backtop


Print 26 comment(s) - last by Viditor.. on Oct 24 at 10:25 PM


PowerEdge 1435SC (Click to enlarge)
AMD says "Dude, you're getting a Dell!"

Today at Oracle OpenWorld, AMD Chairman and CEO Hector Ruiz delivered the opening day keynote address, discussing how the IT industry has entered a new era of choice, and the impacts that will result from this industry phenomenon.

“This era of greater, real choice will benefit one audience in particular: customers,” said Ruiz. “With greater choice comes greater competition. With greater competition comes greater innovation. As IT choices multiply throughout the next decade, customers will see an acceleration of innovation unlike any they have experienced. The future of innovation will rest far more on the choices customers – not their vendors – make.”

Ruiz then welcomed several guests to the stage, one of them being Michael Dell. Ruiz began to roll a video to announce a new AMD partnership, but Dell switched off the video and told all attendees that they would have the news at Dell's speech in the afternoon.

As predicted, Dell unveiled the PowerEdge 6950 and PowerEdge SC1435, both powered by AMD Opteron processors.

The Dell PowerEdge 6950 features quad Opterons running from 2.0 GHz to 2.8 GHz and the PowerEdge SC1435 features dual Opterons ranging from 1.8 GHz to 2.8 GHz. Both PowerEdge servers' 2.8 GHz variants are "SE" Opterons, which were proeviously only available in servers from Sun Microsystems.

“The PowerEdge 6950 and PowerEdge SC1435 continue Dell’s unwavering commitment to meet customer needs and lead the industry in price-performance and performance per watt,” said Brad Anderson, senior vice president, Dell Product Group.

The PowerEdge 6950 sits above the Intel-powered 6850 and provides superior performance with a 20 percent less power consumption, according to Dell's news release. Also stated is that the PowerEdge SC1435 replaces the Intel-based SC1425, but both were listed on the Dell website at time of writing.

Available now worldwide, the PowerEdge 6950 and PowerEdge SC1435 with dual-core, next-generation AMD Opteron processors are priced from $6,499 and $1,299, respectively.

This is the latest design win for AMD in the recent string of agreements with Dell. Earlier in the month, Alienware and Dell introduced AMD LIVE! systems, and in September Dell brought AMD chips to its Dimension desktops.



Comments     Threshold


This article is over a month old, voting and posting comments is disabled

AMD's output
By Desslok on 10/23/2006 7:56:03 PM , Rating: 2
I sure hope AMD can supply all that Dell wants and still provide other suppliers with chips. I know Dell beat AMD down on their prices so their margins have to be razor thin selling to Dell.




RE: AMD's output
By Viditor on 10/23/2006 11:58:34 PM , Rating: 2
quote:
I know Dell beat AMD down on their prices so their margins have to be razor thin selling to Dell

Ummm...how do you "know" this?
I tend to doubt this, but only because AMD said it wasn't so at their last CC...


RE: AMD's output
By JNo on 10/24/2006 5:21:37 AM , Rating: 1
Viditor, we all know and respect your long standing support (plus stake in?) AMD and I too have bought and sold shares. You're right, no-one but Dell and AMD insiders know the prices. However, you'd be crazy to think that Dell, (DELL!!), giant that it is with all its buying power, didn't take AMD to the line on prices, especially given the prices they can probably source Xeons at (er.. why do you think Intel sole supplier for so long?). I don't have to "know" this, it would just be conservative to expect this. Sure, probably still worth it for AMD as they get the volume and the exposure. Confess that the AMD shares look tempting again at $20, but with Intel on the rise in the near-term technologically and intense pricing competition too, it's a tough call...


RE: AMD's output
By Viditor on 10/24/2006 10:16:48 AM , Rating: 5
quote:
However, you'd be crazy to think that Dell, (DELL!!), giant that it is with all its buying power, didn't take AMD to the line on prices, especially given the prices they can probably source Xeons at (er.. why do you think Intel sole supplier for so long?)


I do understand the sentiment (and yes I have both AMD and Intel shares at the moment, though more AMD).
However, to understand why Dell would seek out AMD more on AMD's terms, you have to look back over the last 3 years and see what HP has done to Dell's marketshare because they had both lines.
HP has been wiping the floor on servers with not only Dell, but they've taken a huge chunk of marketshare from IBM...strictly due to Opteron servers.
In addition, HP shot past Dell this quarter in desktops as well (for the first time), again because they have been the only company to offer a choice.

A few Dell directors have already said that sticking with Intel alone has turned out to be a mistake that's cost them dearly...
We all make mistakes, but Mike Dell doesn't strike me as the kind of guy that repeats them. I'm sure they got a good deal, but I can guarantee that they didn't get any better a deal than HP has (or you would have seen HP start to pull some of their lines rather quickly).
quote:
Confess that the AMD shares look tempting again at $20, but with Intel on the rise in the near-term technologically and intense pricing competition too, it's a tough call...

Fair enough...as a point though, the reason for the present valuation is that AMD missed the Gross Margin numbers, which has analysts nervous.
I would submit to you that it's probably logical that this miss was as low as they will most likely go.

1. The prices have settled down from the disruption of the C2D and the need for Intel to dump their outgoing Netburst chips. I really can't see any more big price drops (except the normal ones and the usual EOL sales), and none are planned by either company (Intel lost more than AMD did on this last "price war", but they needed to bite the bullet).

2. AMD's COGS (Cost of Goods Sold) can now do nothing but drop. They are still less than half 300mm, and 300mm as well as 65nm are both coming online quickly (to be close to 100% sometime in 2007). AMD's first iteration of 65nm is a dumb shrink, which means that they should reduce costs by at least 40%/wafer, and another 40%/wafer for every line changeover to 300mm as well.

3. AMD should continue to gain marketshare through the end of this year and on through Q1, with Intel gaining back in Q2 (which is when the supply of C2D hits critical mass).

4. Rev B (K8L) is to be shipped starting in Q2, and it's performance is a big unknown...from the specs, it could be superior to, equal to, or almost as good as C2D.

5. 45nm for Intel begins production in Q2/Q3 for shipping in Q1 08 (according to the foils from IDC and other comments). 45nm for AMD begins shipping in mid 08 according to AMD (~6 months later). However, remember that 45nm is not an architectual change...for comparison, look at the performance change between 90nm and 65nm Pentium D. The temp and power dropped, but performance didn't go up much at all...


RE: AMD's output
By Viditor on 10/24/2006 11:48:49 AM , Rating: 2
BTW JNo, there was no direspect intended in my reply...
What you posted is an extemely common perception, and even some analysts can't shake it. Of course it's those perceptions (and those analysts) that have helped me make money for the last several years... :)
Buy low, sell high...;)


RE: AMD's output
By Desslok on 10/24/2006 12:31:27 PM , Rating: 2
Viditor-On your point about the 40% reduction in cost by going to 300mm and the 65nm shrink. It is not always the case that those savings come right away. From personal experiance I know that in the rush to get the new technology correct cost savings is usally the first thing to be ignored.

I hope I am wrong about this and AMD can give the C2D a run for it's money. The industry needs competition.


RE: AMD's output
By slayerized on 10/24/2006 5:25:29 PM , Rating: 2
The yield ramping process when a company transitions into a new technology is something that needs to be considered. Granted, that a new technology node pontentially yields more dies/wafer and also 200-300 transition does the same..but how mature will the 65nm process be when they roll out chips...will they be able to meet demands? time is the answer. Time has shown that Intel clearly ramped up the 65nm processes extremely fast; can AMD do the same?? Lets wait and watch! and hey the fabs dont come free either..so lets factor all these into account.


RE: AMD's output
By Viditor on 10/24/2006 8:14:31 PM , Rating: 2
quote:
On your point about the 40% reduction in cost by going to 300mm and the 65nm shrink. It is not always the case that those savings come right away. From personal experiance I know that in the rush to get the new technology correct cost savings is usally the first thing to be ignored


A good point, and I would agree. However in this case I don't think it's going to be relevant...
AMD has had production ready 65nm chips since October 2005 (which is when they first showed them to reporters at the launch of Fab 36).
The reason that AMD have delayed as long as they have on volume production is exactly what you stated in your post...AMD can't afford a low yield process. In addition, the longer they used the 90nm process, the less expensive it is for those chips (because of depreciation on the equipment), though 65nm at mature yields is even cheaper.
AMD has been announcing since Jan that the 65nm release will BEGIN at mature yields for the first time in their (or anyone elses) history...


RE: AMD's output
By DallasTexas on 10/24/2006 8:45:24 PM , Rating: 2
In addition, HP shot past Dell this quarter in desktops as well (for the first time), again because they have been the only company to offer a choice.

Seems to me you are stretching the truth here. Dell's woes are well documented to be much more of an operational issue than because they did not offer AMD processors. I tend to accept a more objective viewpoint than your wild assertion that it was because of Dell not using AMD

A few Dell directors have already said that sticking with Intel alone has turned out to be a mistake that's cost them dearly...
That's understandable. It was a blunder but hardly the reason Dell has lost market share. See latest edition of Fortune magazine and Mr Dell tells all.

Confess that the AMD shares look tempting again at $20, but with Intel on the rise in the near-term technologically and intense pricing competition too, it's a tough call...
I bought AMD shares at $7 and at $20 I think it still looks good. They have good technology and good market penetration for future progress although there products today are comparably mediocre.

Fair enough...as a point though, the reason for the present valuation is that AMD missed the Gross Margin numbers, which has analysts nervous. I would submit to you that it's probably logical that this miss was as low as they will most likely go.
I generally agree only because the 4th quarter is a great quarter for everyone. However, if they don't start cranking out 65nm soon, they are toast.


1. The prices have settled down from the disruption of the C2D and the need for Intel to dump their outgoing Netburst chips. I really can't see any more big price drops (except the normal ones and the usual EOL sales), and none are planned by either company (Intel lost more than AMD did on this last "price war", but they needed to bite the bullet).

Check your news inbox. I think price drops continued since you made your claim.

2. AMD's COGS (Cost of Goods Sold) can now do nothing but drop.
Probably because they have written off their old inventory. When you do that, cost is zero.

They are still less than half 300mm, and 300mm as well as 65nm are both coming online quickly (to be close to 100% sometime in 2007). AMD's first iteration of 65nm is a dumb shrink, which means that they should reduce costs by at least 40%/wafer, and another 40%/wafer for every line changeover to 300mm as well.
Makes sense. 40% comes with a shrink and larger wafers. Nothing new here, but OK.

3. AMD should continue to gain marketshare through the end of this year and on through Q1, with Intel gaining back in Q2 (which is when the supply of C2D hits critical mass).

Maybe so but I wouldn't count the chickens quite yet. Price is king and Netburst is selling at rock bottom prices to help stem the share loss.

4. Rev B (K8L) is to be shipped starting in Q2, and it's performance is a big unknown...from the specs, it could be superior to, equal to, or almost as good as C2D.
I guess that covers all possibilities, right? I would assume that K8L will close the performance gap with C2D. The problem is AMD is now on the "close the gap" roadmap versus Intel. No more leap ahead as far as I can see. I optimize compilers for both companies and trust me, I see far.

5. 45nm for Intel begins production in Q2/Q3 for shipping in Q1 08 (according to the foils from IDC and other comments).
I don't know but starting production in Q2/Q3 and then shipping 6months later seems not right. However, a 2 year cycle seems to be the norm at Intel.

45nm for AMD begins shipping in mid 08 according to AMD (~6 months later). However, remember that 45nm is not an architectual change...for comparison, look at the performance change between 90nm and 65nm Pentium D. The temp and power dropped, but performance didn't go up much at all...
Actually, I disagree. You are underestimating the significance of 45nm. It is not your dad's ordinary litho shrink. The 45nm process brings dramatically new technologies to reality. For starters, low-K Dialectric. This is a 10,000x decline in leakage power improvement alone and tri-state (3D stacked) transistors being at least a 2X shrink in transistor size. If Intel cranks out 45nm early, AMD has a rocky road in late 2007 - 2008.


RE: AMD's output
By Viditor on 10/24/2006 9:29:36 PM , Rating: 2
quote:
I tend to accept a more objective viewpoint than your wild assertion that it was because of Dell not using AMD


If you could be more specific about your OWN wild assertions...like why in your opinion (please quote those "objective viewpoints") did Dell lose so much marketshare?

quote:
However, if they don't start cranking out 65nm soon, they are toast

Well, they've been producing 65nm in volume since July/August for a November release...
quote:
Check your news inbox. I think price drops continued since you made your claim

Read the fine print...the price cuts are on EOL products only...
quote:
Probably because they have written off their old inventory. When you do that, cost is zero

Only Intel has had inventory write-offs...AMD has had none (it's part of the quarterly earnings report, Intel had $100M of write-offs, AMD sold every processor they made and are sold out for the 4th quarter already)
quote:
Price is king and Netburst is selling at rock bottom prices to help stem the share loss

As is AMD, which is why they gained share again this quarter
quote:
No more leap ahead as far as I can see. I optimize compilers for both companies and trust me, I see far

And I do development work for many platforms...the Rev B specs (if they all happen as on paper) will put the core of AMD chips at the same level as C2D, and the lower latency of HT and the ODMC will make them faster than C2D.
quote:
I don't know but starting production in Q2/Q3 and then shipping 6months later seems not right

If you study Fab timings (and I have for years now), you will see that 6 months (2 turns) from initial production to shipping is absolutely the standard...it's also what Intel is publishing.
quote:
For starters, low-K Dialectric. This is a 10,000x decline in leakage power improvement alone and tri-state (3D stacked) transistors being at least a 2X shrink in transistor size

Ummm...Firstly Intel isn't going to low-k, they are moving to high-k metal gates. But again, if you reread what I posted you will note that I said "The temp and power dropped, but performance didn't go up much at all...". Reduction of leakage is exactly that. C2D doesn't perform well because it's 65nm, it performs well because of it's architectual design.


RE: AMD's output
By Viditor on 10/24/2006 10:25:57 PM , Rating: 2
Some corrections and validations...

Both AMD and Intel gained marketshare last quarter, but Intel gained more...
The preliminary numbers from Gartner are:
Overall, AMD grew to 23% from 22%...Intel grew from 73% to 76%.
Server - Intel gained 2%
Desktop - Intel gained almost 6%!
Mobile - AMD gained 4%!

As to the validation part,
"According to Gartner, overall, Dell experienced "weak" growth in desk-based PCs, which offset the solid growth in mobile PCs. HP exceled, with strong demand in the home market, fueled by the offering of consumer AMD-based systems "
http://marketnews.ca/news_detail.asp?nid=2259


RE: AMD's output
By Lifted on 10/24/2006 1:16:22 AM , Rating: 3
And why would AMD care about Dell servers enough to ruin their relationships with HP, IBM, and Sun? HP and IBM each sell way more servers than Dell could ever wish to sell, and I think Sun is currently with a percentage or two of Dell. So really, who cares about Dell servers anyway except the people who are forced to buy their crappy servers from the bean counters.


RE: AMD's output
By GoatMonkey on 10/24/2006 8:34:55 AM , Rating: 2
quote:
I sure hope AMD can supply all that Dell wants

I have no numbers to back this up, but I would bet that having Intel currently doing so well in the desktop market would allow AMD to focus their production on the Opteron chips.


RE: AMD's output
By drebo on 10/24/06, Rating: -1
RE: AMD's output
By GoatMonkey on 10/24/2006 11:35:46 AM , Rating: 1
That's good news. I was assuming that AMD's sales had dropped significantly. Apparently that is not the case.


why go amd?
By Pwnt Soup on 10/24/2006 10:03:06 AM , Rating: 2
i work in a lot of schools and other government buildings from state to county and local village, the one thing i see everywhere...DELL pc's and servers. when the gov is buying all they can get from all branches, thats a lot of desktops and servers. it only makes sense for dell too explore every way too make an extra buck and explore every market nitch. and if amd gets them too a better price point, for that market thats the reason for the adition of amd.




RE: why go amd?
By boobot on 10/24/2006 11:10:48 AM , Rating: 5
Great, our National Defense systems are running on Dell's. IF it fails then we get tech support from India!


hmm
By AppaYipYip on 10/23/06, Rating: -1
RE: hmm
By Russell on 10/23/2006 8:21:01 PM , Rating: 5
Sucks to be you, making such a retarded post.


RE: hmm
By boobot on 10/24/2006 1:18:14 PM , Rating: 1
Sarcasm my lonely one.


umm
By shabby on 10/23/06, Rating: -1
RE: umm
By bob661 on 10/23/2006 11:01:57 PM , Rating: 2
quote:
Has dell not heard of the core2duo?
C2D is not a server processor and Dell already has Xeon's from Intel.


RE: umm
By Dactyl on 10/23/2006 11:52:42 PM , Rating: 2
Woodcrest.


RE: umm
By ObscureCaucasian on 10/24/2006 12:25:58 AM , Rating: 4
Umm if you haven't noticed they are 2 and 4 way servers. C2D and Woodcrest can compete in the single socket, but Hypertransport allows AMD to scale to several sockets much better.


RE: umm
By smitty3268 on 10/24/2006 1:43:08 AM , Rating: 2
Woodcrest still takes the lead in most benchmarks on a 2 way server, but I'd go with the Opteron on the quad sockets.


RE: umm
By GoatMonkey on 10/24/2006 8:42:11 AM , Rating: 2
According to this article...
http://www.xbitlabs.com/news/cpu/display/200610240...
...AMD has until Q3 2007 before Intel has a (probably) quad core MP server product.


"It's okay. The scenarios aren't that clear. But it's good looking. [Steve Jobs] does good design, and [the iPad] is absolutely a good example of that." -- Bill Gates on the Apple iPad

Related Articles













botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki