.. Anakin runs in, shouting "We're not supposed to kill him!" Mace turns and said "I don't remember askin' you a G@#DAMN thing" - - [Whacks Palpatine]
Or
.. Anakin runs in, shouting "We're not supposed to kill him!" [Mace whacks Palpatine] , [Big dark force energy explosion] "Oh, I'm sorry. Did that break your concentration? I didn't mean to do that. Please, continue. I believe you were saying something about 'not supposed to kill.'"
I hope Intel pulls off multi-core better than they did hyperthreading. The P4 netburst architecture simply weak, and hyperthreading is really just a patch to make it not suck quite so badly. I "upgraded" from a 1.6GHz AthlonXP to a 2.8GHz P4 Dell and was horribly disappointed with task switching performance. Tried throwing more RAM at it. The P4 with 1GB was still slower than the AthlonXP with 768MB. OTOH, I setup a friend's new Dell with the latest PCI express chipset and was really impressed with the speed, so maybe Intel is finally getting it together... Question is: will the netburst architecture benefit more from dual core than it did from hyperthreading? Its essentially like having an SMP P4 system, so it ought to.
"And in July, DoubleClick Inc.'s DNS (domain name system) was attacked and unable to serve ads for a similar time frame." Citizens will be running in the streets in panic if they're pop-up ads stop working - Oh the humanity!!
I've found the linux kernel's
built-in RAID
capabilities more than adequate for most of my fault tolerance needs. The
best part is I can move the drives to pretty much any system - a new
motherboard, whatever - without having to worry about kernel support or finding
that IDE driver. If a drive fails I can boot its mirror up in any system
and be in great shape. I also use the utility
mdadm to email me
if one of the drives fails. For some linux firewall systems I've built, I
use old crappy 6GB drives, but mirror them so there's no risk if one of them
goes out. Looking at my basement firewall now and...
Is the Opteron even capable of scaling up to 32 or 64-way systems? If the only limitation is the OS, then Sun should seriously look in to making Solaris do it.
I know that Opteron uses an advanced (compared to Intel's Xeon) hypertransport bus to get up to 8-way systems rather easily and with plenty of memory bandwidth for each CPU.
So what kind of system board hacks would be required to get Opteron up to 32-way? How does SPARC accomplish it?
Doesn't it suck when beaucrats make these big research decisions, wasting millions in research dollars? If resonable scientists had been involved in the decicion-making process they wouldn't come within a football field of this project. Frankly, I'd rather they spend the money on in-flight 3D games for the passengers...
The number one problem with "locally built" low cost PCs: They're cheap. There's no way around it. The advantage OEMs have is that they can put quality tested components together at prices impossible for normal folks to match. PCs are an economy of scale, just like any other mass produced product. There's no way you can build up a QC infrastructure to match Dell's. Whenever a friend asks me to build them a PC, I ask for his credit card, order him a Dell and then setup all the software on it. I only build PCs for myself or close loved ones, because a temperamental homemade PC can really erode confidence, believe me. I shutter to imagine what would happen if I had dozens of them out there.
...When the CPU speeds finally hit 1GHz, I guess. The 133Mhz SDRAM probably won't start to cripple the FSB/CPU until the chips are pushing the same speeds as Intel's and AMD's.
According to neurotik's post, this situation could be solved by simply not using a smarthost and delivering SMTP directly, since Verizon apparently does not block such things. So Verison, you're off the hook... for now!
I've never used Verison for any of my client's Internet services, so no big problem for now, but this policy is quite disturbing. Imagine that you are a small business using a Linux box with a dynamic DNS client. That box can host a multitude of services for you company on the Internet, one of the most important of which is email. You're paying only for the DSL or cable Internet access and get to administrate accounts and sendmail policies yourself. The power is in your hands.
Now imagine that suddenly the smarthost that your sendmail server forwards all outgoing email to refuses to deliver messages that you and your coworkers send. And its because they're not From: a valid email address hosted by the ISP!!
Verison is taking away a key capability here. It basically means that Verison's users can no longer run their own email servers. The power is now in Verison's hands. Ready to pay for that email hosting service Verison has? You'd better be.
VA wasn't all that great of a hardware provider compared to the likes of Compaq or Dell. I've used several VA rackmount servers, from Multi CPUs with external RAID to cheap n' dirty 1U IDE servers. The rackmount kits are cheap and shoddy compared to Compaq's hearty bearing rackmounts. And Compaq's Linux support is excellent. Response time is wonderful, the techs actually know the OS, and their proprietary config tools (like SmartStart) and system BIOS have full Linux support. All in all Compaq has beaten VA at their own game. And the hardware is just plain better.
I think VA is making the right move here. They've thrown a few rocks at Goliath, but now its time to run back to the hills of Andover...
To use these new feeds on the Firefox toolbar:
Uhh.. Sony is supposed to be in some kind of deal with nVidia, so the PS3 won't be using an ATI chip...
Mace turns and said "I don't remember askin' you a G@#DAMN thing" - - [Whacks Palpatine]
Or
[Mace whacks Palpatine] , [Big dark force energy explosion]
"Oh, I'm sorry. Did that break your concentration? I didn't mean to do that. Please, continue. I believe you were saying something about 'not supposed to kill.'"
I hope Intel pulls off multi-core better than they did hyperthreading. The P4 netburst architecture simply weak, and hyperthreading is really just a patch to make it not suck quite so badly. I "upgraded" from a 1.6GHz AthlonXP to a 2.8GHz P4 Dell and was horribly disappointed with task switching performance. Tried throwing more RAM at it. The P4 with 1GB was still slower than the AthlonXP with 768MB. OTOH, I setup a friend's new Dell with the latest PCI express chipset and was really impressed with the speed, so maybe Intel is finally getting it together...
Question is: will the netburst architecture benefit more from dual core than it did from hyperthreading? Its essentially like having an SMP P4 system, so it ought to.
Citizens will be running in the streets in panic if they're pop-up ads stop working - Oh the humanity!!
I've found the linux kernel's built-in RAID capabilities more than adequate for most of my fault tolerance needs. The best part is I can move the drives to pretty much any system - a new motherboard, whatever - without having to worry about kernel support or finding that IDE driver. If a drive fails I can boot its mirror up in any system and be in great shape. I also use the utility mdadm to email me if one of the drives fails. For some linux firewall systems I've built, I use old crappy 6GB drives, but mirror them so there's no risk if one of them goes out. Looking at my basement firewall now and...
everything is cool!
I know that Opteron uses an advanced (compared to Intel's Xeon) hypertransport bus to get up to 8-way systems rather easily and with plenty of memory bandwidth for each CPU.
So what kind of system board hacks would be required to get Opteron up to 32-way? How does SPARC accomplish it?
This is wrong. Laws 2 and 3 are out of order (as they are in the movie trailer). The 3 Laws are:
1: A robot may not injure a human being or, through inaction, allow a human being to come to harm.
2 A robot must protect its own existence as long as such protection does not conflict with the First Law.
3 A robot must obey orders given it by human beings except where such orders would conflict with the First or Second Laws.
So basically, a human cannot order a robot to its doom.
Doesn't it suck when beaucrats make these big research decisions, wasting millions in research dollars? If resonable scientists had been involved in the decicion-making process they wouldn't come within a football field of this project. Frankly, I'd rather they spend the money on in-flight 3D games for the passengers...
The number one problem with "locally built" low cost PCs: They're cheap. There's no way around it. The advantage OEMs have is that they can put quality tested components together at prices impossible for normal folks to match. PCs are an economy of scale, just like any other mass produced product. There's no way you can build up a QC infrastructure to match Dell's. Whenever a friend asks me to build them a PC, I ask for his credit card, order him a Dell and then setup all the software on it. I only build PCs for myself or close loved ones, because a temperamental homemade PC can really erode confidence, believe me. I shutter to imagine what would happen if I had dozens of them out there.
...When the CPU speeds finally hit 1GHz, I guess. The 133Mhz SDRAM probably won't start to cripple the FSB/CPU until the chips are pushing the same speeds as Intel's and AMD's.
Now imagine that suddenly the smarthost that your sendmail server forwards all outgoing email to refuses to deliver messages that you and your coworkers send. And its because they're not From: a valid email address hosted by the ISP!!
Verison is taking away a key capability here. It basically means that Verison's users can no longer run their own email servers. The power is now in Verison's hands. Ready to pay for that email hosting service Verison has? You'd better be.