Lancaster 64 = Winchester Revision E, Launch date = 04.18.05

Zebo

Elite Member
Jul 29, 2001
39,398
19
81
Saw this at Xteme http://www.xtremesystems.org/forums/showthread.php?t=52361

--------------------------------------------------------------------------------

http://www.digitimes.com/systems/a20050205A2003.html


Quote:
Vendors participating in the campaign will unveil new notebooks built using AMD Turion 64 2800+ and 3000+ CPUs, with power consumption averaging 25W or 35W, depending on models, the sources noted.



http://www.rage3d.de/index.php?show=3499


Quote:
AMD Turion 64 2800+ (1,60 GHz, 1 MB L2-Cache, 90 nm, Revision E, TDP: 25 Watt)
AMD Turion 64 3000+ (1,80 GHz, 1 MB L2-Cache, 90 nm, Revision E, TDP: 25 Watt)
AMD Turion 64 3200+ (2,00 GHz, 1 MB L2-Cache, 90 nm, Revision E, TDP: 35 Watt)

Remeber these are the bad boys that can clock 24% higher According to AMD... since now they clock 2700Mhz... I'd expect 3300Mhz (course you loose all those nice low power numbers.. but who cares on desktop?)
 

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
They are sckt 939 right??? That is't too far off....I think I better get a 10x multi this time though!!!
 

Zebo

Elite Member
Jul 29, 2001
39,398
19
81
754:) don't worry.. Venice is coming prolly earlier.. - 512 of course:(
 

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
Originally posted by: Zebo
754:) don't worry.. Venice is coming prolly earlier.. - 512 of course:(

WHY???? Why would they add 1mb of l2 cace and hold winchester pr rating...are these still single channel memory controllers???
 

Accord99

Platinum Member
Jul 2, 2001
2,259
172
106
Originally posted by: Zebo
Remeber these are the bad boys that can clock 24% higher According to AMD... since now they clock 2700Mhz... I'd expect 3300Mhz (course you loose all those nice low power numbers.. but who cares on desktop?)
These are the CPUs that used the latest AMD/IBM SS technique, which can allow transistors to switch up to 24% faster. Which does not translate to CPUs being able to clock 24% faster as there are other (even more) important variables to CPU clock speed limits like wire delays.
 

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
I think from reading that post, that nothing looks definitive for it being S754. A bunch of speculation from ppl who really dont know one way or another...It is a 90nm part...the only reason to have S754 would definitely be if there was only a single channle memory controller...
 

Zebo

Elite Member
Jul 29, 2001
39,398
19
81
Originally posted by: Duvie
Originally posted by: Zebo
754:) don't worry.. Venice is coming prolly earlier.. - 512 of course:(

WHY???? Why would they add 1mb of l2 cace and hold winchester pr rating...are these still single channel memory controllers???


Why does 3400 beat 3500 in every bench? I dunno. but Turion is Dothan Killer. All Laptop chips for AMD 64 use skt 754. Lancaster, dubin, oakville etc..


 

Accord99

Platinum Member
Jul 2, 2001
2,259
172
106
Originally posted by: Zebo
Why does 3400 beat 3500 in every bench? I dunno. but Turion is Dothan Killer. All Laptop chips for AMD 64 use skt 754. Lancaster, dubin, oakville etc..
No it's not, Dothan is faster and uses less power.
 

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
What is the link showing me....Also just cause the A64 lappies to date have been sckt 754 doesn't say much...they were all 130nm parts, right??? this may be all new and require newer mobo like our winnies did...
 

Zebo

Elite Member
Jul 29, 2001
39,398
19
81
Originally posted by: Accord99
Originally posted by: Zebo
Why does 3400 beat 3500 in every bench? I dunno. but Turion is Dothan Killer. All Laptop chips for AMD 64 use skt 754. Lancaster, dubin, oakville etc..
No it's not, Dothan is faster and uses less power.

btw How do you fiqure dothan is faster?

Out of 35 benches 2.0Ghz A64 wins 24 over 2.0Ghz Pentium M.. And that with crippling 512 Cache... This chip gets more.

http://techreport.com/reviews/2005q1/dfi-855gme-mgf/index.x?pg=2

Intel fanboy?
 

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
I see the link that say 754 packaging for the mobile sempron but it is not the same chip since none of the l2 caches match out on the chips...Eveyone knew the sempron was going to be sckt 754....The highest l2 cached chip is 256kb..,..Make me more certaion now the top of the line A64 mobile witll be dual channel and 1mb l2 cache and therefore will be a socket 939 variant...
 

Zebo

Elite Member
Jul 29, 2001
39,398
19
81
Duvie..


read thread you'd see a road map in there with Lancaster which clearly says skt 754..Lancaster is the big dog with 1mb cache...Speculation? It show up on AMD road map. Has same cache and process they say it will. And now we know date.
 

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
http://www.amd.com/us-en/Processors/Pro.../0,,30_118_10220_10221%5E11030,00.html


maybe you are right here is the low power Mobile A64s.....It says nothing about 1mb l2 cache chips but does say 90nm tech and is SOI but not SS...It also states 754....

Then there is....

http://www.amd.com/us-en/Processors/Pro.../0,,30_118_10220_10221%5E10269,00.html


here it shows 1mb l2 but it is 130nm (is current chips I think) and sckt 754...no sruprise there though the number convention matching winchesters sckt 939's...


they are all single channel memory controller which makes sckt 754 make since.....

very confusing nonetheless...
 

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
I see it and yes definitely it will be single channel controller....damn the 90nm sckt 939 variant better be right on its heels...usually laptop tech is behind the desktop tech so I would expect to see the ss/soi rev e in the desktop though likely still 512mb l2 cache to be almost immediately or in 0510 chips and newer....Its is likley with the desktop since there is no cache revisions it will go pretty silently...I eman what would they really report??? Maybe just a 3700+ release or something....
 

Accord99

Platinum Member
Jul 2, 2001
2,259
172
106
Originally posted by: Zebo
btw How do you fiqure dothan is faster?

Out of 35 benches 2.0Ghz A64 wins 24 over 2.0Ghz Pentium M.. And that with crippling 512 Cache... This chip gets more.

http://techreport.com/reviews/2005q1/dfi-855gme-mgf/index.x?pg=2

Intel fanboy?
It merely depends on who's benchmarks you read. A 2.4GHz Dothan beats the 2.4GHz 1MB L2 cache S939 3800+ 6-5 here. This would be the closest comparison of the Sonoma platform Dothan versus the Turion.
http://www.hardware.fr/art/imprimer/546

And most importantly, the Dothan based systems are significantly better in power consumption and therefore the Dothan has the performance edge per watt, the key element of the mobile market. Certainly no Dothan killer, at best it'll be more competitive than the mobile XP versus Banias.
 

Zebo

Elite Member
Jul 29, 2001
39,398
19
81
Originally posted by: Accord99
Originally posted by: Zebo
btw How do you fiqure dothan is faster?

Out of 35 benches 2.0Ghz A64 wins 24 over 2.0Ghz Pentium M.. And that with crippling 512 Cache... This chip gets more.

http://techreport.com/reviews/2005q1/dfi-855gme-mgf/index.x?pg=2

Intel fanboy?
It merely depends on who's benchmarks you read. A 2.4GHz Dothan beats the 2.4GHz 1MB L2 cache S939 3800+ 6-5 here. This would be the closest comparison of the Sonoma platform Dothan versus the Turion.
http://www.hardware.fr/art/imprimer/546

And most importantly, the Dothan based systems are significantly better in power consumption and therefore the Dothan has the performance edge per watt, the key element of the mobile market.

While nice you can't compare an OC bus to a non OC bus real well. These will be in laptops not overclcoked. These also wont be at 2.4 but ~2.0. Even using Behardware , under those conditions 2.0 A64 wins again 6:5. Besides Techeport is the most comprehensive test suite on the net..(obviously by some 35 tests vs. 11) most scientific too testing 2x each time and averageing. I trust them infinity more.

What's "significantly better in power consumption"? define?

Last I heard Dothan use 21W if these Turion Chips use 25W at both those points CPU is only a small part of overall system power consumption in a laptop, where things like the hard drive and LCD display can dominate the battery life equation. Again well see. INtel been known to tell fairy tales about TDP while AMD underrates.
 

Accord99

Platinum Member
Jul 2, 2001
2,259
172
106
Originally posted by: Zebo
Last I heard Dothan use 21W if these Turion Chips use 25W at both those points CPU is only a small part of overall system power consumption in a laptop, where things like the hard drive and LCD display can dominate the battery life equation. Again well see. INtel been known to tell fairy tales about TDP while AMD underrates.
And yet the same system power consumption tests, including Tech Report show a significant difference. And CPU measurements articles indicate Intel's TDP for Dothan has significant margin, with a 2GHz Dothan measured at 16W. While the use of a higher TDP rating for the 2GHz model would suggest it does not use less than 25W.
http://www.x86-secret.com/articles/divers/conso/conso-3.htm
And for the thin and light market that AMD hopes to compete in with Turion, CPU power consumption is important with the typical thin and light laptop using ~10W at idle and no more than 40W at full load, every watt counts.
 

Zebo

Elite Member
Jul 29, 2001
39,398
19
81
Originally posted by: Accord99
Originally posted by: Zebo
Last I heard Dothan use 21W if these Turion Chips use 25W at both those points CPU is only a small part of overall system power consumption in a laptop, where things like the hard drive and LCD display can dominate the battery life equation. Again well see. INtel been known to tell fairy tales about TDP while AMD underrates.
And yet the same system power consumption tests, including Tech Report show a significant difference. And CPU measurements articles indicate Intel's TDP for Dothan has significant margin, with a 2GHz Dothan measured at 16W. While the use of a higher TDP rating for the 2GHz model would suggest it does not use less than 25W.
http://www.x86-secret.com/articles/divers/conso/conso-3.htm
And for the thin and light market that AMD hopes to compete in with Turion, CPU power consumption is important with the typical thin and light laptop using ~10W at idle and no more than 40W at full load, every watt counts.

I don't understand how we can look at techreport for those power values on AMD mobile. This will be a mobile chip with totally different power profile of the desktop chip used there. AMD even says so, 25W in above links vs. 67 W for techreports chip.

I'm sure it's will be Like mobile XP's, cherry picked processors running at lower volts. For example the 2400 XP moblie used 1.35V vs. thier desktop parts used 1.65 or (35W vs. ~55W at same speed)

Performance will not change signifigantly, maybe +3-5% in SSE3 stuff but same in other apps.. It's almost same core, with SSE3 and 1MB lvl2 but minus dual channel which cancels one another out.
 

68GTX

Member
Sep 1, 2001
187
0
0
The Pentium-M 770 @ 2.13GHz/533MHz FSB with Dual Channel DDR2 on the Sonoma Platform is the real competition.

Once the 915 based Sonoma desktop mobos are released, we'll have a better idea of the true application performance, and thermal performance potential of these chips using both DDR and DDR2 memory.

I'd love to see some serious competition in this sector, and it will be nice if these Turion chips will work in desktop mobos.
 

Megatomic

Lifer
Nov 9, 2000
20,127
6
81
Duvie and Zebo, you guys are pretty knowledgeable about this stuff so let me ask you: is there going to be a performance delta, at stock speeds, between a Lancaster 3400+ and my Clawhammer 3400+? Both have the same bus speed, cache size, and overall CPU speed. I'm not overclocking the chip so reduced heat load and higher o/c speeds are not pertinent to me at this time. Thanks men. :beer:
 

clarkey01

Diamond Member
Feb 4, 2004
3,419
1
0
Originally posted by: Megatomic
Duvie and Zebo, you guys are pretty knowledgeable about this stuff so let me ask you: is there going to be a performance delta, at stock speeds, between a Lancaster 3400+ and my Clawhammer 3400+? Both have the same bus speed, cache size, and overall CPU speed. I'm not overclocking the chip so reduced heat load and higher o/c speeds are not pertinent to me at this time. Thanks men. :beer:

I think they will run at the same clock, there some minutte 90 nm improvments so the Lanni will have the edge, but the best "3400+" is the newcastle S745, due to its clock of 2.4 Ghz.