per channel (due to the capacitance, as already mentioned). If you only have enough pins for two DDR3 channels, then your high-end processor can only use 2 DIMMs, which simply is not enough. Or you could use those same pins for four to six FBD channels; even if you only have one DIMM per channel (no latency hit) you still come out ahead.
You're overlooking the frequency difference; because FBD is point-to-point, it can run at 5GHz or more, compared to a mere 800MHz for DDR2. Thus with an equal number of pins, FBD gives much more bandwidth.
Yes, with FB you can put all your DIMMs on a single set of signal lines.
That's not how it is used, so I'm not sure why you are emphasizing that. IIRC, Blackford systems have 4 FBD channels (using fewer pins than two DDR2 channels).
instead of having hundreds of people polling a single RSS feed, why not have the server which hosts the RSS feed actually PUSH the updates out to the people who are interested?
FeedTree can operate in that mode, or if the server operator is too lazy to install it, FeedTree will still provide some benefit.
We already have a nice and simple protocol (XMPP) which could be used for this, although admittedly PubSub isn't as final as it could be.
Doesn't this lead to potentially high fanout (with the attendant concentrated bandwidth consumption)? One benefit of FeedTree is that it spreads the work around.
If GMail were to incorporate an RSS reader (the way Thunderbird does), it could potentially update many, many users with a single hit of each RSS site.
There are already several large Web-based aggregators that work this way, but for various reasons many people prefer local aggregators (just as many people prefer local mail clients instead of GMail). FeedTree solves the bandwidth and latency problem for local aggregators.
60GHz signals are absorbed by oxygen (much like 2.4GHz is absorbed by water), so the FCC figured that frequency must be useless and the public might as well be allowed to play with it.
I don't know what weird telco you use, but the RBOCs generally don't structure their service that way at all, especially for "consumer" service. They own all the equipment up to the demarc on the outside of your house and what they put on their side is none of your business. 80% idle capacity? That's not your problem, consumer!
I'm not defending the telcos, just documenting their philosophy.
IIRC it's 622Mbps shared downstream and 155Mbps shared upstream. Depending on how many customers share each fiber, 15-20Mbps per customer sounds reasonable.
I didn't bother to listen to the podcast, but luckily this is Slashdot so no one will hold it against me.
Geoff Huston's "IPv6: Extinction, Evolution or Revolution?" is probably the most insightful thing I've ever read about IPv6 deployment, although the conclusion is pretty negative.
But assuming that IPv6 is worth deploying, Microsoft is way ahead in getting computers IPv6-enabled. Their work on Teredo should make life a lot easier for P2P developers.
When you consider that a SATA RAID 10 array is cheaper (and maybe faster and more reliable) than a SCSI RAID 5 array, maybe the extra peace of mind is worth a little "waste".
This market segmentation is quickly going away, though. Several companies are making reliable SATA drives, and the SATA Raptor is just as fast (and supposedly as reliable) as a 10K RPM SCSI drive.
Intel has open source wireless drivers, open source graphics drivers, etc. Their technology tends to lag behind the cutting edge, but maybe that's the price of freedom.
There is a minimum fee of around $7/domain/year. Apparently a "parked" domain can make over $7/year in ad revenue, so that's why companies are registering them like crazy.
Slashdot Mirror
You don't watch it live, of course. Your computer/set top box "records" the download (possibly faster than real time) and shows the file to you once the whole thing has been received.
Did you notice that the article says Asia? People with 100Mbps FTTH should have no trouble downloading games. I read an article the other day saying that in Japan "heavy users" are people who download more than 2.5GB per day, so downloading a new PS3 game every week or so should be no big deal.
But in the future you'll have no choice. DDR3 can probably only fit one DIMM per channel (due to the capacitance, as already mentioned). If you only have enough pins for two DDR3 channels, then your high-end processor can only use 2 DIMMs, which simply is not enough. Or you could use those same pins for four to six FBD channels; even if you only have one DIMM per channel (no latency hit) you still come out ahead.
You're overlooking the frequency difference; because FBD is point-to-point, it can run at 5GHz or more, compared to a mere 800MHz for DDR2. Thus with an equal number of pins, FBD gives much more bandwidth.
Yes, with FB you can put all your DIMMs on a single set of signal lines.
That's not how it is used, so I'm not sure why you are emphasizing that. IIRC, Blackford systems have 4 FBD channels (using fewer pins than two DDR2 channels).
If this PubSub service is centralized, then it won't be free. If it's decentralized, then it's essentially similar to FeedTree.
instead of having hundreds of people polling a single RSS feed, why not have the server which hosts the RSS feed actually PUSH the updates out to the people who are interested?
FeedTree can operate in that mode, or if the server operator is too lazy to install it, FeedTree will still provide some benefit.
We already have a nice and simple protocol (XMPP) which could be used for this, although admittedly PubSub isn't as final as it could be.
Doesn't this lead to potentially high fanout (with the attendant concentrated bandwidth consumption)? One benefit of FeedTree is that it spreads the work around.
The Internet does not support IP multicast, so FeedTree uses application-level end-system multicast.
If GMail were to incorporate an RSS reader (the way Thunderbird does), it could potentially update many, many users with a single hit of each RSS site.
There are already several large Web-based aggregators that work this way, but for various reasons many people prefer local aggregators (just as many people prefer local mail clients instead of GMail). FeedTree solves the bandwidth and latency problem for local aggregators.
60GHz signals are absorbed by oxygen (much like 2.4GHz is absorbed by water), so the FCC figured that frequency must be useless and the public might as well be allowed to play with it.
I don't know what weird telco you use, but the RBOCs generally don't structure their service that way at all, especially for "consumer" service. They own all the equipment up to the demarc on the outside of your house and what they put on their side is none of your business. 80% idle capacity? That's not your problem, consumer!
I'm not defending the telcos, just documenting their philosophy.
IIRC it's 622Mbps shared downstream and 155Mbps shared upstream. Depending on how many customers share each fiber, 15-20Mbps per customer sounds reasonable.
Tunnel brokers are obsolete and inefficient; you should advise people to use 6to4 instead.
I didn't bother to listen to the podcast, but luckily this is Slashdot so no one will hold it against me.
Geoff Huston's "IPv6: Extinction, Evolution or Revolution?" is probably the most insightful thing I've ever read about IPv6 deployment, although the conclusion is pretty negative.
But assuming that IPv6 is worth deploying, Microsoft is way ahead in getting computers IPv6-enabled. Their work on Teredo should make life a lot easier for P2P developers.
MS is writing their own hypervisor and I suspect Apple will write their own if they want one.
Of course, you can bypass all those shortcomings if you get a good hardware RAID controller.
Sure, if you're made of money.
When you consider that a SATA RAID 10 array is cheaper (and maybe faster and more reliable) than a SCSI RAID 5 array, maybe the extra peace of mind is worth a little "waste".
That's more complex, more expensive, and slower than just using a SATA (or SAS) RAID card.
This market segmentation is quickly going away, though. Several companies are making reliable SATA drives, and the SATA Raptor is just as fast (and supposedly as reliable) as a 10K RPM SCSI drive.
Sometimes Apple uses GCC and sometimes ICC. Both compilers have non-backwards-compatible flags, so I would guess Apple is using them.
What I want is more public bandwidth.
Have you even touched the 5GHz U-NII band that the FCC gave you?
Intel has open source wireless drivers, open source graphics drivers, etc. Their technology tends to lag behind the cutting edge, but maybe that's the price of freedom.
Intel-based Xserves will still include a license for OS X Server. If you don't consider than an incentive, then you're not the target market anyway.
OpenVZ won't benefit from virtualization-assist hardware features.
There is a minimum fee of around $7/domain/year. Apparently a "parked" domain can make over $7/year in ad revenue, so that's why companies are registering them like crazy.