...that corporate world is not going to wait every year until MacWorld to find out what the product roadmap is. Apple will have to ditch the culture of secrecy (they can keep it for the consumer stuff) over their roadmaps. Corporate buyers need long lead times and intro and dicontinuance notices. And corporate IT wants plenty of notice on technology directions from all their key vendors (partially so they can warn off the ones that are about to make a mistake) so Apple's attitude about this would HAVE to change.
AMD is about to introduce their response to Core2, and it seems quite promising based on the hints AMD has provided. Yes but promising in what vector -just performance or price/performance? Barcelona is a big huge die (costly to make) that AMD has already hinted they want to sell at a premium. Just putting the other K8 architecture onto 65nm isn't going to gain them the kind of across-the-board improvement in all segments that Core 2 architecture will for Intel -which is just now phasing out the last of the old NetBurst crap and bringing the Core technology down into budget chips. (Watch for Core-based Celerons mid year). I don't see that AMD's '07 advances help their competitiveness much in the mid and lower priced clients segments and the halo effect of having the fastest floating point quad core selling for premium price servers is not going to go very far. They'll have to really cut their margins in most catagories and even if Barcelona does sell at a great premium, I predict their average margins go down in 2007 (even though they have been rising due to much larger server component in their product mix)
The biggest problem that this technology has is that it is expensive when compared with a compute cluster, which can scale easily and can be more easily programmed. The main time the cluster won't do better are the instances where each core needs results from other cores so frequently that the overhead in message passing is too high. Surely you're joking? A single box constructed with this processor will be vastly less expensive than a compute cluster. Even modern quad core DPs would still require 10 nodes to equal this. The cluster would be more veristile, but it certainly would not be cheaper than this processor.
IHO, Apple is doing this now for several reasons: 1) Steve likes doing major announcements at MacWorld. 2) Unlike a lot of Apple products that can ship right after announcement with no connection needed to 3rd parties, the iPhone was going to need a bunch of pre-release work with at least one telco (Cingular it appears) organization -which is unlikely to be as good as keeping the iPhone secret as Apple employees are. Steve is bowing to the enevitable leak of iPhone details by Cingular if they had to keep the secret for very long. By announcing so far in advance of the product that nobody but a tight few at Cingular knew about it yet, he insured the first word would come from Apple. 3) The iPhone fight with Cisco is probably also deliberate. Recent reports revealing that Cisco's hold the on the trademark is thin probably emboldened Steve and crew to persue something they have probably been planning to to do anyway. Apple's probably been really unhappy about the proliferation of products staring with "i". They'd really like to get a trademark court to say Apple owns "i". If, due to Cisco's sloppyness, they can get a judgement that makes Cisco go away and gives them ammunition to convince smaller players to back off they may be able to enforce it. My guess is that Apple never really wanted to licence "iPhone" from Cisco (certainly not on the extra non-monetary terms Cisco wanted). They wanted to own it outright.
Given that many full installable distros like Ubuntu require a long and complicated edits to configure connection to a WPA PSK AP, it seems a little optimisitc to assume these live CDs can be handed out with strong confidence that newbies, esp with wireless laptops will be able to boot and connect. I've used about 8 of them (from DSL to Knoppix) and while they're pretty good at detecting wired connections (although some recent wired LAN chipsets still don't get supported right away), I've rarely found wireles, esp with encryption to work without a lot of manual configuration - which is really self-defeating if a RO live CD makes you do it everytime. Much as I wish it were'nt true, Windows (Vista -for example) does an excellent job of finding wireless hdw, spotting an access point and correctly guessing the kind of encryption key the AP is looking for - which is ironic, since there's no such thing as a Windows live CD (except for PE which is really an installer boot environment). I don't believe Microsoft can see a licening model for live CD that works as long as CD/DVDs can be copied. (Maybe live CDs that HAVE to be connected to clock a charge-per-minute of run time:-))
So, when DSL can accurately detect and quickly configure wireless (and still run totally out of 128MB or more of RAM) I'll give out tons of these to newbies.
...and by the time AMD starts 45nm, Intel will be shipping 32nm. AMD isn't going to catch up to Intel, if anything Intel's new 2yr process cycle is going to cement AMD being at least one process gen behind.
There's a certain point in the pre-standard development cycle when you know the range of possible ways the standard can go such that you can implement ANY possible final standard with the same physical radio -but with updated firmware. The trick is not to do it so early that you've guessed wrong and have a radio that cannot be upgraded to the final standard.
One assumes that Intel will have made sure their N implementation is upgradeable to the final IEEE standard.
Mac is just a PC with OS/X instead of Windows
on
Leopard Vs. Vista
·
· Score: 1, Troll
A bit more stylish perhaps, but only major architectural difference at the platform level is the use of EFI boot - which the PCs will get to shortly anyway (it's been around for years and is superior to BIOS, but conservative vendors have held off using it). Apple probably wanted it because (other than being more powerful and better) it gives them some control of what OS goes on Macs and lets them also make it more difficult to run OS/X on non-Mac non-EFI systems.
Is there any chance do you think that this could lead to Microsoft launching a series of micro-SCO type initiatives now that their investment in SCO didn't work out?
I'm not sure the SCO effort didn't work for them. This doesn't get a lot of press, because the companies involved generally don't want to talk publicly about it, there are some big Fortune 100 firms that got badly spooked by the SCO suit and even it goes nowhere, opponents of Linux within the IT groups at these orgs (sometimes Windows factions, sometimes traditional UNIX factions) that feel threatened by Linux have promoted a position (now locked into the executive mindset) that Linux is still very legally encumbered and while the SCO suit went nowhere, somebody else's suit might work out.
Never underestimate the timidity and risk aversion of major corporate execs, who are strong skeptics of the competitive advantages of Linux (over Windows/UNIX/mainframe) anyway.
IBM Negates their own advantage.
on
IBM Opts for AMD
·
· Score: 2, Informative
For 4-way MP, one of the reasons that IBM Xeons (like x366) have been excellent performers is the sophisticated memory controller hub chipset ("X architecture", the 'Hurricane' chipset) which has an advanced L4 cache and snoop filter integrated. What IBM now calls the x3850 (basically same as x366) still has this advantage - and combined with the new Intel 'Tulsa' Xeon MP chips with huge L3 caches, IBM should probably outperform their generic design x3655 Opteron (and other Opteron 4-way boxes). The real reason, I suspect, why IBM is doing this is not so much technical (they have felt their Xeon MPs were competative against Opteron 4-ways) but that there is now an AMD market mindshare which had created a disadvantage in some corporate accounts that had adopted the other AMD 4-ways (generally the HP 585) and were happy with it - IBM felt shutout wherever AMD fanboys became dominiant in IT thinking.
In the 2-way DP space the new IBM x3650 with Xeon 5100 series Woodcrest (and any other Woodcrest based DP) will be as good or better than anything AMD throws at the DP space in 2006 -incl the new 2.8GHz F socket stuff.
Does anyone know why Firefox 1.x, despite having a fairly configurable interface, requires you to have an active extension just to make the Google search field larger? (I know there's ways you can edit the user file to change this, but this is a pain and on many of the machines have to use, I haven't bothered. Why should this be so difficult??) Was this some kind of deal with Google, which G wanted users to go to a generic Google page if they wanted to enter (and SEE) more than 15 characters or so of search text? Drives me crazy - hope 2.0 fixes this...
Unified (for each core pair) 4MB L2 cache is going to help a lot there. Also, these aren't successors to the dual cores they're another choice for thread intensive applications that may not have a high a ratio of memory I/O to compute cycles. Real solutions demand muliple options for architecture, not just a sweeping belief that embedded memory controllers make everything better.
Processor does not determine whether system supports SLI - chipset and GPU video configuration does so you could use Yonah without using the default Intel chipset. For instance, several vendor have Nvidia SLI validated solutions for Intel procesors. Can't think of any Core Duo ones yet, but frankly it will make more sense to do this with it's 64-bit big brother Core 2 Duo (Merom) this summer. Dell has Intel SLI desktops planned for the 'Conroe' dekstop version this summer so there's really no reason they can't do a laptop - probably will. This will be so much faster than the wimpy Turion chip it won't be funny. Since Conroe is faster than the fastest single socket cpu AMD has It's abundantly clear that Conroe will be the fastest CPU in town when it hits in a few weeks' time. It pulverises all present CPUs in the majority of CPU-based benchmarks.
I do not want (nor will I think my telco tolerate for long without squizing me for serious money) to have 100s, 1000s, + of external users sucking up connection bandwidth to my phone. However if my phone had a web interface for connection to my PC that made configuration and data transfer easier, OR if I could limit my webphone to a few key users OR if my phone server was actually proxied by servers at the telco, this might make sense.
The Intel Core Duo T2700 at 2.3Ghz would smoke that Turion. Yeah, it's not 64-bit (Merom's this summer) but Alienware can't put more than 2GB in that box anyway, so addressing more than 4GB is pretty meaningless, huh? And the Core cpu and Intel chipset will get much better battery life (anything with that monster screen needs all the battery life help it can get)
If you must buy Dell check out the new Precision workstations (don't know the number) that will be based on the new Intel microarchitecture Woodcrest processors. These are much faster than the AMD Opterons...
'Why don't they go and tell the oil companies what they should charge for their damn gas?'
Well, if the oil companies decided to charge SOME kinds of cars (ones where they had a financial relationship with the automaker) more for gas than others I suspect that WOULD be illegal (Can't quite decide under what law, though:-) ).
For those of us in countries with free speech clauses or updates to their constitutions (such as the US First Amendment - although IANA-consititutional-L:-) ), I believe this should constitute a violation of free speech rights in the same way that if a telco insisted in penalizing/discounting phone tolls depending on whether you were talking to a commercial partner of that telco. (Eg. You may not talk to their 1-900-Weather service, only ours, without penalty).
Who and why I establish IP connections to them is NONE of the last mile provder's business.
Slashdot Mirror
...that corporate world is not going to wait every year until MacWorld to find out what the product roadmap is.
Apple will have to ditch the culture of secrecy (they can keep it for the consumer stuff) over their roadmaps. Corporate buyers need long lead times and intro and dicontinuance notices. And corporate IT wants plenty of notice on technology directions from all their key vendors (partially so they can warn off the ones that are about to make a mistake) so Apple's attitude about this would HAVE to change.
AMD is about to introduce their response to Core2, and it seems quite promising based on the hints AMD has provided.
Yes but promising in what vector -just performance or price/performance? Barcelona is a big huge die (costly to make) that AMD has already hinted they want to sell at a premium. Just putting the other K8 architecture onto 65nm isn't going to gain them the kind of across-the-board improvement in all segments that Core 2 architecture will for Intel -which is just now phasing out the last of the old NetBurst crap and bringing the Core technology down into budget chips. (Watch for Core-based Celerons mid year).
I don't see that AMD's '07 advances help their competitiveness much in the mid and lower priced clients segments and the halo effect of having the fastest floating point quad core selling for premium price servers is not going to go very far. They'll have to really cut their margins in most catagories and even if Barcelona does sell at a great premium, I predict their average margins go down in 2007 (even though they have been rising due to much larger server component in their product mix)
The biggest problem that this technology has is that it is expensive when compared with a compute cluster, which can scale easily and can be more easily programmed. The main time the cluster won't do better are the instances where each core needs results from other cores so frequently that the overhead in message passing is too high.
Surely you're joking? A single box constructed with this processor will be vastly less expensive than a compute cluster. Even modern quad core DPs would still require 10 nodes to equal this. The cluster would be more veristile, but it certainly would not be cheaper than this processor.
IHO, Apple is doing this now for several reasons:
1) Steve likes doing major announcements at MacWorld.
2) Unlike a lot of Apple products that can ship right after announcement with no connection needed to 3rd parties, the iPhone was going to need a bunch of pre-release work with at least one telco (Cingular it appears) organization -which is unlikely to be as good as keeping the iPhone secret as Apple employees are. Steve is bowing to the enevitable leak of iPhone details by Cingular if they had to keep the secret for very long. By announcing so far in advance of the product that nobody but a tight few at Cingular knew about it yet, he insured the first word would come from Apple.
3) The iPhone fight with Cisco is probably also deliberate. Recent reports revealing that Cisco's hold the on the trademark is thin probably emboldened Steve and crew to persue something they have probably been planning to to do anyway. Apple's probably been really unhappy about the proliferation of products staring with "i". They'd really like to get a trademark court to say Apple owns "i". If, due to Cisco's sloppyness, they can get a judgement that makes Cisco go away and gives them ammunition to convince smaller players to back off they may be able to enforce it. My guess is that Apple never really wanted to licence "iPhone" from Cisco (certainly not on the extra non-monetary terms Cisco wanted). They wanted to own it outright.
Given that many full installable distros like Ubuntu require a long and complicated edits to configure connection to a WPA PSK AP, it seems a little optimisitc to assume these live CDs can be handed out with strong confidence that newbies, esp with wireless laptops will be able to boot and connect. I've used about 8 of them (from DSL to Knoppix) and while they're pretty good at detecting wired connections (although some recent wired LAN chipsets still don't get supported right away), I've rarely found wireles, esp with encryption to work without a lot of manual configuration - which is really self-defeating if a RO live CD makes you do it everytime. Much as I wish it were'nt true, Windows (Vista -for example) does an excellent job of finding wireless hdw, spotting an access point and correctly guessing the kind of encryption key the AP is looking for - which is ironic, since there's no such thing as a Windows live CD (except for PE which is really an installer boot environment). I don't believe Microsoft can see a licening model for live CD that works as long as CD/DVDs can be copied. (Maybe live CDs that HAVE to be connected to clock a charge-per-minute of run time :-))
So, when DSL can accurately detect and quickly configure wireless (and still run totally out of 128MB or more of RAM) I'll give out tons of these to newbies.
...and by the time AMD starts 45nm, Intel will be shipping 32nm. AMD isn't going to catch up to Intel, if anything Intel's new 2yr process cycle is going to cement AMD being at least one process gen behind.
Sorry fanboys, facts are facts.
...It's only been 12 months since they changed it to the new one
There's a certain point in the pre-standard development cycle when you know the range of possible ways the standard can go such that you can implement ANY possible final standard with the same physical radio -but with updated firmware. The trick is not to do it so early that you've guessed wrong and have a radio that cannot be upgraded to the final standard.
One assumes that Intel will have made sure their N implementation is upgradeable to the final IEEE standard.
A bit more stylish perhaps, but only major architectural difference at the platform level is the use of EFI boot - which the PCs will get to shortly anyway (it's been around for years and is superior to BIOS, but conservative vendors have held off using it).
Apple probably wanted it because (other than being more powerful and better) it gives them some control of what OS goes on Macs and lets them also make it more difficult to run OS/X on non-Mac non-EFI systems.
Is there any chance do you think that this could lead to Microsoft launching a series of micro-SCO type initiatives now that their investment in SCO didn't work out?
I'm not sure the SCO effort didn't work for them. This doesn't get a lot of press, because the companies involved generally don't want to talk publicly about it, there are some big Fortune 100 firms that got badly spooked by the SCO suit and even it goes nowhere, opponents of Linux within the IT groups at these orgs (sometimes Windows factions, sometimes traditional UNIX factions) that feel threatened by Linux have promoted a position (now locked into the executive mindset) that Linux is still very legally encumbered and while the SCO suit went nowhere, somebody else's suit might work out.
Never underestimate the timidity and risk aversion of major corporate execs, who are strong skeptics of the competitive advantages of Linux (over Windows/UNIX/mainframe) anyway.
(as above - I'm lazy, so don't mark me down a redundant just because someone else has the same clever observation -have YOU read every post?)
This will still be necessary as long as there are AMD fanboys out there still living in denial.
The real sweat shop is in Cupertino...:-)
For 4-way MP, one of the reasons that IBM Xeons (like x366) have been excellent performers is the sophisticated memory controller hub chipset ("X architecture", the 'Hurricane' chipset) which has an advanced L4 cache and snoop filter integrated. What IBM now calls the x3850 (basically same as x366) still has this advantage - and combined with the new Intel 'Tulsa' Xeon MP chips with huge L3 caches, IBM should probably outperform their generic design x3655 Opteron (and other Opteron 4-way boxes). The real reason, I suspect, why IBM is doing this is not so much technical (they have felt their Xeon MPs were competative against Opteron 4-ways) but that there is now an AMD market mindshare which had created a disadvantage in some corporate accounts that had adopted the other AMD 4-ways (generally the HP 585) and were happy with it - IBM felt shutout wherever AMD fanboys became dominiant in IT thinking.
In the 2-way DP space the new IBM x3650 with Xeon 5100 series Woodcrest (and any other Woodcrest based DP) will be as good or better than anything AMD throws at the DP space in 2006 -incl the new 2.8GHz F socket stuff.
Does anyone know why Firefox 1.x, despite having a fairly configurable interface, requires you to have an active extension just to make the Google search field larger?
(I know there's ways you can edit the user file to change this, but this is a pain and on many of the machines have to use, I haven't bothered. Why should this be so difficult??)
Was this some kind of deal with Google, which G wanted users to go to a generic Google page if they wanted to enter (and SEE) more than 15 characters or so of search text?
Drives me crazy - hope 2.0 fixes this...
Unified (for each core pair) 4MB L2 cache is going to help a lot there. Also, these aren't successors to the dual cores they're another choice for thread intensive applications that may not have a high a ratio of memory I/O to compute cycles. Real solutions demand muliple options for architecture, not just a sweeping belief that embedded memory controllers make everything better.
Processor does not determine whether system supports SLI - chipset and GPU video configuration does so you could use Yonah without using the default Intel chipset. For instance, several vendor have Nvidia SLI validated solutions for Intel procesors. Can't think of any Core Duo ones yet, but frankly it will make more sense to do this with it's 64-bit big brother Core 2 Duo (Merom) this summer. Dell has Intel SLI desktops planned for the 'Conroe' dekstop version this summer so there's really no reason they can't do a laptop - probably will. This will be so much faster than the wimpy Turion chip it won't be funny. Since Conroe is faster than the fastest single socket cpu AMD has It's abundantly clear that Conroe will be the fastest CPU in town when it hits in a few weeks' time. It pulverises all present CPUs in the majority of CPU-based benchmarks.
I do not want (nor will I think my telco tolerate for long without squizing me for serious money) to have 100s, 1000s, + of external users sucking up connection bandwidth to my phone.
However if my phone had a web interface for connection to my PC that made configuration and data transfer easier, OR if I could limit my webphone to a few key users OR if my phone server was actually proxied by servers at the telco, this might make sense.
for server and desktop before AMD...
The Intel Core Duo T2700 at 2.3Ghz would smoke that Turion. Yeah, it's not 64-bit (Merom's this summer) but Alienware can't put more than 2GB in that box anyway, so addressing more than 4GB is pretty meaningless, huh? And the Core cpu and Intel chipset will get much better battery life (anything with that monster screen needs all the battery life help it can get)
And would be absolutely useless on a laptop since ESX requires SCSI.
Sorry, ESX DOES run on IDE drives too (all those dev systems y'know) but SATA, alas is still not here yet.
If you must buy Dell check out the new Precision workstations (don't know the number) that will be based on the new Intel microarchitecture Woodcrest processors. These are much faster than the AMD Opterons...
'Why don't they go and tell the oil companies what they should charge for their damn gas?' :-) ).
Well, if the oil companies decided to charge SOME kinds of cars (ones where they had a financial relationship with the automaker) more for gas than others I suspect that WOULD be illegal (Can't quite decide under what law, though
The one you've got there is over 5 months old. Try this one.
For those of us in countries with free speech clauses or updates to their constitutions (such as the US First Amendment - although IANA-consititutional-L :-) ), I believe this should constitute a violation of free speech rights in the same way that if a telco insisted in penalizing/discounting phone tolls depending on whether you were talking to a commercial partner of that telco. (Eg. You may not talk to their 1-900-Weather service, only ours, without penalty).
Who and why I establish IP connections to them is NONE of the last mile provder's business.