Intel bribing THG? Is it possible?

[Crashman]

Quit sidetracking, normal markup at a discount seller on a CPU is around $15, can't blame Intel if resellers want to charge a premium.

I didn't argue anything, I simply told the truth in response to Pop's deceptions/misconceptions. AMD re-released the Opteron as the FX-51 to beat the standard P4, Intel added Xeon features to the standard P4 PGA to beat the FX-51. To discredit one without discrediting the other is ludicrous considering the similar lineage.

Here's an argument, since you mentioned the A64 (3200+); AMD should drop the price of this CPU to make it more competitave in the mainstream, they'd sell a lot more of these things if they made them look like "free performance" priced similarly to say, a 2.6C. That would still leave them an adequate profit margin and allow them to gain market share.

<font color=blue>Only a place as big as the internet could be home to a hero as big as Crashman!</font color=blue>
<font color=red>Only a place as big as the internet could be home to an ego as large as Crashman's!</font color=red>

 

[Ncogneto]

Ahh but you made the argument both processors are server processors didn't you? Where is the SMP support? But I digress, the whole argument about the p4EE or the A64 fx being server processors is moot, that much I will give you. Not buying into the whole socket argumetn either.

Here's an argument, since you mentioned the A64 (3200+); AMD should drop the price of this CPU to make it more competitave in the mainstream, they'd sell a lot more of these things if they made them look like "free performance" priced similarly to say, a 2.6C. That would still leave them an adequate profit margin and allow them to gain market share

I cannot be so bold to say that would leave them an adequate profit margin. Ultimatly it is always supply and demand that will determine the actual price. so much has been made about the the fx that people forget the real threat to Intel is the standard A64. You have to remember that it was clocked 200 MHZ slower than the fx. A 200 to 400 MHZ bump while intel struggles to get prescott out the door is going to give Intel fits. Can Intel really afford to sell that many EE's without destroying there margins on the Xeon's?

It's not what they tell you, its what they don't tell you!

 

[Crashman]

Based on what I've read they should be able to produce the A64-3200+ for around $60-80. If they can sell a processor that cost $30 to produce for only $40, I'm certain they could sell one that cost them $60-80 to make for $200 and make a HUGE profit. Of course they can also sell them for over $300 and make twice as much profit. Your argument for supply and demand pricing is accurate, AMD keeps the price up and the supply down to raise demand to the level of supply. If they doubled supply they could reduce price and still equalize demand. Selling 4,000 a month at $100 profit makes them as much money as selling 2,000 a month at $200 profit, but gains them market share. This type of agressive selling of course lends itself to more demand, more production, more market share. AMD did this very thing at the bottom, and the extremely low profit margins (I was hearing of as little as $3 per unit) was driving their profits into the basement (eventually resulting in losses). If they were to KEEP those tactics, but shift focus to the upper mainstream segment, they could be back in black.

<font color=blue>Only a place as big as the internet could be home to a hero as big as Crashman!</font color=blue>
<font color=red>Only a place as big as the internet could be home to an ego as large as Crashman's!</font color=red>

 

[Guest]

>I say performance comes from the on-die mem controller.

A big part of it, yes, I agree. 64 bit addressing doesnt give you extra performance, it gives you extra capabilities (running processes >2 GB), 64 bit registers on the other hand, can in some cases give you a hefty speedup, in theory up to 4x (like 64 bit multiplies on encryption algorithms), in reality those are rare exceptions. Extra registers help, but not all that much, a few percent at most.

>As for the games chew up 1 gig [-peep-] in virtual memory
>big deal I watch my games from 3 years ago chew 1 gig up of
>virtual space.

Thanks for backing up up my argument. By now you should realize 32 bit cpu's will only allow 2 gigs for those games, and according to you, 3 year old games are already using half of that. Sounds like 64 bit support is really going to be a necessity real soon, don't you think ?

> But since I have chewed up 3 gig of virtual space while
>running the Morrow Wind Editor I really dont need to say
>more.

I doubt this. I doubt you are succesfully running the /3GB switch, and actually managing the full theoretical 3 Gigs. What could be true, however, is that you have used over 3 GB of swap space, divided over several apps; but your morrowind editor will not have used more than 2 gigs by itself. It may have come close though, or it may have build in restriction to make sure you dont exceed those 2 GB and/or crash. Once more, sounds like 64 bit is needed sooner rather than later if you want to work with more complex maps. Doesnt matter how much or how little ram you have, 2 GBs physical+virtual memory doesnt always cut it any longer.

= The views stated herein are my personal views, and not necessarily the views of my wife. =

 

[Guest]

Don't forget one thing: the channel and AMD's warehouses are stuffed with Bartons. If they would release <$200 A64's now, who is going to buy those Bartons ? That is the main reason for such slow ramps, you need time to clean the channel of older chips. Intel does the exact same thing; Prescott will be cheaper to produce than Northwood, but just can't afford to price prescott at or below slower northwoods quickly, because no one will buy the millions of NW's still on shelves everywhere. Sure, they could "give them away" for <$100, but that wouldnt do much good to their bottom line either.

= The views stated herein are my personal views, and not necessarily the views of my wife. =

 

[eden]

Where is the SMP support?

They are the same chips, exactly. Probably even have the same bridges. AthlonMPs had different layout on the chip with the bridges and stuff to lock it for MP.
And besides, are you forgetting the 3HT channels the AFX has?
What exactly is even your question about, with SMP support?

You know very well they could likely find a way if not already doable to insert AFXs in Opteron motherboards and they would work.

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=blue><b>This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </b></font color=blue></A> :lol:

 

[eden]

My issue with calling the P4 EE a Xeon is that you don't even know if the core is the same. Does the P4 EE use some ECC or some circuits in it for such?

I recall that the Xeons not only had a different socket, but they did have internal settings on the core to make them SMP.

If the above is true and the P4 does not have Xeon parts, then I feel people are only associating and not comparing realities. P4, added with 2MB L3, suddenly because Intel has a server chip with 2MB, is suddenly a Xeon counterpart? I'm a bit skeptical on that one. The AFX on the other hand is a direct renaming, nothing more nothing less. It's as easy as buying a Nike shoe, removing the logo tags and put in Adidas.
So really I am not entirely sure if we can compare at the moment. I just need these proofs and then I'd be convinced, the P4 EE is a Xeon stripped if not the same core but set on the Socket 478 format.

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=blue><b>This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </b></font color=blue></A> :lol:

 

[Guest]

>My issue with calling the P4 EE a Xeon is that you don't
>even know if the core is the same

It is. The CPU ID string on some (all ?) review samples gives the same identification as Gallatin.

>Does the P4 EE use some ECC or some circuits in it for
>such?

ECC is a chipset issue (at least with the P4 since it doesnt have ondie MC), it has nothing to do with the core as such, except in the case of Hammer, since it has its MC ondie. IF you mean L2/3 ECC, every x86 cpu since the original pentium has had ECC protected cache. No difference there either.

>I recall that the Xeons not only had a different socket,
>but they did have internal settings on the core to make
>them SMP.

Every P4 core is SMP capable, the capability is just disabled (P4) or limited to 2x (Xeon, no MP). This can be done with external bridges like the Athlon MP,or internally through the microcode, but it doesnt change a thing about the silicon. The silicon is SMP. In fact, I read somewhere that the P4EE microcode still has SMP enabled too, intel was in such a hurry to rush the cpu out, they apparently didnt even take the time to exclude SMP capabilities.

> the above is true and the P4 does not have Xeon parts,
>then I feel people are only associating and not comparing
>realities. P4, added with 2MB L3, suddenly because Intel
>has a server chip with 2MB, is suddenly a Xeon counterpart?
>I'm a bit skeptical on that one.

No need to be. The P4EE is a repackaged Gallatin, plain simple. Doing anything else would be nonsense for intel too, why on earth produce two different cores that share the exact same P4 core, and have the exact same ammount of cache ? Especially given the low volume P4EE will reach, it makes no sense. P4EE==XeonMP just a different package, and binned and validated for higher frequencies and FSB's. I don't see what is wrong with that approach either, it makes sense.

= The views stated herein are my personal views, and not necessarily the views of my wife. =

 

[Guest]

>You know very well they could likely find a way if not
>already doable to insert AFXs in Opteron motherboards and
>they would work.

Doh.. the only motherboards that work with the A64FX are indeed opteron boards.. geez.. Athlon FX==Opteron 1xx. Nothing changed except frequency and the name.

Also, Athlon MP==Athlon XP, just some more extensive validation (like for SMP), other frequencies and FSB binning and a few lasered bridges after the validation. Same core, same processor, no secrets there either.

= The views stated herein are my personal views, and not necessarily the views of my wife. =

 

[NiceyUK]

Omid, if you have such strong held views are you really the best person to be the editor of this website? Should this website not be offering unbiased, educated, honest and truthfull content about the hardware on sale today? People come here to read an honest, unbiased review of new hardware and if your so obviously, blatently, against AMD how can your competative articles be trusted?

 

[jihiggs]

why did you bring up this dead thread to say nothing at all?

wpdclan.com cs game server - 69.12.5.119:27015
now featuring (optional) cheating death!

 

[DeEvolution]

<b>Eeeeeeeeeeeeeeeeek!</b>

Zombies! Zombies!

Someone is raising the dead!

I want my mummy!

Eeeeeek!


"Sad is the elephant upon the ice who went to put on his wooly coat only to realize that he left it in his other trunk." - DeEvolution

 

[darko21]

I think that I like this thread. I cannot say that I agree with this thread, but it is funny. 🙂

You guys need to relax more. It is just a thread. Threads do not have to reflect the views of everyone.

If I glanced at a spilt box of tooth picks on the floor, could I tell you how many are in the pile. Not a chance, But then again I don't have to buy my underware at Kmart.