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.
Chipguy
So P-M has thermal protection functions. Your point is?
I'd rather call it a parametric design feature.
Apologies for a typo in the other part. Are you suggesting Pentium-M has lower diecost than Pentium-4 (?) was what I intended to ask.
K.
chipguy
Timna, e.g.
As far as parametric headroom is concerned, here is an interesting read: http://www.digit-life.com/articles2/cpu/intel-thermal-features-pm.html
As far as diesize is concerned, are you suggesting Pentium-M has lower diesize than Pentium-4?
K.
bobs
Israel design team is known for top-notch designs. However, designs coming from these chaps are also famous for its challenges to fab these.
K.
p.s. By the way, Intel is not fabbing any of its IA-products in Israel, currently.
Joe
Well I was leaning towards understanding tightness rather as "supply management" until Intels quarterly report as well. However, as I understand their latest numbers and statements, Intel is tight indeed, not only in chipsets but in CPU as well - Pentium-M in particular. (It is certainly unusual to see inventories decline in Q2, e.g.).
K.
Joe
You get surprised once, then you overcompensate that it doesn't happen again.
If you can. Maybe the implied assumption of Intels flexibility in terms of wafer assignments between products (processes) should be questioned at this point?
K.
OT mmoy Thanks, i will give it a shot.
K.
OT mmoy
I thought of a combination of key (eg ALT) plus click (eg right-click) on a term opening a browserwindow showing search-engine results for the term.
If there is more folks out there as lazy as i am it might exist. If not, it should, at least. However I did not find it instantly. And hoped I could save the time as well to search for it because you are familiar with many ff-extensions.
K.
OT mmoy
Is there a ff-extension for "key and click" search (ideally with choices for searchengine(s) ) ?
K.
Talking about funny headlines
Intel Could Be Hampered By More-Competitive AMD
http://www.forbes.com/markets/2005/07/20/intel-amd-earnings-0720markets01.html?partner=yahootix
K.
mmoy
That 24 inch monitor is now on sale for $899 at Dell. I'd rather see Dell selling a complete line of Athlon 64 computers and servers.
I agree, in a more general sense without emphasis on Dell: Compelling pricepoints for TFT displays will soak up a lot of spending power.
K.
b2l
Some people value diversity
Count me in. And keep posting.
K.
mmoy
I think that AMD did spend some money on improving GCC which is kind of the cheap way out.
AMD does a lot more than this. Providing math libraries, code optimization software e.g.. Supporting pathscale, and Portland Group Compilers, beyond the ones you mentioned. But still, ISV usually use Intels Compilers. Not only because the Compilers are good for getting suitable code for the vast majority of machines but also for its support, which is reportedly outstanding.
K.
Keith
September, I've been told.
K.
Doug
...which they weren't.
Your alias is derived from dogma, right? :)
K.
upndown
It's dualcore lithography, 2x512KB L2. For X2-products.
K.
Doug
That is what you are seeing. . I see Manchester E6 qualified, as expected:
http://www.investorshub.com/boards/read_msg.asp?message_id=6888923
Next is few months pilot mode of this stepping until yields justify volume ramp.
K.
Doug
Tx. Links work. As for your question, no I don't accuse monarch of lying. They do specify their products to their best knowledge.
Which might or might not be accurate. In this case, monarch is covered very well: AMD does specify its products so as well.
http://www.amdcompare.com/us-en/desktop/Default.aspx
Which, again, I don't consider as a lie. It's just to avoid any unnecessary confusion. Specifying a product with the lowest possible stepping it can have is allright imo.
The few out there familiar with feature-yield products are able to connect the nodes anyway.
K.
doug
Each of your links bring this message up:
Fatal Error
Miva Merchant has encountered a fatal error and is unable to continue. The following information may assist you in determining the cause of the error:
Error Code: MER-DTB-00072
Description: Product '1..' not found
I think this overdoes it a tad. While i believe it is an error, I would not consider it as a fatal one. :)
K.
Doug
Oups. I scrolled down for the link but it didn't appear.
K.
mountain_dazzie
If 4200+ and 4600+ were from "development wavers", how comes they are generally much better available than "mass produced" 4400+ and 4800+?
These are not from development wafers, but lower-feature-yields from Toledo-dies. Toledo is still in a pretty juvenile process-stage, there is a lot of lower feature yields coming out. Beyond this, full cache parts are needed for Opterons as well. In essence, it does not come at any surprise the parts you mentioned are better available.
K.
doug
BH-E4 (Manchester) and JH-E6 (Toledo)
Yup, I'd guess that makes it pretty clear. Manchester ramping for volume as soon as it reaches E-6 stepping and yield justifies ramping. And a handful E-4-steppings from development wafers to show up at the usual suspect sites to confirm capacity calculations of less experienced Semi-Bunnies.
K.
doug
It's just history repeating itself...
Remember our Newcastle-discussion?
Now, it's been a while, so I can tell the tale out of school about it now. Longtime after our discussion here I talked to someone about manufacturing models and mentioned I was caught by surprise to see the physical 512KB L2 part as early as it appeared. He just smiled silently and raised his hand, opening it. The penny dropped instantly.
K.
doug
I must have missed the 27th update.
Btw, there is an update today as well:
http://www.siliconinvestor.com/readmsg.aspx?msgid=21479110
The date I see is still Juni 22nd.
But well, irrelevant.
K.
doug
You know pretty well I do not share the assumption Manchester is in volume already, don't you?
K.
Keith
SUN still doesn´t offer any 2xx DC Opteron configurations
AMD still doesn't offer any 2xx DC Opterons, officially.
At least these are not on its pricelist effective 6/22.
K.
Keith
While too late for back-to-school line-ups
Why is that? It does not need a new design...
The question raised with addition of X2 3800+ by august is more whether back-to-school line-ups of X2-systems are altered or expanded...
Interesting model x2 4000+ seems to be omitted. Now if i might borrow niceguys glasses for a moment i'd say x65-Opterons are selling so well that there is not enough fullcache parts left for launching X2 4000+.
K.
avatar
Circuit City ad says "Closeout. No rainchecks". Could Acer be closing out this Sempron model (with unadvertised upgrades) to make room for Turion?
Could be. However, summer has always been closeout-time. The designs of the new cycle are coming in for back to school.
K.
deleted - obsolete
Cantonese
Thanks for posting it. I believe AMD would not have mentioned customers in its complaint without their blessings. Consequently what the Taiwanese Press is speculating about has probably happened already.
K.
mmoy
Yes. The thingy is six years old.
K.
Tenchu
Afaik AMD uses a PR-agency. At least it did two years ago. Not a clue to whom my kudos go insofar. I did not say its brillant btw. Its just a part of an AMD marketing process i look at as state-of-the-art. Which is, as you say, noteworthy. While I am at it, I see other parts working as well. Henri does an excellent job leading AMDs brands, controlling supply and prices. Noteworthy as well, considering what happened to the XP brand two years ago. Credit where credit is due. You might know I recently ranted here and there about what I considered blunder coming from AMD in this respect, musing about AMD being two generations back versus Intel and such.
K.
upndown
I believe Andand's take ist right on the spot AMD is aiming for in the public - an in jury, if it comes to trial (which I actually don't expect it will).
So, kudos to AMD PR-folks involved in the campaign. What I see them delivering here is clearly on eye's height with the works coming from Santa Clara. AMD seems to have migrated this part of its marketing process to the next node.
K.
mmoy
Many thanks for the elaborations. Widening the focus from just hardware to the ecosystem-level is much appreciated.
I had this in mind when I answered to Chipguy but decided to stay in the frame of his statement, and as well when answering to Rink for the period he was looking at. I believe Longhorn is a different ballpark in terms of threaded apps and 64-bit apps anyway. But then, it is unclear if it will be released by end of next year, so I left this out as well.
K.
p.s: Just curious, do you make use of AMD support for the 64-bit port of FF? If so, are the tools provided helpful? Documentation? Personal support?
Chipguy
I neither consider Pentium M nor Hammer as standing targets. I guess Merom better has some goodies to make up for latency penalty of its memory architecture vs K8.
K.
p.s: Btw, how would the architecture Dothan, Banias and Pentium-3 is derived from be referenced versus "Netburst"?
Rink
IPC and Clockspeed might be pretty equal by H2/06 imo. Give or take couple of points plus or minus in this or that benchmark.
Power consumption in mobile devices, well operating at two cores at comparable performance points, not much difference as well. Now Merom comes with power features of Yonah, i.e. one core is shut off for mobile operation. X2 has not currently, but AMD announced it already. It is not clear if it will be implemented in the design for 90nm/300 (i.e. on new sockets) already or only in 65nm designs (which I don't expect to be available in the period you asked about). I am leaning at the latter, but I do not expect miracles from Merom in terms of mobile power consumption features. Even if one core is shut off, the other is consuming power, at yet unclear characteristics for idle, which is most determining figure for CPU-powerdraw. I would not be completely suprised if X2 and Merom would not be much different at all in this respect (considering it is leakage-determined).
K.
Rink
do you think I'm far off thinking a shrink of X2 is all it might take to get AMD a worthy competitor for Merom? Like between a quarter and half a year after 65nm introduction?
In performance terms, probably yes.
In terms of manufacturing cost, most probably no.
K.
Tenchu
Souvereign posting in many respects. Hat off.
K.
Kate
got all A's but then there was the partying and boys :) and thought business an easier major ;)
Good choice. The suit btw is formally law, but actually it is plain business. Form and content differ very often, you know.
Regards
K.