The future of desktop Dothans?

[jrw]

What do you consider to be the future for desktop Dothans?

Currently they are quite limited to the old 865/875 Asus with the exception of the P4GD1 (which has gone out of production as I understand it) and the low-end P4GPL-X.

Do you think Asus will come up with a new adapter compatible with other boards or maybe even a new board specifically designed to the Dothan (as the Aopen 915 board)?

Does the future hold more PCI-E Dothan-supporting boards?

Will Dothan be king of 1M in 6 months? Will it own the ORB?

[Damien]

they need to make an adapter for LGA

[toledo]

Originally Posted by DamienKC

they need to make an adapter for LGA

Would be great to use a Dothan is a Asus P5WD2 .

I guess Asus CT-479 adapter is only a test to see how gamers and overclockers accept this technology in a desktop environment , it is part of something bigger . I think we will have some news regarding Dothan on desktop soon .

[BrnmccO1]

Originally Posted by toledo

Would be great to use a Dothan is a Asus P5WD2 .

I guess Asus CT-479 adapter is only a test to see how gamers and overclockers accept this technology in a desktop environment , it is part of something bigger . I think we will have some news regarding Dothan on desktop soon .

Yeah, but remember were on the verge of Yonah, I've heard rumors on here from a couple of our friends overseas, there's a couple ES's floating around, but seems no-one can get it to boot except on Intel's 955GX (or something) reference board, as others don't have the micro-code in the BIOS yet to recognize the new chip. Besides the dual core, they will also have apparently fixed some of the major shortcomings of the Dothan chip, namely it's lackluster FP performance. Having two physical cores will also eliminate the disadvantage it has with the Prescott/Northwood P4 in not having Hyperthreading support. As far as Dothan on the desktop, Aopen has a promising new full-size ATX mobo out, but based on the 915G desktop chipset. We just have to keep out fingers crossed, and keep writing to DFI, to bring us an extreme overclocking, SLI mobo, with at least 2.0 volts Vcore, for the socket 479! I can see it now, DFI LanParty UT s479! A Yonah on LN2 with two 7800 GTX in SLI would kick some butt, I think. Or maybe two R580's with Crossfire? Wishful thinking...

[onewingedangel]

479 to 775 adaptor would be a lot more complex than a 479 to 478 adaptor, as 479 is 478 with a different layout and a few diferent voltages, but for the most part its the same, whereas 775 is quite different.

may as well have native boards, or possibly boards where you can choose the socket used much like the asrock boards - 478/479/775 support all on the i945 chipset.

[Duvie]

Originally Posted by stryg

What do you consider to be the future for desktop Dothans?

Currently they are quite limited to the old 865/875 Asus with the exception of the P4GD1 (which has gone out of production as I understand it) and the low-end P4GPL-X.

Do you think Asus will come up with a new adapter compatible with other boards or maybe even a new board specifically designed to the Dothan (as the Aopen 915 board)?

Does the future hold more PCI-E Dothan-supporting boards?

Will Dothan be king of 1M in 6 months? Will it own the ORB?

Is this all that matters to Intel fans in this forum...Dothan is and has weak fpu performance....It wont be the leader of much except synthetic benches. Clock for clock it is faster or equal in gaming to AMD64 s939 chips (That much I can admit.), but at specific retail speed points it wont match FX57 speeds. OC'd to higher speeds perhaps, but then you got to realize ppl can OC FX55's to near 3ghz and FX57 just as higher or higher.... However it is weak in several other areas so the real answer will be how is the single core variant on the yonah chips and will that be able to run on desktops or when will the desktop equivalent be out. We know sometime soon since the P-D liine is a dud and not set to go any higher....

[toledo]

Originally Posted by Duvie

Is this all that matters to Intel fans in this forum...

No,but games really matter for sure (at lest for the majority ). Who cares about synthetic benches ?

Originally Posted by Duvie

Clock for clock it is faster or equal in gaming to AMD64 s939 chips (That much I can admit.),

No way ,its faster not equal. A dothan at 2700Mhz kicks a FX at 3Ghz in games .

[Tony]

Desktop Dothan is coming, sse 1 2 3 and a few other tweaks to boot i bet, much higher fsb by default and ddr2 dual channel support....oh and it has 2 cores as well

3-2-2-4 at ddr600+ on a Yonah is going to be real sweet, we will have it real soon By year end i expect to see benches turning up on the net

[Duvie]

Originally Posted by toledo

No,but games really matter for sure (at lest for the majority ). Who cares about synthetic benches ?



No way ,its faster not equal. A dothan at 2700Mhz kicks a FX at 3Ghz in games .

Why are the ppl in this site so bad at having the facts straight...i now it is faster in some test but how do you account this...

http://techreport.com/reviews/2005q1...f/index.x?pg=7

2.4ghz P-M lost to a 4000+ A64 which we know is 2.4ghz....

In far cry it was in the middle between the 2.2ghz and 2.4ghz venice chips with less cache....

http://techreport.com/reviews/2005q1...f/index.x?pg=8

and again....

http://techreport.com/reviews/2005q1.../index.x?pg=10
http://techreport.com/reviews/2005q1.../index.x?pg=11

basically I can quote every page in this review and it confirms EXACTLY with what I have said....NOw you cannot argue with 2.4ghz cause they have oc'd the bus to 533fsb where the current chips are. Though it is likely the naing convention for this would have been a 790...


Now this one confirm more what you have said, but...

http://www.xbitlabs.com/articles/cpu...mm-780_15.html

You have to wonder why they ran the test at 1024...That doesn't really isolate cpu like 640x480 does...this puts stress on vid card, memory subsystem, graphic card etc...That plays too much into other things then cpu architecture...So look at howmany time the 2.4ghz 1mb and 512kb cache cores are right there with the 2.26ghz P-M chip. NOt conclusive one way or another to reverse what was shown in the Tech Report review....

[Wiggy McShades]

all those tests except the last one have dothan running single chanel ram... and saying 533 bus is an overclock for a dothan is just sad.

[Duvie]

Originally Posted by Wiggy McShades

all those tests except the last one have dothan running single chanel ram... and saying 533 bus is an overclock for a dothan is just sad.

You missed the point like most inthis forum...It wasn't about an oc'd dothan as much as about the performance of a 2.4ghz dothan if there was one...They basically created a 2.4 P-M stock cpu 790. They then used it to show where the scaling will go with this cpu at the higher speeds to be more comparable to the 3800-4000+ chips. They were trying to make it so we could see clock for clock comparisons....placing it on an 800fsb desktop platform with a flaky adapter (from many threads I have read) and using 400ddr and ocing its system bus is not comparing the stock cpus and their designed platform....If that is the case let us rerun all the test with an FX55 and 4000+ chips with 250dividers for use with a recognized standard now for 500ddr. I mean that makes as much sense...

Dothan is a mobile chip first and foremost. The minute it is placed on a desktop platform for all purposes it is OC'd and tweaked...


But still go ahead and provide links and refute what I have said....

I was being conservative when I said faster or equal to...Only true fanatics couldn't accept that wording....LOL!!! I said it and I proved it. More then I can say about most in this forum when it comes to making blank statements with no proof given...nuff said...

[toledo]

Originally Posted by Duvie

Why are the ppl in this site so bad at having the facts straight...i now it is faster in some test but how do you account this...

http://techreport.com/reviews/2005q1...f/index.x?pg=7

2.4ghz P-M lost to a 4000+ A64 which we know is 2.4ghz....

In far cry it was in the middle between the 2.2ghz and 2.4ghz venice chips with less cache....

http://techreport.com/reviews/2005q1...f/index.x?pg=8

and again....

http://techreport.com/reviews/2005q1.../index.x?pg=10
http://techreport.com/reviews/2005q1.../index.x?pg=11

basically I can quote every page in this review and it confirms EXACTLY with what I have said....NOw you cannot argue with 2.4ghz cause they have oc'd the bus to 533fsb where the current chips are. Though it is likely the naing convention for this would have been a 790...


Now this one confirm more what you have said, but...

http://www.xbitlabs.com/articles/cpu...mm-780_15.html

You have to wonder why they ran the test at 1024...That doesn't really isolate cpu like 640x480 does...this puts stress on vid card, memory subsystem, graphic card etc...That plays too much into other things then cpu architecture...So look at howmany time the 2.4ghz 1mb and 512kb cache cores are right there with the 2.26ghz P-M chip. NOt conclusive one way or another to reverse what was shown in the Tech Report review....

ppl in this site are not bad at having the facts straight, they read another sources of reviews and test by themselves

I saw something different at only 2.13 Mhz (using DDR2). Do you really think that at 2.4Ghz will be that way ?



If you want to take a look

http://www.gamepc.com/labs/print_content.asp?id=pm915

[idiotec]

I think alot of people keep underestimated them, and we are close to seeing some huge advancements based on this technology.

Just for laughs, look at what was said about them just 1 year ago:

Dothan Desktops Probably Won't Overclock All That Much

http://www.overclockers.com/tips00593/

[Duvie]

Yeah and gamePC did not jive with articles by many of the other sites...so maybe you are just shopping around scores you want to see....If I remember GamePC used cas2.5 timings on the Athlon FX and most know that the athlon excels with low latency timings and cas 2 timings are quite average.

I perfer to look at more then 1 to come up with a true visual. Running games at that res is adding the system into the variable and is not a true test of the cpu. So this is about Pentium M in a certain platform.

I can agree the numbers look good...Again I was being conservative when I said "Clock for clock it is faster or equal in gaming to AMD64 s939 chips (That much I can admit.)," This was as to not incite a flame fest...But you my friend are a fanatic and that is quite obvious....Dont take everything like I just said something about your mom....LOL!!!!

Originally Posted by toledo

ppl in this site are not bad at having the facts straight, they read another sources of reviews and test by themselves

I saw something different at only 2.13 Mhz (using DDR2). Do you really think that at 2.4Ghz will be that way ?



If you want to take a look

http://www.gamepc.com/labs/print_content.asp?id=pm915

[Astral7]

What is truely amazing is that an engine conceived in the summer and fall of 1990 is still being used as the basis of new designs 15 years later. Not only that but beating the top of the line single core amd offering which costs in excess of $1000.00. And doing this on a last generation graphics port and chipsets that are two generations from current. Merom is coming






Thanks Bob

[DevilsRejection]

yonah is 30% faster than dothan at encoding, this is what people at theinquirer have stated.

we will have to wait a week to get more info at the next IDF

[vapb400]

It looks like im going to hold out till Merom/Conroe, 64bit will soon be mainstream, it's silly not to include it on a new processor.

Maybe I'll look into some Sossaman Xeon's.

[saratoga]

I don't think Dothan has much future on the desktop. Yonah OTOH

[Duvie]

Originally Posted by vapb400

It looks like im going to hold out till Merom/Conroe, 64bit will soon be mainstream, it's silly not to include it on a new processor.

Maybe I'll look into some Sossaman Xeon's.

(this is not meant at OP I quoted)
Your guys beloved INtel (and I like INtel, but just like, not this fanboy e)

(this is)
seems to have been making bonehead blunders for the last few years...
1) prescot
2) no HT on the Pentium-D's
3) no 64bit in yonah

I would say 3 strikes and you are out but Intel is better then that....They will bounce back and be in the lead briefly then the both of them will share the lead back and forth like it has been in the last 5 years....

Who cares who the winner is...Lets us just care about getting the best cpu...

[saratoga]

(this is)
seems to have been making bonehead blunders for the last few years...
1) prescot
2) no HT on the Pentium-D's
3) no 64bit in yonah

Yonah is a pretty old design, so they probably hadn't realized that x86-64 was going to take off when they were designing it. Conroe is really the first new Intel CPU core since Intel adopted x86-64.

[BrnmccO1]

Originally Posted by saratoga

Yonah is a pretty old design, so they probably hadn't realized that x86-64 was going to take off when they were designing it. Conroe is really the first new Intel CPU core since Intel adopted x86-64.

Now, I thought Conroe was just going to be a tweaked Dothan. I might be wrong tho, I can hardly wait for the IDC. This talk about the Dothan just being a re-worked Pentium III on some other review/"news" sites has just got to stop. It was actually designed from the ground up by the guys over at the lab in Israel. Sometimes you need a fresh perspective on things to come with something new. However, the core DOES share a lot of similarities, not with the PIII, but the old Pentium Pro. Also, the u-OPS fusion thing is new, and, I suspect is one big reason it does much work per clock cycle. Also, I'm beginning to wonder if the pipeline is only 10 stages, and not 12 like the Athlon64's??

Well, 90nm didn't last long, let's see how their 65nm process turns out. If they can get the transistor gate leakage current down to an acceptable level being that small; if they can get decent clock speeds out of it without too much vCore; this might really be something. Some big if's there though.

[Astral7]

Originally Posted by BrnmccO1

Now, I thought Conroe was just going to be a tweaked Dothan. I might be wrong tho, I can hardly wait for the IDC. This talk about the Dothan just being a re-worked Pentium III on some other review/"news" sites has just got to stop. It was actually designed from the ground up by the guys over at the lab in Israel. Sometimes you need a fresh perspective on things to come with something new. However, the core DOES share a lot of similarities, not with the PIII, but the old Pentium Pro. Also, the u-OPS fusion thing is new, and, I suspect is one big reason it does much work per clock cycle. Also, I'm beginning to wonder if the pipeline is only 10 stages, and not 12 like the Athlon64's??

Well, 90nm didn't last long, let's see how their 65nm process turns out. If they can get the transistor gate leakage current down to an acceptable level being that small; if they can get decent clock speeds out of it without too much vCore; this might really be something. Some big if's there though.

The Dothan IS an extention and refinement of P6. Merom/Conroe are new.

Keeping the pipelines short is one key: keeping the pipeline "segments" short is even more important. Remember that P6-style microarchitectures are a sequence of loosely coupled pipeline segments with queueing in between the segments. Part of refining that engine is to minimize segment lengths, optimize the queueing, and provide hot bypasses when the queues are empty. Another part of improving that king of engine is to keep power dissipation in mind from day one. Don't just power things down when they're not used; group things together so the overhead of controlling the power is minimized. Pay as much attention to what things cost in heat as to what they cost in performance, because those two are now and forever will be fungible. The old days are gone.

[StrikeRTM]

Originally Posted by Duvie

(this is not meant at OP I quoted)
Your guys beloved INtel (and I like INtel, but just like, not this fanboy e)

(this is)
seems to have been making bonehead blunders for the last few years...
1) prescot
2) no HT on the Pentium-D's
3) no 64bit in yonah

I would say 3 strikes and you are out but Intel is better then that....They will bounce back and be in the lead briefly then the both of them will share the lead back and forth like it has been in the last 5 years....

Who cares who the winner is...Lets us just care about getting the best cpu...

And you're an AmD fan... If you dont like Dothan, nobody forces you to use it or buy it. Dont like it? Well then leave it alone and purchase FX57 for 1000$. I'll go with Dothan 730 or 750 for 20% of FX57 price and will get about the same perfomance for 800$ cheaper than you will.

As was stated before...

Originally Posted by Astral7

What is truely amazing is that an engine conceived in the summer and fall of 1990 is still being used as the basis of new designs 15 years later. Not only that but beating the top of the line single core amd offering which costs in excess of $1000.00. And doing this on a last generation graphics port and chipsets that are two generations from current. Merom is coming

Thanks Bob

Duvie, please stop flaming, as I said two times already...if you dont like noone forces you to use it, what's the problem? AMD fan I think is the problem.

In my opinion...it doesnt realy matter is it AMD or Intel...it's the performance that matters, both companies offer good CPU's, this time cheap power wins the king of the hill mega price. Price \ performance...that's what's it all about.

[rozzyroz]

[quote=Duvie]Why are the ppl in this site so bad at having the facts straight...[quote]

well duvie, people on this site do have the facts straight. a dothan at 2.4 with a 200mhz fsb will walk all over an a64 at the same speeds. a dothan that is crippled with a low fsb and single channel, wont do as good.

because of people on this site, we can see what a powerhouse the dothan really is. the people on this site dont read the facts... they make them.

[DaWaN]

There won't be any S775--->S479 adapter, never i think. But i think there are coming dothan mobo's with desktop chipset and I hope they gonna kick butt with a Yonah