AMD Dual-Core Toledo Socket939 CPU-Z Pciture

[sxs112]

http://img182.echo.cx/img182/7162/capture0023up2tv.jpg

http://img230.echo.cx/img230/1756/sandracpu27vh.jpg

http://img230.echo.cx/img230/4291/sandracpu39ps.jpg

http://img230.echo.cx/img230/8515/sandracpuartibench24ht.jpg

http://img230.echo.cx/img230/8919/sandraramibench13lm.jpg

http://img230.echo.cx/img230/7317/sandraramibench26kq.jpg


update BIOS

Socket939 :banana:

[tictac]

any overclock result @ default voltage & Overvolt voltage? :hehe:

[sxs112]

any overclock result @ default voltage & Overvolt voltage? :hehe:

ATI RS480 reference board can't OC :(

[tictac]

Why dont test it on DFI nForce4 Ultra?

Is it because the processor wont boot on nForce4 board without proper Processor Microcode Update... :(

[saaya]

nice! 1.3v! :slobber:

[sxs112]

Why dont test it on DFI nForce4 Ultra?

Is it because the processor wont boot on nForce4 board without proper Processor Microcode Update... :(

Because... This is Test in the company.. Cannot flow out

[Chris27]

smuggle it out the back door ;)

[babyelf]

then bring the DFI in lol

[tictac]

Because... This is Test in the company.. Cannot flow out

i see... :fact:

anyway thanks for the info :up:

[Cyprio]

When are these due for general release?

[dng29]

I read somewhere that says next week Apr. 21st is the release of the opteron s940 dual core. June is the release of s939 a64 dual core, no specific date. I'm still waiting on the Venice core and this dual core is coming already. i'll probably wait till end of this year or early next year to upgrade to dual core when longhorn comes out. who's going to take the plunge and get dual core s939?

[reject]

can u get a photo of it, at least a blurry pic so we can see if the cores are both in the sime die, i cant remember if the cores are separate or together in one big die :confused:

[Vincentvega18]

Well hed have to take the ihs off for that........

[saaya]

and they are in one big die, i dont think amd would work so hard and build dual core chips with two cores in one die and make them work, and then build a new dual core cpu that has two single dies...

[Noldor]

Why is it only 512kb cache on it? wasnt it supposed to be 1 MB? or is it just not showing up on cpu-z?

[Ferry82]

Wow men you did it again where do you work? :clap:

[saaya]

Why is it only 512kb cache on it? wasnt it supposed to be 1 MB? or is it just not showing up on cpu-z?

wow, i didnt notice that... maybe its a bug? or maybe its a different chip... maybe there will be 1mb and 512kb dual core chips? :eek:


Wow men you did it again where do you work? :clap:
do you honestly think he will answer that? ;) :D

[$a1Ty]

isnt it 512kb cache per core, so thats just showing one core

[[_Silence_]]

Is it possible for u to run SuperPi/Pifast and Sandra?

[panumas05]

Thank for this info. :banana: :banana: :banana:

[matt9669]

ATI RS480 reference board can't OC :(Hmm, they could at the ATI event in Dallas . . . in fact they had OC options I'd never seen before ;)

1.3V default, that is astounding - though I was under the impression that Toledo was 1MB per core, 512KB is nothing to sneeze at . . . thanks for the info bro! :up:

EDIT: posted this to the FP! :toast:

[Cossey]

how hot do they run compared to the single cores

[Gogeta]

Thanks for the info sxs112! I'm guessing they break 100W...

Anyone know when XP 64bit will be widely available? :D

[ibby]

i know what i am buying...

[Revv23]

this is great, at 2200mhz too, means that they wont be much slower then regular A64's...

plus on 939!

I'm very exited, i was a bit worried after seeing intel's DC stuff, but on 939 with dual cores and a good clock like that, looks like AMD will remain on top for at least a little while longer.

[matt9669]

I'm guessing they break 100W...TDP is supposed to be 95W for the Opteron DC up to 2.2GHz, just FYI - I'd suspect these to be no different . . .

[Quanticles]

You're a brave soul.

[onemorekornkid]

benches? :slobber:

[perkam]

benches?

Doesnt look like it. I imagine it would complicate NDA if he gave us ANY performance numbers.

Perkam

[shuRe]

i might be being a bit dumb but if its 2.2 with dual core then that is 2x1.1 rather than 2x2.2

[bxa121]

what about power output.. can you tel us that?

[[_Silence_]]

i might be being a bit dumb but if its 2.2 with dual core then that is 2x1.1 rather than 2x2.2

It has 2 cores, each @ 2.2Ghzs and maybe 2x 512Kb L2 cache. I think the only shared thing is the memory controller, but not sure.

I know it's impossible for u, but it would be very nice to see it under Phase Change =)=)

[terrace215]

http://imagemac.net/uploads/3fd631e001.jpg

Now, MB BIOS has not written.. Therefore CPUZ cannot recognize
Various cores manage 512Kb

Socket939 :banana:

Then what was the sample shown by the Italian site, which was 939 2.4GHz 1MB L2 for EACH core?

Are there going to be 2 sorts of Toledo parts?

http://www.hwupgrade.it/articoli/1193/1.html

"On board sono integrate due cache L2 distinte, ciascuna in quantitativo pari a 1 Mbyte."

Roughly: On board is integrated two DISTINCT L2 caches, EACH equal to 1MB in size.

[matt9669]

Then what was the sample shown by the Italian site, which was 939 2.4GHz 1MB L2 for EACH core?

Are there going to be 2 sorts of Toledo parts?

http://www.hwupgrade.it/articoli/1193/1.html

"On board sono integrate due cache L2 distinte, ciascuna in quantitativo pari a 1 Mbyte."

Roughly: On board is integrated two DISTINCT L2 caches, EACH equal to 1MB in size.My first inclination is that the Toledo chips are 512MB L2/core, and the Opteron DC (codename?) are 1MB L2/core, which would make sense considering a server/workstation environment makes excellent use of L2 . . .

[ibby]

where can i buy 1 from ?
:D

[terrace215]

My first inclination is that the Toledo chips are 512MB L2/core, and the Opteron DC (codename?) are 1MB L2/core, which would make sense considering a server/workstation environment makes excellent use of L2 . . .

No, no. The Italian site was previewing a 2.4GHz socket 939 DC A64 part (so they claimed), not a DC Opteron part. Of course, like the current FX, and San Diego parts, there is little except the 939/940 packaging to distinguish 1xx Opteron from them.

Perhaps AMD will have both parts? DC A64 2 x 512K L2, and DC A64 2 x 1MB L2... who knows, maybe they'll change their mind and brand the latter a DC A64 FX.

[matt9669]

The Italian site was previewing a 2.4GHz socket 939 DC A64 part (so they claimed)Hmm, Opteron DC's are only slated to 2.2GHz initially - but you're right, perhaps we will see an FX-style incarnation at 1MB/core . . .

AMD has said the FX line will continue to be single core, so they would have to give it a new name it seems like . . .

[terrace215]

Hmm, Opteron DC's are only slated to 2.2GHz initially - but you're right, perhaps we will see an FX-style incarnation at 1MB/core . . .

AMD has said the FX line will continue to be single core, so they would have to give it a new name it seems like . . .

Yeah, AMD did say that... but was that before or after Intel revealed the Smithfield EE brand?

That might change their mind.

[Pjotr]

']It has 2 cores, each @ 2.2Ghzs and maybe 2x 512Kb L2 cache. I think the only shared thing is the memory controller, but not sure.

This is all correct.

[Pjotr]

Yeah, AMD did say that... but was that before or after Intel revealed the Smithfield EE brand?

That might change their mind.

True, but as far as the latest info goes, the dual core 939 CPU will get a new brand name, just like Intel goes Pentium D for their desktop dual core part.

[[_Silence_]]

This is all correct.

Hmm even the 512Kb Cache? Thought that the first Dual Cores were all 1Mb Cache each CPU, or is it only San Diego's?

[Pjotr]

']Hmm even the 512Kb Cache? Thought that the first Dual Cores were all 1Mb Cache each CPU, or is it only San Diego's?

The Opteron dual cores are called Egypt, Italy and Denmark (8, 2, 1 series) and have 1 MB L2 per core. The Toledo (dual core A64 new brand name) is 512 kB per core.

http://www.amd.com/us-en/Processors/ProductInformation/0,,30_118_608,00.html

[matt9669]

The Opteron dual cores are called Egypt, Italy and Denmark (8, 2, 1 series) and have 1 MB L2 per core. The Toledo (dual core A64 new brand name) is 512 kB per core.

http://www.amd.com/us-en/Processors/ProductInformation/0,,30_118_608,00.htmlThanks Pjotr, I know I had seen the roadmap and the codenames for the Opterons but I was too lazy to go look it up :lol2:

[shuRe]

i know its a bit crazy wanting more than 4.4ghz combined, but does anyone have any idea on how these might overclock with the two cores? would they overclock simultaneously or seperatly, would they need real good air cooling to get responable temps?

[Gogeta]

TDP is supposed to be 95W for the Opteron DC up to 2.2GHz, just FYI - I'd suspect these to be no different . . .

I was looking at the opteron stepping comparison charts yesterday on AMD and noticed the E4 revision 200 series were approacing 90W+, so I assumed 2 cores would most likely exceed 100.

Anyways, good to know AMD has done dual core right. :)

[Orthogonal]

Why must we be incessantly teased by these new and unreleased products.

[matt9669]

The cores cannot be clocked separately, and while the cooling requirements won't be drastically higher than single core (according to AMD) I would suggest very good air cooling :cool:


Why must we be incessantly teased by these new and unreleased products.This behind the scenes info bothers you?

[[_Silence_]]

The Opteron dual cores are called Egypt, Italy and Denmark (8, 2, 1 series) and have 1 MB L2 per core. The Toledo (dual core A64 new brand name) is 512 kB per core.

http://www.amd.com/us-en/Processors/ProductInformation/0,,30_118_608,00.html


I had a totally wrong ideia :S Thks =)

[aCidbAbY]

nice!!!!!!! looking forward to seeing the new OC's from this hardware:D

[Bennah]

i know what i am buying...

A P4 660 :p: You like your P4's dont you Ibby ;)

[R.Rabbit]

i know its a bit crazy wanting more than 4.4ghz combined, but does anyone have any idea on how these might overclock with the two cores? would they overclock simultaneously or seperatly, would they need real good air cooling to get responable temps?
it would be a pretty good guess to say that these oc simultaniously because the share a memory controller, and being a new core design(1.3v @ 2.2ghz ;) ), they probably oc pretty well as long as your cooling can take the wattage these babies will put out, as for if you need good air to get reasonable temps.. AMD would say so, but they have a different definition of reasonable temps! so who know..

[kryptobs2000]

I'd think these should overclock not much worse than single cores as long as temps don't become a problem. And it dosn't look like they'd be much hotter, putting out only 95w compared to an venice or so, which as stated, is close to 100w as well. Hell.. maybe it's possible they'd even overclock better. I really doubt that but they do run alot cooler (comparably).

[J-Mag]

I'd think these should overclock not much worse than single cores as long as temps don't become a problem.

Why would you think this? There are obviously more transistors, so that means there are more possible points of failure. As long as the chips are manufactured with the same process single cores on avg will always clock higher than dual cores, IMO. However, if running only single threaded benchmarks, you might see similar clocks since the other half of the transistors won't be touched.

[Fairydust]

Just when I finally decided to buy a San Diego when they become available, this pops up. Guess I'll get a Venice 3200+ and save my money for the Dual cores.

[kryptobs2000]

Why would you think this? There are obviously more transistors, so that means there are more possible points of failure. As long as the chips are manufactured with the same process single cores on avg will always clock higher than dual cores, IMO. However, if running only single threaded benchmarks, you might see similar clocks since the other half of the transistors won't be touched.


Cause a single one of these (afaik you can't really have a 'single' one since they are on the same die but yunno what I mean) should be better than a single venice/san diego. So... just speculation.. dind't mean to sound like gf4ti :p:

[Pjotr]

And the new name for the Athlon 64 dual core CPU is:

Athlon 64 X2


Hexus.net (http://www.hexus.net/) can exclusively reveal that the CPU is to be named the AMD Athlon 64 X2. It will be available at least in the following models:

AMD Athlon 64 X2 4400+
AMD Athlon 64 X2 4600+
AMD Athlon 64 X2 4800+

The flagship AMD Athlon 64 X2 4800+ will have both its cores operating at 2.4GHz, the same frequency as the AMD Athlon 64 4000+, and we'll be surprised if this doesn't rape the already impressive Pentium Extreme Edition 840 we reviewed here.
I have to agree. And what a simple, yet genious name! :slobber:

[kryptobs2000]

I don't like the pr values at all. how does a dual core 2.4ghz get to 4800+? That makes no sense. It should simply be called a X2 4000+ or something.

[terrace215]

I don't like the pr values at all. how does a dual core 2.4ghz get to 4800+? That makes no sense. It should simply be called a X2 4000+ or something.


I don't think it matters.

X2 is cool, though.

[terrace215]

The Opteron dual cores are called Egypt, Italy and Denmark (8, 2, 1 series) and have 1 MB L2 per core. The Toledo (dual core A64 new brand name) is 512 kB per core.

http://www.amd.com/us-en/Processors/ProductInformation/0,,30_118_608,00.html

Where do you get the 512KB per core from?

And in that case, WHAT was tested here, at 2.4GHz, 2 x 1MB L2, socket 939 ???

http://www.hwupgrade.it/articoli/1193/1.html

[Fairydust]

I don't like the pr values at all. how does a dual core 2.4ghz get to 4800+? That makes no sense. It should simply be called a X2 4000+ or something.

Well to me a 4800+ at 2.4 x2 would be welcome. The current PR names are simply for comparison with Intel chips and have no relation to actual processor speed. As Intel entered the realm of useless designations, AMD can return to designations resembling the real speed, although you have to divide by 2.

[$a1Ty]

for the people discussing the temps here, remember that although the watts given out will increase so has the size of the die, we'll have to wait and see how hot they run, and also depending on the program your running will be rare that both cores are at 100%

[kryptobs2000]

Well to me a 4800+ at 2.4 x2 would be welcome. The current PR names are simply for comparison with Intel chips and have no relation to actual processor speed. As Intel entered the realm of useless designations, AMD can return to designations resembling the real speed, although you have to divide by 2.


I just realized 4800+/2 = 2400mhz lol. I see where they get it now. I was thinking they were trying to compare it to a 4800mhz p4 or something. Which it would probably beat anyways :p:

[ozzimark]

i think it may actually be in comparision to an a64 chips for once.. the fact that it's 4800+ for a 2400mhz chip is awfully suspicious ;)

[Orthogonal]

i think it may actually be in comparision to an a64 chips for once.. the fact that it's 4800+ for a 2400mhz chip is awfully suspicious ;)

I never even put that simple observation together but that makes total and perfect sense. :toast:

[BleakVoid]

According to the roadmap (http://www.amd.com/us-en/Processors/ProductInformation/0,,30_118_608,00.html), it would seem that the A64-FX will be using the Toledo core as well...

Though I do know what I want this year...2x 246 X2 Italy cores... :slobber:

[robberbaron]

According to the roadmap (http://www.amd.com/us-en/Processors/ProductInformation/0,,30_118_608,00.html), it would seem that the A64-FX will be using the Toledo core as well...

Though I do know what I want this year...2x 246 X2 Italy cores... :slobber:

You ought to look at the roadmap again. There is no connection between the A64 FX and "Toledo."

[ozzimark]

the line between the a64 and toledo doesn't connect :confused:

edit: snap, i was beat!

[The Coolest]

How about a picture of the chip itself?
If it was dual core, windows should've recognized it as two processors, thus CPU-z should've shown more than just one CPU.
Also the 20FB1 is also weird. On 130nm, FB* CPUID would represent an s939 256KB L2 part, the Paris (physical 256KB L2). But this one has a strange BrandID, regular desktop A64s have BrandID of 4, this one has 5, it also has 512KB of L2 instead of 256KB.
Its odd. we'll have to just wait and see what that chip is.

[Cossey]

the pr rating may also give and idea of cost next to the curent single cores. the pr ratings are all high so expects costs to be higher than the top of the range single cores

[sxs112]

new update

http://www.xtremesystems.org/forums/showthread.php?t=59344&page=1&pp=25

[furyfax]

Nice!

[The Coolest]

Ok, with this update everything makes more sense.
I'm just not sure its a Toledo. There are two types of s939 Dual Core chips, one with 1MB of L2 and the other (this one) with 512KB. One of them is Toledo, the other one is something else, I wonder which is which.
Any chance we could see a pic of the CPU itself, or at least the OPN code and stepping?

[furyfax]

I would like to see a SUPERPI test, if that is possible..

[sxs112]

I would like to see a SUPERPI test, if that is possible..

http://img24.echo.cx/img24/7958/dualcore11xg.jpg

[furyfax]

Is the MP locked ?

[sxs112]

Is the MP locked ?

This TEST is with old BIOS

[actionB52]

i guess the p-rating points toward what we can expect, when multithreading is supported in comparison with a singlecore.

only my speculation.

greetings

[Chris27]

I just realized 4800+/2 = 2400mhz lol. I see where they get it now. I was thinking they were trying to compare it to a 4800mhz p4 or something. Which it would probably beat anyways :p:


I thought AMD was just continuing from the 4200+ since dual cores are "better" they continue from where single cores left off? anyware I hope they are nott too expersive :p:

[Ferry82]

What is the rating from this cpu? :stick:

A: 4400+
B: 4600+
C: 4800+

[andyhuang0308]

My guess....A
don't ask me why I answer this....
By the way... It is really cool while working... roadnews....XD

[longshot]

It's alleady widely available if you know where to look or have a MSDN subscriber ID :)


Thanks for the info sxs112! I'm guessing they break 100W...

Anyone know when XP 64bit will be widely available? :D

[ozzimark]

not sure if it's been posted anywhere else, but...

http://h71016.www7.hp.com/dstore/MiddleFrame.asp?page=config&ProductLineId=431&FamilyId=1256&BaseID=14621&jumpid=ex_r2910_frooglesmb/servers



Processor
Infomation Learn more



(2) AMD® 875 Opteron™ 2.2GHz, Dual Core Processors


edit: it was taken down :(

[karelke]

What is the rating from this cpu? :stick:

A: 4400+
B: 4600+
C: 4800+

4800+ runs @ 2,4; so 2,2 = 4600+

[furyfax]

This is what I believe:
Athlon 64 X2 4400+ = 2x 2200mhz with512kb L2
Athlon 64 X2 4600+ = 2x 2200mhz with 1024kb L2
Athlon 64 X2 4800+ = 2x 2400mhz with 512kb L2
"Athlon 64 X2 5000+" = 2x 2400mhz with1024kb L2 and so on...

[The Coolest]

Actually it'll be X2 44, X2 46, X2 48, etc. Like the FX, no 00+

[karelke]

This is what I believe:
Athlon 64 X2 4400+ = 2x 2200mhz with512kb L2
Athlon 64 X2 4600+ = 2x 2200mhz with 1024kb L2
Athlon 64 X2 4800+ = 2x 2400mhz with 512kb L2
"Athlon 64 X2 5000+" = 2x 2400mhz with1024kb L2 and so on...

Maybe, or Toledo is 512kb only (and 1MB for the Opteron)

So we have 2 options here:

Athlon 64 X2 4400+ = 2x 2200mhz with 512kb L2
Athlon 64 X2 4600+ = 2x 2200mhz with 1024kb L2
Athlon 64 X2 4800+ = 2x 2400mhz with 512kb L2

(like the Sempron)

or

Athlon 64 X2 4400+ = 2x 2000mhz with 512kb L2
Athlon 64 X2 4600+ = 2x 2200mhz with 512kb L2
Athlon 64 X2 4800+ = 2x 2400mhz with 512kb L2

[Chris27]

according to this roadmap that came out today
http://www.c627627.com/AMD/Athlon64/

they have 1mb L2 per core

the 4400+ is 2ghz
the 4600+ is 2.2ghz
and the 4800+ is 2.4ghz

If the 4400+ is priced vs the pentium D at 2.8ghz @ $240 then they should be around $200 (i think) which is a really good deal considering two 3200+ venice cores would cost around $380 :banana: .

[LowRun]

:D

According to this news dual core A64 could take advantage of Hyper Threading: http://www.x86-secret.com/index.php?option=newsd&nid=871

[furyfax]

And for those of us who don't speak french, here's a babelfish translation:

By attentively studying the behavior of Athlon 64 Dual Core, we realized of a very interesting characteristic relating to HyperThreading. Indeed, if the architecture even of K8 cannot support the technology of INTEL, it seems indeed that registers CPUID of the processor are not of this opinion. If one studies moreover meadows the response to instruction CPUID with EAX = 1, one obtains the following values:

Athlon 64 Simple Core - CPUID (EAX = 1)
Hex (EDX) : 0x078bfbff
Bin (EDX) : 0b00000111100010111111101111111111
Hex (EBX) : 0x00000800

Athlon 64 Dual Core - CPUID (EAX = 1)
Hex (EDX) : 0x178bfbff
Bin (EDX) : 0b00010111100010111111101111111111
Hex (EBX) : 0x00020800

The bit 28 which is emphasized here in register EDX corresponds to the support of Hyper Threading. However, it is seen clearly that this bit is now positioned to 1 whereas it was to 0 for Athlon 64 Individual Core traditional. In the same way, the bits [ 23:16 ] of register EBX, which indicate the number of CPU logical supported on CPUs INTEL with Hyperthreading, are also to "2" in the case of Athlon 64 Dual Core. It seems thus that AMD chose to activate the bit "HyperThreading" on these CPUs Athlon 64 Desktop in order to profit from optimizations already misent in place per many programmers for HyperThreading of INTEL. This will thus make it possible Athlon 64 Dual Core to also benefit from work already carried out. At present, we do not know yet if them Opterons Dual Core will also have this active bit.

EDIT: LOL! OK, I see now that the translation was not neccesary! :clap:

[The Coolest]

Yep, I've known about this feature for a couple of weeks now, its a real smart move by AMD, to make programs that support Intel's HT to also support multi core CPUs without any updates to the program itself, so the user benefits from having a dual core CPU, and the programmer of the said program doesn't have to rush to get a version that actually supports multi-core CPUs out to make it compatible.

[terrace215]

:D

According to this news dual core A64 could take advantage of Hyper Threading: http://www.x86-secret.com/index.php?option=newsd&nid=871


Interesting. Makes sense. So folks don't get confused, this would mean that a DC A64 can run two threads at once, even if the second thread is set up via HT technology.

(Not that a DC A64 could run 4 threads at once.)

[c627627]

The apparent existance of both 2x512Kb and 2x1Mb cache Toledos really threw a monkey wrench into the clear Roadmap we thought we had.

[terrace215]

The apparent existance of both 2x512Kb and 2x1Mb cache Toledos really threw a monkey wrench into the clear Roadmap we thought we had.

Your roadmaps have often been too wildly optimistic in terms of AMD speedgrade introductions, but it looks like you are getting back towards reality with this latest one, although FX-61 in Q106 seems highly unlikely, if it ever arrives at all.

[[XC] moddolicous]

Can we see a pic of the chip itself to see if it looks any different than a regular A64? U can blur out the steppings and all, I just want to see if the IHS looks any different (not sure y it would).

[andyhuang0308]

ZDA4400DAA5BV......a message comes from no where?....
ADBHE XXXXXXXX
PS:botton line: 1.30V
another roadnews...m....

[STEvil]

One dualcore is based on Venice (2x512).

The other dualcore is based on San Diego (2x1024).

Cant wait for laptops with dualcore to start showing up.....

[Fzero]

hmm im wondering how these will do in graphic benchmarks they seem very powerfull and i love how low the voltage is cant wait to see more info on them :)

[c627627]

Your roadmaps have often been too wildly optimistic in terms of AMD speedgrade introductions, but it looks like you are getting back towards reality with this latest one, although FX-61 in Q106 seems highly unlikely, if it ever arrives at all.

Believe it or not, that roadmap is not based on emotion.

Here's the thought process on the particular point you raised so a counter argument could be offered:

Since it appeared that we would see a 3 GHz 90nm single-core Opteron in Q1 2006, it was safe to assume that we would see a 3 GHz Athlon 64 FX and by extension 3.2 GHz Athlon 64 FX.

Ed Stroligo of overclockers.com seemed to think so and he's got a pretty good conservative track record.

His views from a couple of weeks ago mentioning this very argument are here: "Here Comes 3" (http://www.overclockers.com/tips00753/)
.
.

[sxs112]

http://img215.echo.cx/img215/4972/pic14ly.jpg
New BIOS

Athlon64 X2 Dual Core Processor 2200+(4400+)

[shuRe]

gwan sxs112 if you can post all these screenies of the toledo then you can do a little bit of overclocking on stock cooler. plz :)

[c627627]

Thank you sxs112.
http://www.c627627.com/AMD/Athlon64/
(Refresh)

[furyfax]

gwan sxs112 if you can post all these screenies of the toledo then you can do a little bit of overclocking on stock cooler. plz :)

Don't think he can OC because of the mobo..

[shuRe]

Don't think he can OC because of the mobo..
my bad

[The Coolest]

sxs112, why don't you give GCPUID a shot, see how it works with your new chip?
Download GCPUID here (http://www.thecoolest.zerobrains.com/apps/GCPUID64.zip)

[sxs112]

http://img162.echo.cx/img162/4127/toledo21iw.jpg

[reject]

the cache is shared between the cores, physically its one block of cache from what i have seen, and the cores use HT links to access it

[The Coolest]

No its not, they are two seperate cores.
How hard is it to believe? they are two seperate K8 Venice/San Diego cores connected at the memory controller.

[ozzimark]

the cache is shared between the cores, physically its one block of cache from what i have seen, and the cores use HT links to access it
sorry. but there is no way that's right. why would they take up die space making an slow HT link between the cache and core now when they never did before?
it's two seperate caches. i think all the cores will be made with 1mb+1mb, but any of them with a defect in the cache (likely, due to size) get cut down to 512kb+512kb

[Chris27]

hmm so according to the newer roadmap only the 4800+ has 2mb of L2 and both the 4600+ and the 4800+ are 2.4 ghz.

[The Coolest]

sorry. but there is no way that's right. why would they take up die space making an slow HT link between the cache and core now when they never did before?
it's two seperate caches. i think all the cores will be made with 1mb+1mb, but any of them with a defect in the cache (likely, due to size) get cut down to 512kb+512kb
Look at the picture in the post above yours. You're right, these are 2 seperate processing cores, with their own dedicated L1 and L2 caches, and the only place they meet is the memory controller.
I have to disagree with you on your statement that all cores will be 1mb+1mb, when clearly, already its not the case.
20F30 is a CPUID reading of a dual core 1MB+1MB CPU, the new DC Opteron have a 20F10 CPUID reading. The chip we see in this thread is 20FB1, which is not similar to either of these CPUIDs.
The 20F30 and 20F10 are like 2 San Diegos stuck together, this 20FB1 is like 2 Venices stuck together, there is no disabled L2 cache on this chip. I don't think I've ever seen an ES (and this is a semi-ES as it doesn't have a model number programmed into it, that's why the BIOS reads it as 2200+) that had disabled cache, or been disabled by any other means.

[Pjotr]

Look at the picture in the post above yours. You're right, these are 2 seperate processing cores, with their own dedicated L1 and L2 caches, and the only place they meet is the memory controller.

No, they meet at the crossbar, the System Request Queue. This in turn is connected to the HT links and the memory controller. So CPU-CPU cache snooping traffic never goes out to the HT links, like in a dual CPU system. This improves performance slightly.



MC HT
\ /
SRQ
/ \
core1 core2

[Pjotr]

:D

According to this news dual core A64 could take advantage of Hyper Threading: http://www.x86-secret.com/index.php?option=newsd&nid=871

Not exactly. An AMD dual core CPU will be able to *appear* as a single core with HT support. This is because some morons develop software for HyperThreading, but fail to spawn threads for multi-CPU. By appearing as a HT CPU an AMD dual core would be able to take advantage for HT-aware Dual-UNaware software.

[The Coolest]

Pjotr>> Yes you're right on the fact that there's System Request Queue between the cores and the memory controller. But what I wanted to say basically is that the CPUs are still two completely seperate cores.
HyperThreading is relatively old tech, dual core is relatively new, not all developers that were writing their programs for multi CPU support, never knew about Intel and AMD releasing multi-core CPUs in the future, so this feature allows older software to still have a noticable performance increase over running on a single core.
So I wouldn't call developers morons just because they didn't know about this.

[Napoleonic]

my question is, how can we overclocked it?

independent clock speed for each core, or just one for both? ummm I think 1, since it only have 1 mem controller :) and independent clock speed adjustment for each core is somewhat not logic

[terrace215]

Believe it or not, that roadmap is not based on emotion.

Here's the thought process on the particular point you raised so a counter argument could be offered:

Since it appeared that we would see a 3 GHz 90nm single-core Opteron in Q1 2006, it was safe to assume that we would see a 3 GHz Athlon 64 FX and by extension 3.2 GHz Athlon 64 FX.

Ed Stroligo of overclockers.com seemed to think so and he's got a pretty good conservative track record.

His views from a couple of weeks ago mentioning this very argument are here: "Here Comes 3" (http://www.overclockers.com/tips00753/)
.
.


To sell a part at 3.2, it needs to run at 3.4, at least, with no problems. This is known as a "guard band". I'm not sure 90nm will deliver that, especially by Q106.

It's also pure speculation, not really a "roadmap". Perhaps you should cite the evidence you have for each prediction on your website with a link to an explanation somewhere else on your site. That way, folks can distinguish between supposedly leaked roadmaps and speculation.

Oh, and I wouldn't rely on the Ediot for anything. :)

[terrace215]

New BIOS

Athlon64 X2 Dual Core Processor 2200+(4400+)

Well this is strange.

If 4400 = 2 x (2.2GHz, 512K L2)
and (speculating) 4800 = 2 x (2.4GHz, 512K L2)

What the heck is 4600? Are they really going to release 2.3GHz cores? I guess that's possible, and perhaps makes some sense, even.

The alternative of: 4600 = 2 x (2.4, 512K L2), 4800 = 2 x (2.4, 1MB L2) is possible, but it doesn't make as much sense from the model number viewpoint.

[[_Silence_]]

Well this is strange.

If 4400 = 2 x (2.2GHz, 512K L2)
and (speculating) 4800 = 2 x (2.4GHz, 512K L2)

What the heck is 4600? Are they really going to release 2.3GHz cores? I guess that's possible, and perhaps makes some sense, even.

The alternative of: 4600 = 2 x (2.4, 512K L2), 4800 = 2 x (2.4, 1MB L2) is possible, but it doesn't make as much sense from the model number viewpoint.

I think 4600+ is 2.4Ghzs 512Kb cache and 4800+ 2.4Ghzs 1Mb Cache.

[The Coolest]

Well this is strange.

If 4400 = 2 x (2.2GHz, 512K L2)
and (speculating) 4800 = 2 x (2.4GHz, 512K L2)

What the heck is 4600? Are they really going to release 2.3GHz cores? I guess that's possible, and perhaps makes some sense, even.

The alternative of: 4600 = 2 x (2.4, 512K L2), 4800 = 2 x (2.4, 1MB L2) is possible, but it doesn't make as much sense from the model number viewpoint.
Its very possible that this is what's going to happen:
4200+ = 2.0GHz 2x512KB <-- Might not happen at all.
4400+ = 2.2GHz 2x512KB
4600+ = 2.4GHz 2x512KB
4800+ = 2.4GHz 2x1MB

I think that at least for now, both of the dual core CPUs are going to be called Toledo, as we don't have any other info on this.

[Orthogonal]

or, it could be that
4400+ = 2.2 GHz 2x512KB
4600+ = 2.2 Ghz 2x1MB
4800+ = 2.4 Ghz 2x512mb
and maybe a higher version later on with 2x1MB

This is analagous to the new Venice and San Diego cores w/
3500+ = 2.2 Ghz 512k
3700+ = 2.2 Ghz 1MB
3800+ = 2.4 Ghz 512k

Just MHO.

[terrace215]

or, it could be that
4400+ = 2.2 GHz 2x512KB
4600+ = 2.2 Ghz 2x1MB
4800+ = 2.4 Ghz 2x512mb
and maybe a higher version later on with 2x1MB

This is analagous to the new Venice and San Diego cores w/
3500+ = 2.2 Ghz 512k
3700+ = 2.2 Ghz 1MB
3800+ = 2.4 Ghz 512k

Just MHO.

Not if you accept the italian site CPUZ shot, which featured:

2.4GHz 2x1MB

It would be odd to be sampling a part this far in advance that wasn't going to be part of the launch (and thus probably the 2400+/4800+ A64 X2)

[Orthogonal]

Of course I could be way off my rocker. If AMD is going to have the 4800+ as their new Flagship, it would most certainly have 2x1MB, if any at all have that much.

[Pjotr]

HyperThreading is relatively old tech, dual core is relatively new, not all developers that were writing their programs for multi CPU support, never knew about Intel and AMD releasing multi-core CPUs in the future, so this feature allows older software to still have a noticable performance increase over running on a single core.

Yes. But HyperThreading is rather new, while dual core is over ten years old. It's not a new phenomenon.


So I wouldn't call developers morons just because they didn't know about this.

I'm a developer and I would. Anyone programming threads based on number of CPUs *should* ask the system how many CPUs there are. They should *not* check if there is HT and forget to check for multiple CPUs. That's just stupid and I would slap my developers if they did that. :slapass:

[The Coolest]

Have you actually take the time to read how dual core works?
on the Intel side, which has HT, makes it a bit more complex:
First find out how many processors windows detects.
Then find out how many cores a single chip has.
Then find out how many virtual CPUs a single chip has.
Then find out which CPU is a virtual and which one is physical.
Build a list of which virtual CPU belongs to which physical, then make softare make the correct decision on what specific CPU to start a new thread....

[Adamantine]

Except programmers are stereotypically lazy and will ask if the machine uses HT and not do any other checks. Which is obviously stupid/bad programming.

[Pomme]

Except programmers are stereotypically lazy and will ask if the machine uses HT and not do any other checks. Which is obviously stupid/bad programming.

if you program multithreaded, this shouldn't be a problem ?

compilers don't have to be adjusted either, do they ?or am i mistaking here ?

[Rys]

So CPU-Z v1.28.4 shows the core number in the dropdown? I didn't see that with 1.28.3 when I tested.

X2 4800+ is definitely 2.4GHz with 1MiB L2, per core. You'll see next week ;)

[perkam]

I thought A64 had QuantiSpeed not Hyper threading..which was for Intel :confused:

But other than that, can some PLS PLS PLS PLS PLS PLS PLEASE tell me if its confirmed that Dual Core will be coming 939 ???? :eek:

Perkam

[The Coolest]

This chip is a s939 CPU.

[perkam]

This chip is a s939 CPU.

I know that !!! But will we have to change chipsets just to use it???

Perkam

[jjcom]

No, if the MB makers release a BIOS that will work with the dual cores all should be well. The dual core chips should work on any S939 board. Should.

jjcom

[Rys]

If the board can support an FX processor, it can support a dual-core CPU.

[The Coolest]

I know that !!! But will we have to change chipsets just to use it???

Perkam
Basically all s939 are supposed to support them. I'm pretty sure that with a BIOS update there won't be much boards that won't be able to run them.

Rys>> YGPM

[PCcrazy]

If the board can support an FX processor, it can support a dual-core CPU.

And i believe all 939 board support the FX though, so dual core should support all 939 board out there...

[terrace215]

So CPU-Z v1.28.4 shows the core number in the dropdown? I didn't see that with 1.28.3 when I tested.

X2 4800+ is definitely 2.4GHz with 1MiB L2, per core. You'll see next week ;)


Cool. So does that mean the X2 4600+ is 2.4GHz 512K L2, I guess?

EDIT: Or perhaps more likely is:

X2 4800+ = 2.4 1MB
X2 4600+ = 2.2 1MB
X2 4400+ = 2.2 512K

That makes more sense. All produced using the same die for DC Opteron. The 2.2 512K part has 1/2-cache disabled.

And if one core is bad, those become San Diego 4000+, 3700+ and "Venice" 3500+.

[c627627]

The reason it does not make more sense is what, then, would a hypothetical X2 4x00+ = 2.4 512K be?

[Grayskull]

Can't O/C? Send me a PM and let me know what board you have.


ATI RS480 reference board can't OC :(

[terrace215]

The reason it does not make more sense is what, then, would a hypothetical X2 4x00+ = 2.4 512K be?

Why, 4600+.

There are different ways of getting to the same PR.

[Hallowed]

sxs: In your opinion does Dualcore make any difference in your normal PC use? Windows feel snappier, etc?

Thanks for the new info again man, very appreciated.

[andyhuang0308]

It's much more like u r using an INTEL HT CPU on your PC, the benefit is much better while multi-tasking. But with very low heat, much lower then you even can image while FX and P4 are out there. I know the test was done under air-cooling with a normal Semprone (754) cooler. There is a reason to use RS480 M/B testing this, but couldn't say much about it. It's Bios limitation not to allow OC on this M/B, not the chipset itself.
OK. above are all I have heard from someone.... but all facts.

[perkam]

Cool. So does that mean the X2 4600+ is 2.4GHz 512K L2, I guess?

EDIT: Or perhaps more likely is:

X2 4800+ = 2.4 1MB
X2 4600+ = 2.2 1MB
X2 4400+ = 2.2 512K

That makes more sense. All produced using the same die for DC Opteron. The 2.2 512K part has 1/2-cache disabled.

And if one core is bad, those become San Diego 4000+, 3700+ and "Venice" 3500+.

Should more along the lines of:

A64 X2 4200+ / 2.2 Ghz 512K Dual Core
A64 X2 4400+ / 2.2 Ghz 1MB Dual Core
A64 X2 4600+ / 2.4 Ghz 512K Dual Core
A64 X2 4800+ / 2.4 Ghz 1MB Dual Core
A64 X2 5000+ / 2.6 Ghz 512KB Dual Core
A64 X2 5200+ / 2.6 Ghz 1MB Dual Core

Which is consistent at all levels...

Perkam

[terrace215]

Should more along the lines of:

A64 X2 4200+ / 2.2 Ghz 512K Dual Core
A64 X2 4400+ / 2.2 Ghz 1MB Dual Core
A64 X2 4600+ / 2.4 Ghz 512K Dual Core
A64 X2 4800+ / 2.4 Ghz 1MB Dual Core
A64 X2 5000+ / 2.6 Ghz 512KB Dual Core
A64 X2 5200+ / 2.6 Ghz 1MB Dual Core

Which is consistent at all levels...

Perkam

That's not how it is. We have two data points: 4400+ = 2.2 512K, 4800+ = 2.4 1MB. The only question is the 4600+.

Also, that doesn't necessarily make sense. The contributions of +200MHz and +512K L2 may not be equal "on average" to performance.

See the 3500+ Venice(2.2/512K) / 3700+ San Diego(2.2/1MB) / 3800+ Venice (2.4/1MB) series for a real-world example.

Given that, IMO, all of the initial X2 parts will be the same (large) die as DC opteron (with the 4400+ having 1/2-cache *disabled*), it would make more sense for the 4600+ to be 2.2GHz 1MBL2.

[shadowing]

Holy moly... I just realized the memory speed of that thing. :slobber:

[c627627]

Hm... we may find out tomorrow that sxs112 had an early sample whose PR rating was incorrect.

http://www.theinquirer.net/?article=22676

[M1nDH4cK]

http://www.theinquirer.net/?article=22685

So it's finally settled. Huge prices though. The FX-55 still retains the leadership as far as gaming performance is conceirned.

[OC Detective]

Looks like Andy was wrong about it being a 4400+? Some new glasses perhaps?

[andyhuang0308]

Doesn't matter to me... I just a visitor.... ^_^

[Napoleonic]

Anandtech already made benches with X2 processor.......

[Rys]

They used an Opteron, which is a different socket with different memory. Nobody has officially reviewed an X2, yet.

[Woody-san]

AMD X2 Dual Core manufacturer published pricing in Tokyo:

Socket 939 (95WTDP/80A):
1) 4800 2.4GHz. 1MB + 1MB ~US$1000 (Y111K).
2) 4600 2.4GHz. 512KB + 512KB ~US$880 (Y89K).
3) 4400 2.2GHz. 1MB + 1MB ~US$630 (Y65K).
4) 4200 2.2GHz. 512KB + 512KB ~US$590 (Y60K).

Availability will be in June 2005(corrected!) Samples are available to select customers ...

:woot:

[Napoleonic]

June 2006???? :confused:

[Kunaak]

that mips shot from stock is just beautiful :)

[dion.nex]

Does anyone know about opteron 270 .. which have agreement from corel company

[Revv23]

Does anyone know about opteron 270 .. which have agreement from corel company


holy thread bump!

what do you want to know.