• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Fudz on K10 : 45W TDP, Budapest, Shanghai

NoobyDoo

Senior member
Some Barcelona K10s to have 45W TDP

... At 65 nanometre native quad core AMD can hit 45W TDP ... The core that can hit these speeds will be clocked to 1.9 GHz or less ... Some parts will have 68W TDP ...this means that these parts will be supper overclockable after all ...

Budapest is K10 with HT3

After Barcelona comes Budapest ... in socket AM2+ flavour and will be known as Opteron 1000 series. It will have Hypertransport 3.0 ...

AMD 45 nm Shanghai up and running

AMD already has some prototypes of Shanghai native quad core at 45 nm and 6 MB of L3 cache. This unit won't go in production till sometimes next year but at least the marchitecture works.
Sources close to company also confirmed that AMD can 45 nanometre parts without changing the fabs so the usual suspect Dresden fab will get a line to produce these parts.

Intel to challenge K10 with 45 nm parts

Intel knows that AMD has a very powerful CPU and that K10 has a very good performance per watt figures. It also scores really well for its clock speed.
But Intel can go beyond 3GHz as soon as it goes to 45 nanometre. Harpertown the first Intel's 45 nm server part is scheduled for Q4 2007 but the real game starts with Yorkfield in Q1 2008.
AMD should not be that far away with its 45 nanometre production as it is currently on the schedule to start the product in first half of 2008. With Shanghai and Montreal out the game is on, one more time.


 
Too bad Armageddon is at Q2 '08 and all this would amount to nothing.**






** Note: precise time of Armageddon may vary.
 
Originally posted by: NoobyDoo
Some Barcelona K10s to have 45W TDP

... At 65 nanometre native quad core AMD can hit 45W TDP ... The core that can hit these speeds will be clocked to 1.9 GHz or less ... Some parts will have 68W TDP ...this means that these parts will be supper overclockable after all ...

Budapest is K10 with HT3

After Barcelona comes Budapest ... in socket AM2+ flavour and will be known as Opteron 1000 series. It will have Hypertransport 3.0 ...

AMD 45 nm Shanghai up and running

AMD already has some prototypes of Shanghai native quad core at 45 nm and 6 MB of L3 cache. This unit won't go in production till sometimes next year but at least the marchitecture works.
Sources close to company also confirmed that AMD can 45 nanometre parts without changing the fabs so the usual suspect Dresden fab will get a line to produce these parts.

Intel to challenge K10 with 45 nm parts

Intel knows that AMD has a very powerful CPU and that K10 has a very good performance per watt figures. It also scores really well for its clock speed.
But Intel can go beyond 3GHz as soon as it goes to 45 nanometre. Harpertown the first Intel's 45 nm server part is scheduled for Q4 2007 but the real game starts with Yorkfield in Q1 2008.
AMD should not be that far away with its 45 nanometre production as it is currently on the schedule to start the product in first half of 2008. With Shanghai and Montreal out the game is on, one more time.


You have another source says that AMD has had K10 working on their 45nm process.
Fudzilla doesn't count . When he doesn't list a source.
I bet this 45nm process is HighK with metal gates also right?

AMD hasn't demoed K10 on 65 nm yetyou except this as a fact. Your amusing!
 
Originally posted by: Nemesis 1
You have another source says that AMD has had K10 working on their 45nm process.
Fudzilla doesn't count . When he doesn't list a source.
I bet this 45nm process is HighK with metal gates also right?

AMD hasn't demoed K10 on 65 nm yetyou except this as a fact. Your amusing!

Huh? The K10 HAS been demoed, just under extreme NDAs so very little info has gotten out.

On the 45nm side, 45nm prototypes are not completely out of this realm, the problem is bringing them into full production (since this requires AMD to retrofit a fab, likely fab30, which requires massive capital expenditures). Another thing to think about is clockspeed. I doubt AMD can get much out of K10 in regards to clock speed unless its pipeline was lengthened, so Intel will have a huge advantage just by being able to hike up its clocks closer to 4GHz. The asynchronous northbridge/CPU core configuration will likely help overclockers (if they are given full control over it) but AMD will probably need to stick to synchronous clock speeds at stock.
 
Agena, K10 pictured running

We told you before K10, Agena works just fine. We had a chance to witness it and we can even prove the existence of this 65 nanometre part. The Agena processor is an AM2 version and fits in 940 pin boards.
... you can clearly see that this is a 65 nanometre part with 1.273V, with four cores.
It works at 1.9 GHz ... with multiplier at 9.5 and the bus speed of 200.9 MHz. ... Agena at 1.9 GHz has 4x64KB L1 Data cache memory, 4x64KB L1 for Code, 4x 512 KB L2 cache and finally all cores share the 2048KB L3 cache.

 
Interesting You post a link saying Agena runs at 2.5 and O/C to 3.0 ghz easily on stock volts . Than you show proof of something running @ 1.9 ghz. Thats hardly proof of this outstanding O/C from a 4 core die.
 
Originally posted by: NoobyDoo
Agena 1.9 GHz overclocks to 3.05 GHz

... With the default multiplier set to 9.5, you can raise the FSB all the way up to 320 MHz.
You only need to increase the voltage from 1.273 to 1.297V and you hit over 3 GHz or 3042 MHz to be precise. This is an 1142 MHz overclock or 62 per cent from the stock speed and the machine was stable at all times. ...

Not stock volts, though. I wonder what the power draw is at that clock... If the 2.3GHz part really gives the 2.66GHz kentsfield a run for its money then a 3GHz part may beat whatever Intel can get out of Penryn, TDP permitting. Of course, Penryn is probably a Late Q3/Early Q4 part, and only at the high-end, so AMD's competition for half a year may be plain-old Kentsfields.
 
Very interesting. A member at XS has confessed to sending those slides to fudsilla . He said it didn't take long to photoshop it . Its a fake . Name of cpu is wrong yet it shows L# cache. CPU_Z wouldn't show the l3 cache.

How is it that AND K10 will have a 6 month advantage over Penryn ?

Are K10's out on the market yet? When will they be out ?

AMD has hugh fab capacity . NOT

You have K10 beating Penryn . But we have seen performance numbers for Penryn on an old chipset . Running @ DDR2 - 800 with 5-5-5-12 timing . I wonder what those numbers will look like at DDR3 1333. with 5-5-5-12 timings on a chipset designed for Penryn. Skulltrail will use regular DDR3 memory not FB dimms

I just talked to a close relative that works at intel . He said Skulltrail will have 2 front side buses using DDR3 1600. and the Processors used with it will be running at the 3.66ghz range.
Skull trail will keep Intel in the performance lead .

As for the performance of K10 . Lets wait and see . AMD isn't in a hurry to show its performance . Why should they its not like their bleeding money all over the place.
 
Originally posted by: Furen
Originally posted by: NoobyDoo
Agena 1.9 GHz overclocks to 3.05 GHz

... With the default multiplier set to 9.5, you can raise the FSB all the way up to 320 MHz.
You only need to increase the voltage from 1.273 to 1.297V and you hit over 3 GHz or 3042 MHz to be precise. This is an 1142 MHz overclock or 62 per cent from the stock speed and the machine was stable at all times. ...

Not stock volts, though. I wonder what the power draw is at that clock... If the 2.3GHz part really gives the 2.66GHz kentsfield a run for its money then a 3GHz part may beat whatever Intel can get out of Penryn, TDP permitting. Of course, Penryn is probably a Late Q3/Early Q4 part, and only at the high-end, so AMD's competition for half a year may be plain-old Kentsfields.

Agena is simply not due to Q4 2007.

EDIT: Linkage

http://www.cpilive.net/v3/inside.aspx?scr=n&NID=1320
 
Originally posted by: yacoub
Originally posted by: Nemesis 1
ZOMG EVERYTHING AMD IS FAKE RUN TO INTEL THEY ARE TEH WIN I AM CAPTAIN NEGATIVE


I am not being negitive at all. Your being overly optimistic. If K10 was all that . AMD would show some real benchmarks. Like they did with Hammer. AMD needs to show K10 performance NOW . They are bleeding money fast . By the end of qt 2 they will have less than 600 million in cash reserves. K10 still won't be out .

Qt. 3 Intel does another price cut. If X2 3600 is such a good deal now . What will Qc2d be at $266? What are the yields going to be like on a native Q core . I bet not very good. When K10 is finally released the ramp up will be slow. and painfull.

By the end of qt.3 AMD will have no cash reserves. The only reason were not seeing K10 benchies is its not all that good. Now this is based on common sense . Not some fanboy desire for K10 to be = to the second coming of Christ.

 
Originally posted by: Nemesis 1
Originally posted by: yacoub
Originally posted by: Nemesis 1
ZOMG EVERYTHING AMD IS FAKE RUN TO INTEL THEY ARE TEH WIN I AM CAPTAIN NEGATIVE


I am not being negitive at all. Your being overly optimistic.

ahahaha, wait how am I being overly optimistic when I haven't said anything in this thread about a yet-to-be-released product, yet you have commented on it and make it seem like the world is going to end? 😀
 
Originally posted by: Nemesis 1
Very interesting. A member at XS has confessed to sending those slides to fudsilla . He said it didn't take long to photoshop it . Its a fake . Name of cpu is wrong yet it shows L# cache. CPU_Z wouldn't show the l3 cache.

I looked over at XS and found nothing. Link please.
 
Originally posted by: yacoub
Originally posted by: Nemesis 1
Originally posted by: yacoub
Originally posted by: Nemesis 1
ZOMG EVERYTHING AMD IS FAKE RUN TO INTEL THEY ARE TEH WIN I AM CAPTAIN NEGATIVE


I am not being negitive at all. Your being overly optimistic.

ahahaha, wait how am I being overly optimistic when I haven't said anything in this thread about a yet-to-be-released product, yet you have commented on it and make it seem like the world is going to end? 😀


yacoub, please don't feed the Trolls...🙂
 
Originally posted by: Nemesis 1
Originally posted by: yacoub
Originally posted by: Nemesis 1
ZOMG EVERYTHING AMD IS FAKE RUN TO INTEL THEY ARE TEH WIN I AM CAPTAIN NEGATIVE


I am not being negitive at all. Your being overly optimistic. If K10 was all that . AMD would show some real benchmarks. Like they did with Hammer. AMD needs to show K10 performance NOW . They are bleeding money fast . By the end of qt 2 they will have less than 600 million in cash reserves. K10 still won't be out .

Qt. 3 Intel does another price cut. If X2 3600 is such a good deal now . What will Qc2d be at $266? What are the yields going to be like on a native Q core . I bet not very good. When K10 is finally released the ramp up will be slow. and painfull.

By the end of qt.3 AMD will have no cash reserves. The only reason were not seeing K10 benchies is its not all that good. Now this is based on common sense . Not some fanboy desire for K10 to be = to the second coming of Christ.


Nemesis, fanbois are a good thing in moderation. :Q They could possibly keep undeserving companies alive (if they actually vote their money where their mouth is) thru tough times so that the rest of us can enjoy the fruits of competition. 😎
 
Back
Top