Clawhammer not 'til 2H03!

[SteveC]

According to The Inquirer, AMD said it's delaying the Clawhammer (desktop) until the second half of next year. The Sledgehammer (server) is still scheduled for the 1st half however. How much is this going to hurt them, and will it give Intel a big advantage?

Steve

 

[blakerwry]

That's too bad. I'm stating to feel AMD slip back to it's former position.

As long as they can keep up quality and cost/benefit (ie: bang for the buck) I will enjoy using their products.

I will certainly buy a AMD over a C3 processor without hesitation. And for desktop use I think AMD is still great, more than adaquate for most people. Their k7, k75, Tbird, and T-bred lines have been excelent and beat Intel's processors into the ground until recently.

I expect that their current technology will lead them past 3.5gHz, but without a new chip to compete with Northwood's potential I think high end graphical/workstations are going to go intel.

I've heard that Dell wants to use AMD in their servers, no doubt to further decrease their prices(and increase their profits). This struck me as odd because Dell states on their commercials that they use Intel Pentium processors... I would think that Dell would have to break some strong ties with intel by offering AMD chips... but of course there is still awhile before Opteron comes out, they could still change their minds.

 

[Tea]

I strongly suspect that the delay, if it eventuates, will be for marketing rather than purely technical reasons. AMD seem to be feeling very upbeat indeed about the 166MHz FSB Athlon XP family, and (in my estimation) are expecting them to last quite a while as the top dog on the desktop. On the other hand, they don't really have a serious entry in server space at all right now, so they have nothing to loose and everything to gain by getting the server chip out ASAP.

It occurs to me that they expect to be entering the server domain with their own, in-house chipset (or are they collaborating with someone like ServerWorks? I haven't been following this), which it's reasonable to suspect, given the excellence and stability of past AMD chipsets, will be pretty much a goer right out of the blocks. On the other hand, it would surprise me not at all to discover that AMD are dubious about the ability of VIA, Nvidia and SiS to deliver really solid, 100% stable new chipsets for the desktop market without a little longer to spend on development. So there might be a synergy there.

Thirdly, it seems that although there have been no tales of quality problems with the new process, volume is ramping up no faster than predicted. Some of their past process changes, notably the one from 0.35 to 0.25u that introduced the K6-266 and took them through the whole K6-II and K6-III family, went better than predicted and they started delivering ahead of schedule. It looks as though the current process change is not doing that. Given that (according to this theory) they are not getting a headstart on production due to a process bonus (like the one that kick-started the K6-II), I guess that's another reason not to hurry too much.

I notice that the Register has been running stories of "delays" and "shortages" of the new Athlons. This is complete bunkum: AMD themselves said way back in August that they were not expecting to ship retail quantities of the new XPs to the general market until November. (This was at an official training session Kristi and I attended.) The Reg simply has it wrong. If we get to mid November and don't see the new parts, then and only then will there be a "delay". Thus far, they have delivered exactly what they said they would deliver.

 

[Prof.Wizard]

That's too bad. I'm stating to feel AMD slip back to it's former position.

It's extremely sad to know, but I agree...
Even if the new Athlons with FSB166 do fare good, the hat trick would have been the ClawHammer...

 

[P5-133XL]

I look at this action with sadness and delight.

It is sad that they can't seem to keep their schedule: If they can release it quickly then They would be able to gain a foothold and market share. It would produce competition with Intel and thereby lower prices, encourage innovation, and give people more choice.

I look at this with delight too. It tells me that AMD is serious about producing something with quality in mind. I would be very disappointed if AMD released it and it had problems or was not competative with Intel. First impressions are very important and if AMD failed then I'm not sure they would be able to recover.

 

[blakerwry]

Do all of us here feel that VIA releases things too quickly?

It seems like their 1st generation of anything gives a few more features but almost no improvement in performance. It's not until the 'A's start rolling that we see performance and or stability.

ex:
kt133 offered a socket solution but had no perf. benefit over kx133. it was not even stable
kt133a offered much needed stability and performance increases

kt266 was initially slower with DDR than the kt133a was with SDR, atleast it was stable.
kt266a allowed VIA to compete with the SIS and AMD 760 chipsets which had been way ahead for many months.

kt333 offers marginal speed improvements over kt266a

kt400 offeres even less to the table
kt400a will hopefully actually show performance benefits from 400mhz DDR

 

[Tea]

Good summary, Blakewry. And if you take it a little further back, there was the original VP3 an the much improved MVP3.

 

[Buck]

Hence, KT266A boards are still good solutions for new systems. Only when KT333 prices are the same as KT266A boards is the upgrade worth it, which has already occurred in some situations.

 

[Jake the Dog]

so going by that theory, the KT400A should be a good chipset then.

 

[Buck]

Nice avatar Jake.

 

[timwhit]

Personally I liked the old one much better.

 

[Tea]

Not at all, Tim. Two nice new avatars from Jake and Buck. I vote approval.

 

[Jake the Dog]

Nice avatar Jake.

likewise Buck. obvisouly you were my inspiration

 

[Buck]

Next we'll see Tony come in as Arthur English.

 

[Tannin]

I did think about it.

 

[SteveC]

From the Inquirer again:
Clarification: AMD asks us to point out that Hammer schedules haven't slipped from its previous advice, as we originally suggested in this article. A spokesman from the company told us that desktop versions of Hammer are still planned to ship (for revenue) in Q1 2003 with systems on shelves at the turn of Q1 2003, not the second half of 2003 as we stated.

It looks like they were wrong, and it's still on schedule. Good news!

Steve

 

[LiamC]

AMD could be being clever with the use of "hammer". Opteron is a hammer chip and is slated for Q1 2003 shipment. But this is a server chip, using 8~12 layer mobo's, dual channel DDR, and a different socket -> Mega$$$. Whilst it qualifies, I don't think this will the Hammer Athlon chip we are hoping for.

 

[SteveC]

In the revised article, they specifically said the Clawhammer would be out towards the end of Q1, and the Opteron in the 1st half. Now, I don't know if these will be paper launches, or if you'll actually be able to buy them at that time.

 

[LiamC]

SteveC, there is no mention of "Claw" anywhere in that article.

In the AMD conference call, it was specifically stated that "Claw" would not be out until 2H '03. In the same CC, AMD stated that they would ship Opteron in Q1 2003.

 

[SteveC]

I found this article at ExtremeTech which says:
On Thursday, an AMD spokeswoman said that only the company's product strategy had shifted, not the company's production timetable. Officially, samples of the 64-bit Athlon desktop processor are scheduled to begin in the first quarter of 2003, with production due for sometime in the first half. However, AMD's schedule calls for 64-bit Athlon systems to be available in either "late Q1 or early Q2", the spokeswoman said, painting a clearer picture of the 64-bit Athlon's availability. Opteron processors for servers and workstations are scheduled to be manufactured in the first half of 2003.

It looks like they will have the Clawhammer out, but it'll probably be in very limited quantities until the second half.

Steve

 

[LiamC]

I hope you are right and I am wrong

 

[SteveC]

You were right, LiamC, and I was wrong. The Clawhammer (Athlon 64) won't be out until September after all. The Sledgehammer chips are to be released on April 22.

 

[Pradeep]

A Barton 3000+ with 512KB cache should keep you going till then. Means that all the bugs in the mobos should be worked out by then two.