You can still overclock Athlon/Duron the same way you do Celeron/PIII by increasing FSB speed.
There is no difference between AMD and Intel in this regard - they BOTH lock the multiplier. While it might have been cool to allow it to be adjusted, there is no way to prevent remarkers from abusing this.
Given that AMD has been selling 900MHz Thunderbirds downbinned to 700MHz, and that Duron appears equally overclockable (same way as Celeron), I'd still rate AMD as very overclocker friendly!
Intel P4 (aka Willamette) has 20 stage pipeline, and it remains to be seen whether the high clock rates this enables makes up for the hits it'll take due to latency and branch mispredict penalty.
AMD's brand new PowerNow! technology is way more advanced than Intel's power/speed levels because it can automatically dynamically change the power and CPU speed depending on processor load! For example, during DVD playback it'll use less power during slow moving scenes that are less CPU intensice to decompress. In practice this can make the difference between having the battery power to see the whole movie or now. I have no idea how they do this, but it works!
Right now PowerNow is K6 only, but later this year there will be a "mobile Duron" chip that combines a Duron variant of the enhanced Corvette core (that will debut in Mustang) with PowerNow!
The buggy Coppermine chipsets and motherboards were all Intel parts, so Intel was to blame. In Gateway's case they are using a Jabil motherboard (using an AMD 750 chipset), and have isolated the problem to either being the motherboard or the power supply. The AMD chipset has been around for a very long time, and has a reputation for great stability - not surprisingly Gateway do not think it is a suspect.
In some cases it's close, but Athlon is ahead of PIII on pretty much everything. On computationally intensive stuff like Photoshop, it's not even close - Athlon scorches PIII (for that matter Duron scorches PIII on floating point too):
The independent review sites such as Anandtech, Toms hardware, Aces, etc all have pretty much the same results, although the details obviously differ based on the exact system configuration.
The only benchmarks I've seen where PIII beats Athlon are memory benchmarks where the PIII is running on an i840 platform with PC800 Rambus memory, or on an overclocked BX platform. For many people these arn't realistic comparison platforms, since PC800 Rambus memory is prohibitively expensive and most people do not overclock their systems.
Anyone looking for an affordable high performance memory solution should wait until Sept/Oct for AMD's 760 chipset which supports DDR. Given Athlons 200/266MHz DDR FSB, the speedup you get with DDR should be much greater than PIII gets with Rambus memory, since PIII is limited by it's 100/133MHz FSB.
This spells it out a bit clearer that both Gateway and AMD agree it's a Gateway issue - either the motherboard or power supply.
Originally Gateway said it's be fixed by July 10th, but that has more recently changed to July 18th. Given the specificity of the date, I assume that they now know what the issue is, and are waiting either for rev'd motherboards from Jabil, or for a shippment of an appropriately spec'd power supply.
I'd assumed it was Gateway's motherboard that was to blame, since I've seen AMD approved motherboards go from 900MHz max at one rev., to 1GHz at a higher rev. (e.g. MSI 6340 1.0 -> 1.0A).
I'd have thought they'd have been able to at least isolate it to the power supply or motherboard very quickly, but I guess it may still take them time to source an appropriate new power supply if that was the problem.
Rambus would be dead without Intel. The Dramurai don't like it, nor does AMD. Even Intel admits it's not suitable for large memory sizes (Foster server variant of Willamette is DDR), and has been backing away from it elsewhere with it's MTH chips. There's even the secret Almador DDR chipset for Willamette. Rambus's only hope at RDRAM being a mass market product is if Intel stick with RDRAM for Willamette and if Willamette is a success. Intel's Rambus warrants and patent deals are a small price to pay if that's what it takes to push RDRAM into the market place.
DDR will not be more lucrative for Rambus if they lose the patent battle. The reason they want RDRAM to succeeed over DDR (as they have stated - this isn't my or anyone's else second guess!) is becuse they do have control over the RDRAM patents but it's a lot more unsure whether they'll prevail with SDRAM/DDR. 2% of X is a lot higher than 3-4% of zero.
Rambus absolutely do care whether RDRAM or DDR wins. GO back and read the initial quote from Rambus. They admit they want DDR fees to push people to RDRAM. If they didn't care which won, they'd have equal fees for DDR (or less, since it doesn't use as many patents as RDRAM).
Hitachi licenced because they are about to be purchased and didn't want a lawsuit hanging over them. Toshiba is a tiny player in the DRAM market, and there is various speculation about why they signed. Even Rambus friend Intel says the patents are not cut and dry (funny why Intel would care about DDR since they claim they're going to be all RDRAM, isn't it!).
Benchmarks on Micron's DDR reference platform show DDR to be better performing than RDRAM.
Fees for RDRAM are about 2%. The Toshiba agreement has SDRAM less, but DDR is more even though it's obviously using fewer of Rambus's (disputed) patents then RDRAM. Rambus themselves (see the original article) have admitted that they are using higher DDR fees to try to coerce people to use RDRAM.
Agreed that a 3 or 4% licencing fee won't make DDR anywhere near as expensize as RDRAM, and even if they were the same cost DDR is still the higher performance solution. Still, the arrogance of Rambus feeling free to admit they are trying to subvert technical excellence with licencing fees is pretty disgusting!
AMD may have started out as an Intel licencee, as Intel wanted a second source. When Intel felt AMD was starting to compete with them they refused to licence their designs and bus to AMD. Result - Athlon and Duron with its Alpha based EV6 bus - no Intel IP there, and funnily enough free of Intel IP AMD is now the speed champion!
So gripe at AMD's history if you want to, but today's Athlon and Duron are the fastest processors you can buy.
Duron is not intended as a mobile/low power part. It's meant to compete with Celeron (and even PIII) on the desktop.
AMD's next mobile (low power) part will be a version of their next generation "Mustang" core due out later this year.
AMD are selling processors as fast as they can make them and ramping up production in their Dresden fab as fast as they can. The timing of the mobile Mustang should coincide with their ability to produce them in sufficient volume to satisty the market.
Certainly AMD has disappointed as an investment in the past, which is exactly why the stock is so ridiculously cheap (forward P/E of 15 or so, vs 40-50 for Intel).
Last quarter AMD earned $1.15/sh completely blowing away street estimates of less that 50c, and by the estimates I am hearing it will provide another MAJOR positive surprise when current quarter earnings are out on July 19th...
If you wait to invest in AMD until it has a history of positive results you will miss out on the big gains to be made by those who realize that profits will follow performance.... Note that when AMD announced it's 1GHz Athlon, that for the first time in history Intel had to price the 1GHz PIII UNDER AMD's part!!!
Right now your PC processor choices are PIII/Athlon or Celeron/Duron. In both cases the AMD processors are clock-for-clock faster as well as cheaper, in addition to being available in higher speeds. The latest Dell catalog I received still only has 866MHz PIII's, while Gateway has 900, 950 and 1GHz Athlons. Duron is available up to 700MHz (actually there are 750's on sale in Japan!), while Celeron is still 700 MHz.
There's always going to be those who hold out for tomorrow's Mustang (AMD) or Willamette (Intel), or whatever, but for anyone buying today AMD's choices provide bother better performance as well as value.
The only benchmarks I've seen where Intel beats AMD are one or two where Rambus RDRAM makes a difference, but with cheap DDR SDRAM a few months away, it's hard to justify paying today's crazy RDRAM prices. If you're buying an SDRAM PC and want top performace, then AMD is your only choice.
There's a new standard "AGP Pro" for AGP slots that provide more power.
It's nothing to do with AMD/Intel or specific motherboards. Certain video cards need an AGP Pro slot to get adequate power - you need to check the specs before buying.
You're comparing MSRP with street prices
on
AMD's Duron Birthed
·
· Score: 2
Wait a few weeks and check for Duron prices on pricewatch.com.
While you're at it, compare with PIII prices, because that's what Duron's performance is closest too, not Celeron. Remember Duron has a full speed 200MHz FSB, while celeron is held back to 66 MHz.
Even if the SDRAM/DDR royalties are a bit more than the 2% Rambus gets from RDRAM, that's not going to make RDRAM competetive with DDR. RDRAM costs a LOT more than DDR, and the reasons are packaging, yield and testing costs.
The only affect of Toshiba buckling is that it weakens the case of the rest of the "Dramurai" that are battling Rambus.
This really has nothing to do with AMD (who incidently is already a Rambus licencee). DDR will win vs Rambus based on performance and cost, and if DDR prices are raised by royalty fees it'll still be way cheaper than RDRAM, and will still be used by both AMD and Intel.
RDRAM will soon be dead, whether or not Rambus succeeds in getting fat off SDRAM/DDR royalties.
AMD already have around 17% market share, and onece Fab30 in Dresden is fully ramped it will be running 5000+ WSPW, which at current die sizes will give them enough capacity to take 30%+ market share. AMD is actually gaining market share at the HIGH end, since their process technology is currently well ahead of Intels (all.18 micron, with copper interconnect in Dresden, moving to.13 micron by next year). Intel are still struggling to upgrade their fabs from.25 micron to.18, and are still using aluminum interconnect - it'll be a year before they move to copper.
Go to any computer store and all the fastest computers (900MHz+) are going to be AMD.
On top of the process advantages, the Athlon/Thunderbird is a modern core design, and has plenty of head room for clock speed upgrades. Intel's new Willamette core may conceivably start sample in a few months time, but won't be start to be available in any sort of volume to replace the PIII until Q1 2001. Intel expect it will take 3-8 quarters (huge range!) for Willamette to fully replace PIII. In the meantime, AMD Mustang will be out in Q4, and Sledgehammer in late 2001.
BTW, AMD's 760-MP (aka 770) dual way SMP chipset is due out in Q3 to coincide with DDR availability. Micron (whose chipset technology is licenced to Via) are coming out with high end SMP chipset support for Thunderbird.
For slot-A/socket-A mobo information, you can check http://www.slota.com (socketa.com points to the same page).
The KZ133 has been sampling for a long time, and will be available in volume along with socket-A motherboards to coincide with the Duron/Thunderbord launch in early June. There are at least 3 socket-A motherboards that should be available at launch (the FIC mobo is already on pricewatch.com - search for "Duron" or "Thunderbird").
There will also initially be some slot-A Thunderbirds to ease the transition, and these will work with existing AMD 750 based slot-A motherboards, although don't expect to be able to buy a slot-A Thunderbird in retail - they are likely to only show up in OEM computers.
The AMD 760 chipset is for DDR and will be available in Q3. There is also the 760-MP (aka 770) which will support 2-way SMP as well as DDR.
There are also many other AMD chipsets coming, such as Micron's DDR/SMP ones, Via's KZ266 for DDR, ALI (Acer Labs) support, etc.
Of the top 10 PC manufacturs, 9 use AMD parts (incl. Compaq, Gateway, IBM, HP...), with Dell being the only hold out.
Guaranteed volume may be a factor of why Dell hasn't yet adopted AMD parts, but another major reason is that it would jeapordize Dell's special relationship with Intel.
Still, performace and price are likely to force Dell to use AMD processors this year in order to remain competetive. Thunderbird is likely to be around 10% faster than PIII at the same clock speed, but will in fact also be available at much higher clock speeds than PIII.
Ignoring Dell's limited edition 1GHz PIII made possible by their Intel relationship (1GHz PIII won't be in volume production until Q4), the fastest PIII currently available is the 866MHz.
1GHz Thunderbird vs 866MHz PIII is a big performance gain for Dell to expect it's customers to give up. The latest Computer Shopper is even advertizing the as yet unannounced 1.1GHz Thunderbird, which may be a surprise announcent at the launch on June 5th.
Intel are not going to be able to match Thunderbird clock speed or performance until Willamette is launched towards the end of this year. At that point AMD Mustang will be out, and may well again be the faster processor (no benchmarks for either Willamette or Mustang are available, but AMD technology roadmap indicate they are about a year ahead of Intel in the move to.13 micron process technology - Mustang may well be a ".15".18/.13 micron process hybrid).
Duron has 64K L2 cache, but it's exclusive so there's no duplication of what's in L1 and L2. Total Duron cache is 192K = 128K L1 + 64K L2, which is more than Celeron (I or II).
Itanium (aka Merced) is years late, and is now slated to launch sometime this year at a rather underwhelming 800MHz, although Intel themselves admit they are currently having problems getting it much faster than 600MHz in the lab. It uses up ungodly amounts of power, and has terrible heat problems.
While Itanium is supposedly IA-32 compatible, this compatability comes at the cost of performance. Itanium runs 32 bit software (i.e. all currently available software, including Windows itself) *SLOWER* clock for clock than a 32 bit processor. In other words, you'll probably get better performace from an entry level 600MHz Duron than a 600MHz IA-64 !!! A 1GHz Thunderbird (or PIII if you can find one) will simply blow Itanium away on 32 bit performance.
Itanium is such a poor implementation of the IA-64 architecture that Intel's technology partner HP has done a complete Itanium replacement IA-64 implementation called McKinley, that they have licenced back to Intel. McKinley will be the first IA-64 implementation to offer decent performance and clock speeds (although still suffering from poor 32 bit performance), but is not scheduled to be released by Intel until late 2001.
Itanium, or even McKinley, are not going to be mainstream desktop parts due to being outperformed by 32 bit chips such PIII and Athlon, and more modern chips such as Intel's upcoming Willamette and AMDs Thunderbird, Mustang and 64/32 bit Sledgehammer. AMD Sledgehammer (slated for late 2001 release) includes dual 32-bit cores and will provide stunning 32 bit speed as well as 64 bit mode - it therefore offers a smooth upgrade path allowing 64 bit applications to run alongside 32 bit ones running at top speed.
IMO we havn't seen the last of the Itanium/Merced delays. I expect that Itanium will never see a full volume release, but will rather be an extremely low volume stop-gap solution until the better performing McKinley is available sometime next year.
Slashdot Mirror
Are you sure about 15 for Athlon?
I was going by Dirk Meyer's presentation on AMD's web site, but I was relying on my memory which is perhaps a bad thing!
That depends on what clock speed you're talking about. The lower speed parts overclock better because there's more headroom in the core voltage.
The Duron, which is cheaper and only exists in "slower" (still 600MHz+) speed grades is much more overclockable though.
You can still overclock Athlon/Duron the same way you do Celeron/PIII by increasing FSB speed.
There is no difference between AMD and Intel in this regard - they BOTH lock the multiplier. While it might have been cool to allow it to be adjusted, there is no way to prevent remarkers from abusing this.
Given that AMD has been selling 900MHz Thunderbirds downbinned to 700MHz, and that Duron appears equally overclockable (same way as Celeron), I'd still rate AMD as very overclocker friendly!
AMD Athlon pipeline is only 10 or 12 stages.
Intel P4 (aka Willamette) has 20 stage pipeline, and it remains to be seen whether the high clock rates this enables makes up for the hits it'll take due to latency and branch mispredict penalty.
AMD's new socket-A interface is used by Duron, Thunderbird (i.e. new Athlon) and will also be used by their Mustang due out later this year.
If you buy a socket-A Duron or Athlon now, you'd also have the option of upgrading to a dual SMP 760-MP DDR based mobo later this year.
AMD's brand new PowerNow! technology is way more advanced than Intel's power/speed levels because it can automatically dynamically change the power and CPU speed depending on processor load! For example, during DVD playback it'll use less power during slow moving scenes that are less CPU intensice to decompress. In practice this can make the difference between having the battery power to see the whole movie or now. I have no idea how they do this, but it works!
Right now PowerNow is K6 only, but later this year there will be a "mobile Duron" chip that combines a Duron variant of the enhanced Corvette core (that will debut in Mustang) with PowerNow!
The buggy Coppermine chipsets and motherboards were all Intel parts, so Intel was to blame. In Gateway's case they are using a Jabil motherboard (using an AMD 750 chipset), and have isolated the problem to either being the motherboard or the power supply. The AMD chipset has been around for a very long time, and has a reputation for great stability - not surprisingly Gateway do not think it is a suspect.
http://www.amd.co m/products/cpg/athlon/benchmarks/benchmarks.html
In some cases it's close, but Athlon is ahead of PIII on pretty much everything. On computationally intensive stuff like Photoshop, it's not even close - Athlon scorches PIII (for that matter Duron scorches PIII on floating point too):
Photoshop benchmark
The independent review sites such as Anandtech, Toms hardware, Aces, etc all have pretty much the same results, although the details obviously differ based on the exact system configuration.
The only benchmarks I've seen where PIII beats Athlon are memory benchmarks where the PIII is running on an i840 platform with PC800 Rambus memory, or on an overclocked BX platform. For many people these arn't realistic comparison platforms, since PC800 Rambus memory is prohibitively expensive and most people do not overclock their systems.
Anyone looking for an affordable high performance memory solution should wait until Sept/Oct for AMD's 760 chipset which supports DDR. Given Athlons 200/266MHz DDR FSB, the speedup you get with DDR should be much greater than PIII gets with Rambus memory, since PIII is limited by it's 100/133MHz FSB.
http://www.techweb.com/wire/stor y/TWB20000630S0011
This spells it out a bit clearer that both Gateway and AMD agree it's a Gateway issue - either the motherboard or power supply.
Originally Gateway said it's be fixed by July 10th, but that has more recently changed to July 18th. Given the specificity of the date, I assume that they now know what the issue is, and are waiting either for rev'd motherboards from Jabil, or for a shippment of an appropriately spec'd power supply.
Interestingly AMD's web site:
http://www1.amd.com/athlon/power
Lists 52 approved power supply for 900MHz ATX, but only 37 for 1GHz ATX...
Thanks for pointing that out.
I'd assumed it was Gateway's motherboard that was to blame, since I've seen AMD approved motherboards go from 900MHz max at one rev., to 1GHz at a higher rev. (e.g. MSI 6340 1.0 -> 1.0A).
I'd have thought they'd have been able to at least isolate it to the power supply or motherboard very quickly, but I guess it may still take them time to source an appropriate new power supply if that was the problem.
Rambus would be dead without Intel. The Dramurai don't like it, nor does AMD. Even Intel admits it's not suitable for large memory sizes (Foster server variant of Willamette is DDR), and has been backing away from it elsewhere with it's MTH chips. There's even the secret Almador DDR chipset for Willamette. Rambus's only hope at RDRAM being a mass market product is if Intel stick with RDRAM for Willamette and if Willamette is a success. Intel's Rambus warrants and patent deals are a small price to pay if that's what it takes to push RDRAM into the market place.
DDR will not be more lucrative for Rambus if they lose the patent battle. The reason they want RDRAM to succeeed over DDR (as they have stated - this isn't my or anyone's else second guess!) is becuse they do have control over the RDRAM patents but it's a lot more unsure whether they'll prevail with SDRAM/DDR. 2% of X is a lot higher than 3-4% of zero.
Rambus absolutely do care whether RDRAM or DDR wins. GO back and read the initial quote from Rambus. They admit they want DDR fees to push people to RDRAM. If they didn't care which won, they'd have equal fees for DDR (or less, since it doesn't use as many patents as RDRAM).
Hitachi licenced because they are about to be purchased and didn't want a lawsuit hanging over them. Toshiba is a tiny player in the DRAM market, and there is various speculation about why they signed. Even Rambus friend Intel says the patents are not cut and dry (funny why Intel would care about DDR since they claim they're going to be all RDRAM, isn't it!).
Benchmarks on Micron's DDR reference platform show DDR to be better performing than RDRAM.
Fees for RDRAM are about 2%. The Toshiba agreement has SDRAM less, but DDR is more even though it's obviously using fewer of Rambus's (disputed) patents then RDRAM. Rambus themselves (see the original article) have admitted that they are using higher DDR fees to try to coerce people to use RDRAM.
Agreed that a 3 or 4% licencing fee won't make DDR anywhere near as expensize as RDRAM, and even if they were the same cost DDR is still the higher performance solution. Still, the arrogance of Rambus feeling free to admit they are trying to subvert technical excellence with licencing fees is pretty disgusting!
AMD may have started out as an Intel licencee, as Intel wanted a second source. When Intel felt AMD was starting to compete with them they refused to licence their designs and bus to AMD. Result - Athlon and Duron with its Alpha based EV6 bus - no Intel IP there, and funnily enough free of Intel IP AMD is now the speed champion!
So gripe at AMD's history if you want to, but today's Athlon and Duron are the fastest processors you can buy.
Duron is not intended as a mobile/low power part. It's meant to compete with Celeron (and even PIII) on the desktop.
AMD's next mobile (low power) part will be a version of their next generation "Mustang" core due out later this year.
AMD are selling processors as fast as they can make them and ramping up production in their Dresden fab as fast as they can. The timing of the mobile Mustang should coincide with their ability to produce them in sufficient volume to satisty the market.
Certainly AMD has disappointed as an investment in the past, which is exactly why the stock is so ridiculously cheap (forward P/E of 15 or so, vs 40-50 for Intel).
Last quarter AMD earned $1.15/sh completely blowing away street estimates of less that 50c, and by the estimates I am hearing it will provide another MAJOR positive surprise when current quarter earnings are out on July 19th...
If you wait to invest in AMD until it has a history of positive results you will miss out on the big gains to be made by those who realize that profits will follow performance.... Note that when AMD announced it's 1GHz Athlon, that for the first time in history Intel had to price the 1GHz PIII UNDER AMD's part!!!
Right now your PC processor choices are PIII/Athlon or Celeron/Duron. In both cases the AMD processors are clock-for-clock faster as well as cheaper, in addition to being available in higher speeds. The latest Dell catalog I received still only has 866MHz PIII's, while Gateway has 900, 950 and 1GHz Athlons. Duron is available up to 700MHz (actually there are 750's on sale in Japan!), while Celeron is still 700 MHz.
There's always going to be those who hold out for tomorrow's Mustang (AMD) or Willamette (Intel), or whatever, but for anyone buying today AMD's choices provide bother better performance as well as value.
The only benchmarks I've seen where Intel beats AMD are one or two where Rambus RDRAM makes a difference, but with cheap DDR SDRAM a few months away, it's hard to justify paying today's crazy RDRAM prices. If you're buying an SDRAM PC and want top performace, then AMD is your only choice.
There's a new standard "AGP Pro" for AGP slots that provide more power.
It's nothing to do with AMD/Intel or specific motherboards. Certain video cards need an AGP Pro slot to get adequate power - you need to check the specs before buying.
Wait a few weeks and check for Duron prices on pricewatch.com.
While you're at it, compare with PIII prices, because that's what Duron's performance is closest too, not Celeron. Remember Duron has a full speed 200MHz FSB, while celeron is held back to 66 MHz.
Duron is a unbelievable price/performance deal!!!
Even if the SDRAM/DDR royalties are a bit more than the 2% Rambus gets from RDRAM, that's not going to make RDRAM competetive with DDR. RDRAM costs a LOT more than DDR, and the reasons are packaging, yield and testing costs.
The only affect of Toshiba buckling is that it weakens the case of the rest of the "Dramurai" that are battling Rambus.
This really has nothing to do with AMD (who incidently is already a Rambus licencee). DDR will win vs Rambus based on performance and cost, and if DDR prices are raised by royalty fees it'll still be way cheaper than RDRAM, and will still be used by both AMD and Intel.
RDRAM will soon be dead, whether or not Rambus succeeds in getting fat off SDRAM/DDR royalties.
AMD already have around 17% market share, and onece Fab30 in Dresden is fully ramped it will be running 5000+ WSPW, which at current die sizes will give them enough capacity to take 30%+ market share. AMD is actually gaining market share at the HIGH end, since their process technology is currently well ahead of Intels (all .18 micron, with copper interconnect in Dresden, moving to .13 micron by next year). Intel are still struggling to upgrade their fabs from .25 micron to .18, and are still using aluminum interconnect - it'll be a year before they move to copper.
Go to any computer store and all the fastest computers (900MHz+) are going to be AMD.
On top of the process advantages, the Athlon/Thunderbird is a modern core design, and has plenty of head room for clock speed upgrades. Intel's new Willamette core may conceivably start sample in a few months time, but won't be start to be available in any sort of volume to replace the PIII until Q1 2001. Intel expect it will take 3-8 quarters (huge range!) for Willamette to fully replace PIII. In the meantime, AMD Mustang will be out in Q4, and Sledgehammer in late 2001.
BTW, AMD's 760-MP (aka 770) dual way SMP chipset is due out in Q3 to coincide with DDR availability. Micron (whose chipset technology is licenced to Via) are coming out with high end SMP chipset support for Thunderbird.
For slot-A/socket-A mobo information, you can check http://www.slota.com (socketa.com points to the same page).
The KZ133 has been sampling for a long time, and will be available in volume along with socket-A motherboards to coincide with the Duron/Thunderbord launch in early June. There are at least 3 socket-A motherboards that should be available at launch (the FIC mobo is already on pricewatch.com - search for "Duron" or "Thunderbird").
There will also initially be some slot-A Thunderbirds to ease the transition, and these will work with existing AMD 750 based slot-A motherboards, although don't expect to be able to buy a slot-A Thunderbird in retail - they are likely to only show up in OEM computers.
The AMD 760 chipset is for DDR and will be available in Q3. There is also the 760-MP (aka 770) which will support 2-way SMP as well as DDR.
There are also many other AMD chipsets coming, such as Micron's DDR/SMP ones, Via's KZ266 for DDR, ALI (Acer Labs) support, etc.
Of the top 10 PC manufacturs, 9 use AMD parts (incl. Compaq, Gateway, IBM, HP...), with Dell being the only hold out.
.13 micron process technology - Mustang may well be a ".15" .18/.13 micron process hybrid).
Guaranteed volume may be a factor of why Dell hasn't yet adopted AMD parts, but another major reason is that it would jeapordize Dell's special relationship with Intel.
Still, performace and price are likely to force Dell to use AMD processors this year in order to remain competetive. Thunderbird is likely to be around 10% faster than PIII at the same clock speed, but will in fact also be available at much higher clock speeds than PIII.
Ignoring Dell's limited edition 1GHz PIII made possible by their Intel relationship (1GHz PIII won't be in volume production until Q4), the fastest PIII currently available is the 866MHz.
1GHz Thunderbird vs 866MHz PIII is a big performance gain for Dell to expect it's customers to give up. The latest Computer Shopper is even advertizing the as yet unannounced 1.1GHz Thunderbird, which may be a surprise announcent at the launch on June 5th.
Intel are not going to be able to match Thunderbird clock speed or performance until Willamette is launched towards the end of this year. At that point AMD Mustang will be out, and may well again be the faster processor (no benchmarks for either Willamette or Mustang are available, but AMD technology roadmap indicate they are about a year ahead of Intel in the move to
Duron has 64K L2 cache, but it's exclusive so there's no duplication of what's in L1 and L2. Total Duron cache is 192K = 128K L1 + 64K L2, which is more than Celeron (I or II).
Itanium (aka Merced) is years late, and is now slated to launch sometime this year at a rather underwhelming 800MHz, although Intel themselves admit they are currently having problems getting it much faster than 600MHz in the lab. It uses up ungodly amounts of power, and has terrible heat problems.
While Itanium is supposedly IA-32 compatible, this compatability comes at the cost of performance. Itanium runs 32 bit software (i.e. all currently available software, including Windows itself) *SLOWER* clock for clock than a 32 bit processor. In other words, you'll probably get better performace from an entry level 600MHz Duron than a 600MHz IA-64 !!! A 1GHz Thunderbird (or PIII if you can find one) will simply blow Itanium away on 32 bit performance.
Itanium is such a poor implementation of the IA-64 architecture that Intel's technology partner HP has done a complete Itanium replacement IA-64 implementation called McKinley, that they have licenced back to Intel. McKinley will be the first IA-64 implementation to offer decent performance and clock speeds (although still suffering from poor 32 bit performance), but is not scheduled to be released by Intel until late 2001.
Itanium, or even McKinley, are not going to be mainstream desktop parts due to being outperformed by 32 bit chips such PIII and Athlon, and more modern chips such as Intel's upcoming Willamette and AMDs Thunderbird, Mustang and 64/32 bit Sledgehammer. AMD Sledgehammer (slated for late 2001 release) includes dual 32-bit cores and will provide stunning 32 bit speed as well as 64 bit mode - it therefore offers a smooth upgrade path allowing 64 bit applications to run alongside 32 bit ones running at top speed.
IMO we havn't seen the last of the Itanium/Merced delays. I expect that Itanium will never see a full volume release, but will rather be an extremely low volume stop-gap solution until the better performing McKinley is available sometime next year.