StarDestroyer.Net BBS

http://slashdot.org/article.pl?sid=03/03/15/2059226

That's muuuuuuuuuuuuuuuuuuch better than September. I guess they needed to do this to stave off competition from Intel. Win64 for AMD probably will not be done yet, though Linux would be, so...

More specifically, Opteron will be out in April. Athlon64 may not be out until September.

Apparently x86-64 provides excellent performance even with a basic recompile (a bunch of extra registers) in Counterstrike Server (on the order of 30%) but we'll see how well AMD does. Major companies still will have to adopt it.

I desperately want to see AMD come out with a competent laptop chip, something like Banias. Probably won't happen, though.

Huk! Bintel is falling behind again

It's not too much of a stretch to call this AMD's last gasp. The software industry isn't going to support multiple incompatible 64-bit instruction sets on cost grounds. AMD has never been able to capture enough of the business market to be a viable software platform for non-home apps, so everyone is going to write for IA64 and leave X86-64 in basically the same position as the old Alpha.

Enlightenment wrote:It's not too much of a stretch to call this AMD's last gasp. The software industry isn't going to support multiple incompatible 64-bit instruction sets on cost grounds. AMD has never been able to capture enough of the business market to be a viable software platform for non-home apps, so everyone is going to write for IA64 and leave X86-64 in basically the same position as the old Alpha.

x86-64 natively supports 32-bit, while IA64 does not. The Itanium 2's 32-bit performance blows due to its P5-based 32-bit processor core which sits alongside the 64-bit one. AMD offers a much better solution. Their problem is that their marketing blows. Even if AMD delivers it on time, they still have to gain some sort of industry recognition. So far, they just haven't been able to do that. They had a chance when the Athlon first debuted and began slaughtering the Pentium III, but the initial buzz quickly fizzled out as Intel ramped up the clockrate in the Pentium 4, as well as their marketing exposure.

Durandal wrote:x86-64 natively supports 32-bit, while IA64 does not. The Itanium 2's 32-bit performance blows due to its P5-based 32-bit processor core which sits alongside the 64-bit one. AMD offers a much better solution.

Providing a much better solution for the 'legacy' software is merely going to drag out AMD's funeral a bit further. The industry will be forced onto a 64-bit platform one way or the other; concentrating on the 32-bit codebase while ignoring the IA64 gorilla is merely going to give AMD an increasing share of what is going to be a fast-disappearing market. The only way AMD is going to survive the forced transition to 64-bit is if its chips can run IA64 binaries.

Hameru wrote:Huk! Bintel is falling behind again

We'll see. They have a lead in desktop processors (everyone laughed when P4 was out and despite it's clock speed it was getting trounced - now who's laughing?). They have a lead in mobile procesors (for x86, Banias is untouchable at the memont). They very well may get a lead in 64-bit computing with EPIC over x86-64.

phongn wrote:

Hameru wrote:Huk! Bintel is falling behind again

We'll see. They have a lead in desktop processors (everyone laughed when P4 was out and despite it's clock speed it was getting trounced - now who's laughing?). They have a lead in mobile procesors (for x86, Banias is untouchable at the memont). They very well may get a lead in 64-bit computing with EPIC over x86-64.

True. It seems that Intel makes very good mobile chips. But this whole thing with people semi-seriously competing with Intel was good while it lasted.

What market are these chips for? I dont forsee most business upgrading any machines soon. P2's still rule at many offices. I suppose when those machines finally die, we will see a round of upgrades.

TrailerParkJawa wrote:What market are these chips for?

Outside of serious number crunching applications, the only purpose for 64-bit chips is as a component of Microsoft's medium-term strategy to force people onto Longhorn. Longhorn is the code name for next-generation Windows, which is based heavily on Palladium and .NET software-as-service. Longhorn is not designed to be backwards compatible with existing software and will begin a major shift in computing towards software that is rented on a monthly basis with all user-created data stored on Microsoft-controlled central servers.

Here's an overview of how things will shape up in the next three years.
http://www.aaxnet.com/editor/edit029.html

TrailerParkJawa wrote:What market are these chips for? I dont forsee most business upgrading any machines soon. P2's still rule at many offices. I suppose when those machines finally die, we will see a round of upgrades.

I presume you're talking about AMD64 and Itanium? If so, the market is mostly for enterprise stuff . . . big business. At least for the moment. Eventually everybody is gonna want a 64-bit computer, but for the moment they're targeting business. Unfortunately, they couldn't have picked a worse moment to bring the technology to the market. Right now, the economy sucks and, for the moment, enterprise customers are quite happy to stick with their 32-bit servers.

So, at the moment, it means that the one who gets the most customers is really going to be the one with the better sales pitch. And currently the one with the better sales pitch is Intel with it's newest Itanium chips.

Enlightenment wrote:Outside of serious number crunching applications, the only purpose for 64-bit chips is as a component of Microsoft's medium-term strategy to force people onto Longhorn. Longhorn is the code name for next-generation Windows, which is based heavily on Palladium and .NET software-as-service. Longhorn is not designed to be backwards compatible with existing software and will begin a major shift in computing towards software that is rented on a monthly basis with all user-created data stored on Microsoft-controlled central servers.

Here's an overview of how things will shape up in the next three years.
http://www.aaxnet.com/editor/edit029.html

Er, wouldn't the bandwidth requirements for such a system be hideously prohibitive?

The vast majority of Internet users are still on dialup.

GrandMasterTerwynn wrote:

TrailerParkJawa wrote:What market are these chips for? I dont forsee most business upgrading any machines soon. P2's still rule at many offices. I suppose when those machines finally die, we will see a round of upgrades.

I presume you're talking about AMD64 and Itanium? If so, the market is mostly for enterprise stuff . . . big business. At least for the moment. Eventually everybody is gonna want a 64-bit computer, but for the moment they're targeting business. Unfortunately, they couldn't have picked a worse moment to bring the technology to the market. Right now, the economy sucks and, for the moment, enterprise customers are quite happy to stick with their 32-bit servers.

So, at the moment, it means that the one who gets the most customers is really going to be the one with the better sales pitch. And currently the one with the better sales pitch is Intel with it's newest Itanium chips.

Well AMD chips are cheaper, and that's kind of common knowledge.

Uraniun235 wrote:The vast majority of Internet users are still on dialup.

The emphasis for the .NET grand plan is on businesses at the moment. Most substantial businesses have dedicated broadband connections these days. Adolph Gates will simply wait until the vast majority of Internet users are on broadband before bringing Farenhiet 451 to the home.

There are lots of customers at the business level that cant allow any external net access. ie) banks

I guess the whole .NET and Palladium thing might be a boon for Linux. It will be interesting to see how it turns out.

--- edit

by net access I meant microsoft.

not talking to other banks on secure lines.

Or people will mass copy windows 2000 CDs

MKSheppard wrote:Or people will mass copy windows 2000 CDs

What's the point if you can't get the "newest" software to run on Win2K?

TrailerParkJawa wrote:There are lots of customers at the business level that cant allow any external net access. ie) banks

Microsoft has generously allowed very large insitutions to host their own .NET servers on their internal networks. This move is specifically intended to placate all entities with enough political clout to force Microsoft to play nice.

All other groups that are too small to resist will be given the 'privilage' of having their data held hostage on the Microsoft Hailstorm.

I guess the whole .NET and Palladium thing might be a boon for Linux. It will be interesting to see how it turns out.

No, centralized computing (which is what Palladium and .NET are really about) is going to kill Linux. Centralized computing is an answer to so many problems--corporate whistleblowers, music sharing, unauthorized DVD player software, individual privacy--that it simply will not be allowed to fail. Everyone from music, movie, software and ebook publishers to control-freak PHB and the brownshirts desparately want MS-oriented centralized computing to happen because of the power it will give them to control the market and society. There is no organized opposition at the corporate level and it's simply not possible for the public to stand against such a determined assault consisting of both the government security aparatus and hundreds of billions of dollars of capital in the global economy. If Linux ever poses a threat it will be banned or driven from the marketplace.

Crayz9000 wrote:What's the point if you can't get the "newest" software to run on Win2K?

And what's the point if all available hardware won't run Win2K either due to the transition to 64-bit CPUs or a TCPA chip intended to disalow the execution of unauthorized software?

Enlightenment wrote:

Uraniun235 wrote:The vast majority of Internet users are still on dialup.

The emphasis for the .NET grand plan is on businesses at the moment. Most substantial businesses have dedicated broadband connections these days. Adolph Gates will simply wait until the vast majority of Internet users are on broadband before bringing Farenhiet 451 to the home.

That will be years, if only because there are still wide swaths of land where broadband is simply unavailable, let alone in demand.

Centralized computing is an answer to so many problems--corporate whistleblowers, music sharing, unauthorized DVD player software, individual privacy--that it simply will not be allowed to fail.

Except that the people who do those last three won't buy into the new stuff. They'll be the kind of people who hang on to their pre-Palladium/.NET hardware and software so they can continue to do what they like and want to do.

What's the point if you can't get the "newest" software to run on Win2K?

You run the old software to do what you want to do and simultaneously run the new software on a seperate machine to do what you "need" to do.

There's still an unaddressed issue with bandwidth. .NET works just dandy for dinky shit like Word documents, but what about people working with Photoshop? Or video editing tools? CGI production for Hollywood? You can't seriously tell me that all of that data will be sent over slow internet connections when the day's work is done. Internet servers are a poor substitute indeed for local hard drives.

All other groups that are too small to resist will be given the 'privilage' of having their data held hostage on the Microsoft Hailstorm.

Or, they'll just hunker down and make do with the latest pre-.NET/Palladium software and hardware.

And what's the point if all available hardware won't run Win2K either due to the transition to 64-bit CPUs or a TCPA chip intended to disalow the execution of unauthorized software?

Old hardware. You can still find SIMMs if you look hard enough.

And come on. Aren't there non-US based motherboard manufacturers that would produce non-TCPA boards?

I think AMD should have waited until Microsoft released their 64-bit OS. Linux may support it, but most people use Windows, anyway.

Whereas Intel waits a bit longer, floods the market with it's HT chips.

Shit I hope this doesn't kill AMD. Intel with a monopoly will lead to slower development and horrific pricing.