The first, what happens if encryption makes it impossible to really tell what anything is? How does a non-net-neutral ISP then determine tiered prices for the content? Does encryption effectively enforce Net Neutrality?
Encryption just shoots yourself in the foot, since an ISP can just put all encrypted traffic into the lowest-speed or highest-cost tier. So instead of the ISP penalizing VoIP, now they will penalize all your traffic.
That dark fiber doesn't run to your house. Think about it; the evil is all in the last mile.
Re:Sun made the move in April
on
IBM Opts for AMD
·
· Score: 2, Informative
The x3455 is basically an Opteron version of the x336.
Re:without HyperTransport, AMD would be dead
on
IBM Opts for AMD
·
· Score: 1
how is cache coherency handled [in Woodcrest/Blackford]?
Bus transactions that need to be seen by the other processor will be reflected onto the other FSB by the northbridge. So even though there are two FSBs, the effective bandwidth is not quite twice because of the broadcast traffic. (This is essentially the same scheme used by the PowerPC 970 and AMD K7, BTW.)
And how will the pricing of a 2-socket Woodcrest system compare to that of a two-socket dual-core Opteron system?
You can bet that Intel will make it comparable or cheaper.
Re:That's an easy one.
on
IBM Opts for AMD
·
· Score: 4, Informative
all this gain is LOST when useing 2 CPUs (differant sockets) running in 64 bit mode, you will thrash the cache just like usual, the fancy microcode optimizer disables, it all falls apart.
Sorry, but there is no evidence for this. Real-world benchmarks show that 2-socket, 64-bit Woodcrest systems have good performance (usually better than Opteron).
Unless you are mmap'ing those large files, you don't need 64-bit. Why, I remember back in the 90s when BeOS supported lottabyte files on a 32-bit processor...
While I understand that 4x4 will be the desktop part, designed for single-socket systems, I expect it will arrive with similar Opteron-2xx series processors that have 4 cores.
No, 4x4 is two sockets with dual-core processors in them. Since you already have that, 4x4 won't benefit you. Basically 4x4 is a way to trick gamers into buying quad Opteron systems under a different name.
People who use computers for work and who want four cores already bought Opteron workstations. That's why 4x4 and Kentsfield are targeted at the enthusiast (aka more money than sense) market.
Wouldn't that mean that it probably wouldn't pay in any computer dvd player?
According to Wikipedia (I know, grain of salt), ARCCOS "deliberately creates a number of sectors on the DVD with corrupted data that causes DVD copying software to produce errors. Normal DVD players do not ever read these sectors since they follow a set of instructions encoded on the disc telling them to skip it. Less sophisticated DVD ripping programs do not follow these instructions but try to read every sector on the disk sequentially, including the bad ones."
If lots of discs were published using this method, couldn't DVD Shrink be modified to not choke.
DVD Shrink is no longer being developed and is not open source, so it can't be fixed. There are tools that can copy corrupted DVDs, but they're not as convenient. As with most DRM, the goal of corrupted DVDs is to make them harder to copy, not impossible to copy.
Why? Sometimes Xeons are even cheaper than desktop processors. (e.g. Xeon 5160 is less than a Core 2 Extreme)
the AMD 4x4 platform concept - which is basically releasing Athon64's that can do dual-cpu but without the "server" chip premium.
I don't think you're taking into account market segmentation; 4x4 will undoubtedly have an "enthusiast" premium that is just as much as the "server" premium on Opterons.
How AMD plans to prevent dual CPU servers based on A64's is another matter...
Just make A64s the same price as Opterons and laugh all the way to the bank.
What do you mean by cheap hardware? Do you mean fully designed, implemented and commercially available set top boxes, HUDs, etc.
Yes. Consider examples such as the Linksys WRT54G, TiVo, Sharp Zaurus, or Neuros products; these have probably enabled ten times as much cool hacking as developer boards have. Besides low price, complete products allow incremental hacking; the product has some useful features when you buy it and thus hackers can focus on adding features rather than starting from scratch (or a fairly low-level base platform, such as MontaVista).
Just the other day I was reading about all the complexity and licensing issues that Neuros is having with TI chips, and I thought "if only they had GStreamer".
It remains to be seen what effect this will have on the GStreamer community, as cheap DaVinci hardware may never find its way into hackers' hands.
Adding heterogeneous multiprocessing support to GStreamer may also help when it comes time to run it on the PS3.
Isn't the virtual console a big part of Wii? And aren't virtual console games downloaded instead of bought on disc? And do you think there's any chance of buying, selling, or renting used virtual console games?
Slashdot Mirror
The first, what happens if encryption makes it impossible to really tell what anything is? How does a non-net-neutral ISP then determine tiered prices for the content? Does encryption effectively enforce Net Neutrality?
Encryption just shoots yourself in the foot, since an ISP can just put all encrypted traffic into the lowest-speed or highest-cost tier. So instead of the ISP penalizing VoIP, now they will penalize all your traffic.
That dark fiber doesn't run to your house. Think about it; the evil is all in the last mile.
The x3455 is basically an Opteron version of the x336.
how is cache coherency handled [in Woodcrest/Blackford]?
Bus transactions that need to be seen by the other processor will be reflected onto the other FSB by the northbridge. So even though there are two FSBs, the effective bandwidth is not quite twice because of the broadcast traffic. (This is essentially the same scheme used by the PowerPC 970 and AMD K7, BTW.)
And how will the pricing of a 2-socket Woodcrest system compare to that of a two-socket dual-core Opteron system?
You can bet that Intel will make it comparable or cheaper.
all this gain is LOST when useing 2 CPUs (differant sockets) running in 64 bit mode, you will thrash the cache just like usual, the fancy microcode optimizer disables, it all falls apart.
Sorry, but there is no evidence for this. Real-world benchmarks show that 2-socket, 64-bit Woodcrest systems have good performance (usually better than Opteron).
Both formats have mandatory managed copy.
Performance in full virtualized mode is noticeably worse than in paravirtual mode, so people are fighting over paravirtual interface standards.
For ZFS I think SAS JBODs are a better solution that AoE. And AoE will get squeezed even more when SAS SANs arrive.
Unless you are mmap'ing those large files, you don't need 64-bit. Why, I remember back in the 90s when BeOS supported lottabyte files on a 32-bit processor...
Is Monster Cable making flash cards now?
While I understand that 4x4 will be the desktop part, designed for single-socket systems, I expect it will arrive with similar Opteron-2xx series processors that have 4 cores.
No, 4x4 is two sockets with dual-core processors in them. Since you already have that, 4x4 won't benefit you. Basically 4x4 is a way to trick gamers into buying quad Opteron systems under a different name.
4x4 means 4 cores and 4 GPUs. I guess 4+4 just didn't sound as cool.
People who use computers for work and who want four cores already bought Opteron workstations. That's why 4x4 and Kentsfield are targeted at the enthusiast (aka more money than sense) market.
Wouldn't that mean that it probably wouldn't pay in any computer dvd player?
According to Wikipedia (I know, grain of salt), ARCCOS "deliberately creates a number of sectors on the DVD with corrupted data that causes DVD copying software to produce errors. Normal DVD players do not ever read these sectors since they follow a set of instructions encoded on the disc telling them to skip it. Less sophisticated DVD ripping programs do not follow these instructions but try to read every sector on the disk sequentially, including the bad ones."
If lots of discs were published using this method, couldn't DVD Shrink be modified to not choke.
DVD Shrink is no longer being developed and is not open source, so it can't be fixed. There are tools that can copy corrupted DVDs, but they're not as convenient. As with most DRM, the goal of corrupted DVDs is to make them harder to copy, not impossible to copy.
It's probably just like ARCCOS: the disc is corrupt enough that DVD Shrink chokes on it, but not corrupt enough to confuse a set-top DVD player.
it seems wasteful to buy Xeons for a workstation.
Why? Sometimes Xeons are even cheaper than desktop processors. (e.g. Xeon 5160 is less than a Core 2 Extreme)
the AMD 4x4 platform concept - which is basically releasing Athon64's that can do dual-cpu but without the "server" chip premium.
I don't think you're taking into account market segmentation; 4x4 will undoubtedly have an "enthusiast" premium that is just as much as the "server" premium on Opterons.
How AMD plans to prevent dual CPU servers based on A64's is another matter...
Just make A64s the same price as Opterons and laugh all the way to the bank.
For a four-core system you want Xeon 51x0; it's the multi-socket version of Conroe.
A Woodcrest system should have very similar power consumption to a 4x4 system. For example, check out AnandTech's comparison.
It looks like Windows will be faster on an Apple machine than on any other factory-built desktops.
There's no evidence for this. You can buy a Dell or HP that has the exact same components as a Mac Pro.
What do you mean by cheap hardware? Do you mean fully designed, implemented and commercially available set top boxes, HUDs, etc.
Yes. Consider examples such as the Linksys WRT54G, TiVo, Sharp Zaurus, or Neuros products; these have probably enabled ten times as much cool hacking as developer boards have. Besides low price, complete products allow incremental hacking; the product has some useful features when you buy it and thus hackers can focus on adding features rather than starting from scratch (or a fairly low-level base platform, such as MontaVista).
On x86-64. On 32-bit systems Vista will load anything to maintain backwards compatibiity.
Just the other day I was reading about all the complexity and licensing issues that Neuros is having with TI chips, and I thought "if only they had GStreamer".
It remains to be seen what effect this will have on the GStreamer community, as cheap DaVinci hardware may never find its way into hackers' hands.
Adding heterogeneous multiprocessing support to GStreamer may also help when it comes time to run it on the PS3.
Isn't the virtual console a big part of Wii? And aren't virtual console games downloaded instead of bought on disc? And do you think there's any chance of buying, selling, or renting used virtual console games?
I mean, its not like you can attach this thing to a SAN and carve off LUNS to utilize the space from other devices in the fabric.
Actually, you can do that with iSCSI.
The way I see it, an 11-disk RAIDZ stripe gives you 10% redundancy - 10 disks worth of data and 1 disk worth of parity.