Cisco Systems today introduced its next-generation Internet core router, the CRS-3, with about three times the capacity of its current platform. "The Internet will scale faster than any of us anticipate," Cisco Chairman and CEO John Chambers said during a webcast announcing the product. At full scale, the CRS-3 has a capacity of 322Tbit/sec., roughly three times that of the CRS-1, which was introduced in 2004. It also has more than 12 times the capacity of its nearest competitor, Chambers said.
Post a Comment
Member since:
2005-12-15
Member since:2009-10-04
Member since:2005-07-06
Member since:
2005-08-18
Member since:2008-07-15
You really expected the marketing speak to be accurate?
This router certainly has an impressive speed capability, but what good will this really do? How many people are going to be able to afford this router, even businesses probably won't spend the money unless they're doing high-performance network clustering, and it would have to be local networking to get that speed anyway. Wan speeds are only as good as both ends of the connection, even if some in-between points are faster than both ends the speed you get is limited to what both ends can upload and download themselves. This is more likely to change how some businesses do clustering, but change the internet? Not really, unless they intend to give these routers to ISPs along with the cabling to use them to this capacity, and then give everyone a DSL or Cable or Fios modem that can handle speeds like this, not to mention role out the proper wiring into people's homes and business offices... you get the idea.
Member since:2006-01-01
Member since:2006-07-26
It's not just the end points though. The connection is as slow as the slowest point the data flows through, including the end points. These routers are intended more for back bone connections across vast distances more then for ISPs. Such as between phone companies, universities, and governments.
Member since:2006-01-01
You're right about evolution, but someone still has to invest in R&D to get us there.
Also, the example in the ComputerWorld article is a bad one. There are two better examples of the volume of data this device can process in the press release (http://tinyurl.com/yljrcro):
* Every man, woman and child in China can make a video call simultaneously [that's 1 billion people!]
* Every motion picture ever created can be streamed in less than four minutes
Many ISPs have servers with local copies of websites: this is cheaper for them, because they pay reduced access fees to backbones, and faster for their customers. This can't be done with newer Internet technologies, such as video and audio teleconferencing, DRM Internet-based TV (esp. Hulu) and BitTorrent. These are all technologies that have really taken off in the last five years and can't be cached locally by ISPs - they must travel over the backbone.
Edited 2010-03-10 14:22 UTC
Member since:2005-08-18
True, but that in itself does not make a product groundbreaking.
* Every motion picture ever created can be streamed in less than four minutes
I don't put much faith in marketing speak, especially since the last point is pure nonsense.
Look, I'm sure it's a great product but that does not make it ground breaking and capable of changing the internet forever.
Bullshit is bullshit, no matter how you spin it.
Member since:2006-01-01
I guess I am trying to look beyond the marketing gabble and at the more practical implications of this. Saying that the device can stream the world's entire movie archive is a bit silly, but it does make me think that the transmission of private Internet TV channels, streaming DVD rentals etc. could become more financially viable in the near future.
In 2004, YouTube was brand new, Skype had only just started (first release in 08/2003), and BitTorrent had taken off only a few years prior (2002-3). CRS-1 and the Juniper clones made these kinds of direct high bandwidth communications possible.
Companies are already developing hologram technology (not the crappy CNN variety but single-camera 3D transmissions) but these require insanely high bandwidth rates - only affordable at ISP level. Who knows where that will lead in a few years time when CRS-3 and eventual clone devices start being deployed?
Edited 2010-03-10 14:58 UTC
Member since:
2009-05-11
Member since:2006-02-27
Member since:2006-01-01
Think about this in terms of CPUs.
Intel got its Pentium 4 processor clock speed up to 3.4GHz in 2004. Because nobody could work out how to make significantly faster CPUs without causing overheating and memory faults, Intel turned to dual cores - in effect, they cheated by loading two or more slower CPUs together to create a combined faster clock speed (e.g. Core 2 Duo 2.4GHz = "4.8Ghz" theoretically).
No single Intel core has gone much above the 3.4GHz clock speed - in essence, we've hit a physical limitation that hasn't been solved in six years.
Meanwhile, in the same time period Cisco has TRIPLED the throughput of its CRS router. To give you an idea of how expensive and difficult developing this type of carrier device is, take a look at Juniper Networks' competing T-series. From the time that the CRS-1 launched in 2004, it took Juniper another three years (until 2007) to get the T1600 model, which matches the earlier Cisco device's throughput, to market.
Member since:
2006-06-28
Member since:2005-07-28
