Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
None of those things would explain why the 3.0-3.2/3000+-3200+ level parts ARE available to be configured, and not the faster stuff.
They all might have something to do with an overall delay, but the odd low-speeds-only availability is the more puzzling thing, IMO.
I can see how that would explain the long build times, but not how it would explain NOT OFFERING high speed parts from either AMD or Intel *at all*.
Venice 3800+ o/c'd to ~5000+ (3.2GHz)
http://www.xtremesystems.org/forums/showthread.php?t=56313&page=12&pp=25
The delay is across all lines, but you cannot even select fast Intel or AMD processors at this point. As I said, sales rep claims it will soon be remedied, but his explanation ("stock issue") doesn't ring true.
Can't be hard drives or ram, as that wouldn't explain low frequency CPUs being available, and not high-frequency parts, across both Intel and AMD lines.
Power supplies? Hmmm. Still seems unlikely...
Perhaps there's a holdup with Intel high-frequency parts, and due to another "contractual issue" HP won't offer high-frequency AMD parts until Intel can supply theirs?
Talked to HP Sales rep. He knew of the issue, and said they had a "temporary stock issue" with parts for their desktop lines. He didn't know exactly when it would be fixed, but said it should be cleared up "very soon".
Not sure how they could have found themselves with a stock issue across both Intel and AMD product lines...but I can't think of why they would intentionally do this either.
It's both Intel and AMD products, and this issue was noted more than a week ago, over at SI:
http://www.siliconinvestor.com/readmsg.aspx?msgid=21116923
As for a possible reason... perhaps HP is refreshing BOTH their Intel and AMD lines? Certainly, AMD is moving all desktop parts to 90nm Rev E over the next month. Intel appears to have fewer changes going on over that time period.
It is hard to imagine that they would not notice a bug in hpshopping.com for over a week.
after every Itanium cancellation, you say it was expected.
Bingo!
Fujitsu-Siemens is NOT a "pure Intel shop".
http://www.fujitsu-siemens.com/sme/products/deskbound/workstations/celsius_v.html
I'll give you an 'A' for Fanaticism, but unfortunately you flunk at Debate Skills.
Given your track record with the facts, you'll have to forgive me for discounting your pronouncements. :)
LOL! I see, when wbmw is wrong, being right is beside the point. Good one.
so you would have to assume that many of the discussed features will be going into Athlon 64 and Opteron.
No, one would assume that those features (different transistors, layout tweaks) are in *Low Power* mobile A64 and *Low Power* Opteron, but not in desktop varieties which need to clock higher, and can afford more power dissipation.
So what's the possible benefit of a different layout for Turion?
Making maximal use of transistor-based power savings tweaks often *requires* changing the layout.
What is AMD likely to gain that justifies the greater costs of new and future layout teams?
Lower power, more competitive mobile parts.
I sense your "point" is about to morph again, now that your "same transistor" speculation has been shown to be...
Wrong!
By DON CLARK
Staff Reporter of THE WALL STREET JOURNAL
January 7, 2005; Page B3
LAS VEGAS -- Advanced Micro Devices Inc., in another attack on its giant rival, is taking aim at Intel Corp.'s high-profile Centrino chip brand for mobile computers.
AMD announced that it will soon begin selling a line of chips called Turion, which will be its first microprocessors to be designed from scratch for mobile computers. Up to now, AMD has adapted chips for desktop computers for the mobile market.
That strategy has left AMD at a disadvantage with Intel, whose Pentium M microprocessor was developed specifically for lightweight laptop computers. That microprocessor -- the heart of what Intel calls its Centrino technology "platform" -- has become popular for its combination of high performance and low power consumption, which extends laptop battery life.
......
AMD, of Sunnyvale, Calif., hopes to make up for lost time. Marty Seyer, a vice president of AMD's microprocessor unit, said Turion will be based on a new transistor structure that will bring power savings and other benefits. While AMD plans to market the brand aggressively, he said there are no plans to emulate Intel's requirement that PC makers buy two kinds of chips in addition to the Pentium M to qualify for the Centrino logo and Intel marketing funds.
"Turion is the anti-Centrino," Mr. Seyer said.
An Intel spokesman declined to comment.
AMD, which announced the brand at the Consumer Electronics Show here, isn't disclosing many more specifics about the chips, except that they will be delivered in the first half. Like some other AMD chips -- and unlike Intel's desktop and laptop chips -- Turion chips will be able to process 64 bits of data at a time, which allows chips to tap into more memory than 32-bit chips. The impact of that advantage has been slight so far, because Microsoft Corp. isn't expected to deliver a 64-bit version of the Windows operating system until later this year.
I don't believe AMD has the resources to staff layout teams to update the layouts of all their different brands each time they do a stepping or process change. I think Turion shares the same layout - and thus the same transistors - as E-step Opteron and Athlon 64 parts,
You're completely wrong! What you "believe" is at odds with reality.
Duh.
From AMD:
AMD, of Sunnyvale, Calif., hopes to make up for lost time. Marty Seyer, a vice president of AMD's microprocessor unit, said Turion will be based on a new transistor structure that will bring power savings and other benefits. While AMD plans to market the brand aggressively, he said there are no plans to emulate Intel's requirement that PC makers buy two kinds of chips in addition to the Pentium M to qualify for the Centrino logo and Intel marketing funds.
wbmw... wrong again!
No, I mean like IBM dumping Itanic after HP canned Itanic workstations and dumped its IPF engineers on Intel.
Translation: No.
I've noticed that he's most unruly in the weeks after an OEM (HP, IBM, etc) drops a "surprise" on the future of IPF. <ggg>
Rink, don't you realize, that was apparently (as of today, anyway) his "point" all along! LOL.
Wow, your "point" sure does evolve quickly when confronted with the facts: <ggg>
"There are no differences except for fuse-enabled stuff"
...
"As I was saying, <cough> except for process differences that lower the power and other critical path work which together completely change the power and speed bin yields <cough>, there are no differences."
Ever considered a career in politics?
"Why do people want Advance Micro Devices (AMD:NYSE - commentary - research) when Intel (INTC:Nasdaq - commentary - research) has proven again and again to be the best bet?"
Gosh, I don't know, maybe because that isn't true?
http://finance.yahoo.com/q/bc?s=AMD&t=5y&l=off&z=l&q=l&c=INTC
that 2.2GHz speeds at 25W is probably not a high volume proposition.
As a top bin for Q3 at that power rating, probably not as high volume as lower speed bins, but what is your point? It's not like the forthcoming 2.26GHz Dothans (nor the existing 2.13GHz parts) are a high volume proposition either, and those use more power!
Reading comprehension difficulties?
not as thin as desktop process target
In fact, the whole presentation is about the differences that were used to make Low Power K8s.
Turion continues with these process differences.
So your statement that all K8s are the same, and the implication that low power parts are somehow cherry-picked from the same parts that might be used for the desktop performance segment is...
Wrong! (again!)
and I've already made my point.
Indeed you did, but your "point" was incorrect!
Kaet,
so he's a hypopcrite when it comes to AMD but a leader when his articles are about Itanium...
Nobody's perfect. :)
To be fair, it's Mr. Reynolds who issued the unfair quote, but it is still Ashlee's responsibility to check these things when his name is on the article.
Hypocrisy haunts The Register's article.
Why is AMD being so coy?
"The answer is that the battery life isn't so good," Reynolds said, adding that Turion-powered systems could have up to one-third less battery life than laptops running on Intel's ultra low voltage products.
Excuse me? An article complaining about fairness, and the importance of apples-to-apples compares the battery life of Turion laptops to ULTRA LOW VOLTAGE P-Ms that run at 1.0 - 1.2 GHz with roughly half the performance???
The objection regarding performance comparisons using discrete vs. integrated graphics solutions was valid, but their own apples-to-oranges battery life comparisons are equally objectionable.
Thanks for that link, proving wbmw wrong:
•Thicker gate oxides (GOX)
– Reduced static gate leakage caused by tunneling
– Reduced gate capacitance
– More reliable (Great for servers)
• Longer nominal channel length reduces static leakage
• Dual Gate Oxide : two gate oxide thicknesses on die
– Thinner GOX for core transistors – not as thin as desktop process target
– Thicker GOX for on-chip decoupling capacitors – reduces gate leakage
– Thicker GOX transistors in I/O power domain
• Three threshold voltages
– High Vt transistor has low leakage, low performance
– Medium Vt has higher leakage and 10% higher drive current
– Low Vt has very high leakage and another 10% higher drive current
– Vt levels (and therefore leakage) are targeted for low power
It really is amazing that he claims all A64s are the same, but for fuses.
Nope. And Yonah has SSE3, but no iAMD64.
Features such as deep sleep modes aren't "transistor properties", you silly boy, and they most certainly can be fused.
Nice try.
Except I didn't say the transistor changes were to enable different sleep modes.
Slower, less leaky transistors can be swapped for faster, leakier transistors, if one wants to trade clockspeed for power usage, as is the case with Turion.
Hence, your statement that the only differences between Turion and regular A64 are controlled by fuses is... (wait for it...)
Wrong! (again!) <ggg>
Wrong. AMD isn't diverging on two different cores, Doug. They are putting these enhancements into the silicon and only enabling certain ones in Turion through one-time programmable fuses. Otherwise, Turion == Athlon 64.
LOL!!!
Yeah, right! Wow, imagine changing transistor properties through a fuse!
You really need to stop embarrassing yourself with this stuff.
From the point of view of optimizing *code*, and code-level features, yes, the architectures are the same, which is a great advantage that Intel lacks. With AMD, you still get SSE3 and AMD64 on Turion. With Intel, oops, sorry, can't do that.
Acer is replacing their already Athlon 64 based Ferrari laptops with Turion based ones. It's hardly anything to get excited about.
They are? The 5000 and 5020 are new lines. Do you have a link saying the Ferrari branded laptops will be discontinued?
Acer seems to think their new Turion laptops outperform their Centrino models. <ggg>
You're right, and wbmw is wrong (again), of course.
It isn't just a Rev E part, either. It's had substantial changes to support lower power, lower power sleep states, a new mode to allow lower power access to system memory for integrated graphics (while the processor remains asleep), etc.
It makes wbmw uncomfortable to contemplate this, so you get the usual misinformation and denials.
I guess you must know better than the Acer spokesman, then. <ggg>
And what makes you think they can manufacture more than one?
The fact that Acer is demoing two models that run these ML-40s, and listing them in the specs.
LOL, Turion is vapor. And then after it is introduced
some day it will be silicon waiting on the shelf for
more than token OEM support.
Paul, you've confused Turion with Montecito.
I looks to me that AMD is attempting an entrance into a market that they aren't truly capable of competing in.
LOL!!! Yeah, right!
before laptops based on the first set of parts are due to hit the market.
Wrong again! First laptops hit the market by the end of March.
It isn't 400MHz in 4 months. It's 200MHz more than the 2.2GHz Turion notebooks on display at CeBIT *right now*. Even those will outperform the 2.26GHz Pentium M.
Especially on 64-bit software :)
It has the ceramic forced-air furnace market pretty much to itself. :)
As for the Turion threat, consider AMD's highest end MT sku, 1.8GHz. The Pentium M low end starts at 1.7GHz, and goes up to 2.26GHz. Even if you are optimistic and hope that AMD's 35W cores can compete perfectly fine with Pentium M, then we are still talking a top end sku of 2.0GHz for the Turion ML.
You are comparing the Q3 PM lineup to the Q1 Turion lineup!
(Why am I not surprised?)
The ML-40 (2.2GHz 35W) Turion arrives in a month or two (already on display at CeBIT), along with the MT-37 (2.0GHz 25W).
In Q3, the quarter in which the 2.26GHz PM appears, the ML-42 (2.4GHz 35W) and MT-40 (2.2GHz 25W) arrive.
Is that what happened to you? <ggg>
Well it has only been 2 months since the prototype was announced. Let's see how it's doing in 6 months.
design win made more than a year ago
???
BTW, I spelled it wrong; 'Vassili' is correct.
Here's some more information: http://www.google.com/search?q=FIC+vassili