Dan, I don't know where you get your information from. Do you just make it up?
Re: "P4 desktop uses up to 94w"
developer.intel.com
Look at page 63. Intel's Pentium 4 at 2.53GHz (fastest part available) has a TDP of 59.3W, less than 25% more than the 1.6GHz Athlon, but far more powerful.
developer.intel.com
According to this datasheet for the Pentium 4-M (page 86), the 2GHz part (fastest available), has a power dissipation of 32W. It also runs at 1.3V. Using voltage scaling alone, a 2GHz Pentium 4 would dissipate 39.4W, so clearly Intel has done some additional power reduction for their mobile parts (has AMD?).
Re: "Mobile Athlon, available now, uses 25w for 1200mhz to 1533mhz. Mobile Athlon low power, available now, uses 16w, for 900mhz to 1300mhz."
Where do you see these parts? They are not on AMD's web page; there have been no such product announcements; I see no vendors carrying them. Is this simply one of your lies, or do you have a link to show me?
amd.com^609,00.html
amd.com^6308,00.html
amd.com^2987,00.html
amd.com
These links include AMD's price sheet (low power Athlon XP a no-show), AMD's Mobile Athlon XP processor FAQ (no mention of low power parts), AMD's Mobile Athlon XP tech docs (note: there are no tech docs), and AMD's processor press release page (no mention of low power mobile parts).
Since there are no tech docs available, I can only use voltage scaling to estimate AMD's mobile power dissipation, and I come up with 35W, as I showed in my previous post. AMD doesn't seem to mention the power of their mobile processors. Maybe they are embarrassed. Maybe it's even higher than 35W, because AMD's .13u process is still broken. Maybe you need to check up on these things before bringing up processors that don't exist.
Re: "6 months from now, Intel will introduce a brand new architecture, dedicated to mobile applications, that runs 66mhz faster (1600 vs. 1533) than the chip AMD has now... What do you suppose AMD will be shipping in 6 months, when Intel catches up to where AMD is today?"
Look at AMD's current release rate for processors. They released a 1.33GHz Athlon 4 in January, a 1.4GHz Athlon 4 in March, a 1.47GHz Athlon XP April, and a 1.53GHz part in July. That's roughly 2 months between processor releases. Therefore, in 6 months, I expect 3 more releases from AMD, putting them at about 1.73GHz, or a mere 133MHz above the scheduled Banias launch. Given Intel's execution, I might expect a 1.7GHz Banias to be "pulled in" to compete with AMD's slightly faster parts. What I think will be compelling, though, would be the 30% less power dissipation from the Banias parts. Given equal performance, I think lower power is likely to sell in the mobile market, but that's just MHO. I won't try and claim it as fact.
Re: "Intel has nothing on its roadmap that can support thin and light notebooks (16w) - the most lucrative and fastest growing segment. P3 for thin and light topped out 800 or 900 mhz and is being discontinued. Athlon for thin and light starts at 900 and goes to 1300."
More lies, as far as I can see. AMD has no such low power parts, as I've indicated above. If they get some out soon, good for them, since they are likely to be very competitive against the Pentium III-M. However, I expect low power versions of Banias in the H1 2003 time frame to compete with lower power processors from AMD. Unless AMD launches their low power parts in the next 6-12 months (very probable, but so far, there has been no indication that they will), they will have absolutely no opportunity to gain market share in the mobile segment, as far as I can see.
wbmw |
