LOL well I certainly hope it's true, if not then TheINQ will end up with egg on their faces (not like they don't have enough eggs on their faces already :lol: ).
But this rumor can neither be proved or disproved unless AMD makes an official statement or we wait till launch date to find out. :?
But I bet Charlie 'bunny AMD fanboy' Demerjian would be all giddy right about now, you can tell from the wording he is estatic atm. :lol: :lol:
Viperabyss
Distinguished
- Mar 7, 2006
- 573
- 0
- 18,980
I think there's more needed to be done in order to give K10 a speedboost. IMO, AMD's current 65nm process may be hard to scale compared to Intel's offering, is mainly due to the process, not the memory controller. As we can see from K8, the highest stock speed released by AMD so far is 3.0Ghz. Enthusiasts generally cannot clock K8 higher than 3.6 without the use of exotic cooling (Phase, LN2, etc).
As a result, IMO, AMD still needs to improve their process before they can benefit from this memory controller upgrade.
m25
Distinguished
- May 23, 2006
- 2,363
- 0
- 19,780
If AMD will fail competing with intel in the future, that will only be due to technical/financial issues, because with Intel's aggressive development cycle and their detailed roadmap up to 2009, I don't think AMD in in the position to underestimate anything.
As for the 2.8 K10 clockrates; AMD's 65nm SOI has been around for a while, and despite the problems they had in the beginning, launching a totally new architecture that is supposed to be 70% more efficient than the older, almost at the same clock rates as the older one is pretty satisfactory.
As for the 2.8 K10 clockrates; AMD's 65nm SOI has been around for a while, and despite the problems they had in the beginning, launching a totally new architecture that is supposed to be 70% more efficient than the older, almost at the same clock rates as the older one is pretty satisfactory.
MU_Engineer
Splendid
- Feb 18, 2006
- 3,868
- 4
- 23,165
Well, if that's true, then that (probably fake) CPU-Z shot of a Barcelona may not have been that far off of reality. But once you think about it, this does sound plausible enough. The 90 nm dual-core parts went up to 3.0 GHz. The K10 is a rework of the K8, but nothing that should greatly depress clock speeds (e.g. NetBurst -> Core 2.) I'm perhaps thinking that the 2.6-2.9 GHz quads will be 125 watt parts, but the fact that AMD might have clock speed parity with Intel while claiming a big IPC advantage will be good for their sagging bottom line as well as performance enthusiasts.
Of course, this is the Inquirer and it may or may not be true. We shall wait and see.
Of course, this is the Inquirer and it may or may not be true. We shall wait and see.
[quote="TrackIf it dosent end up being a huge success we may have to buy our GPU/CPUs from Intel-nVidia.[/quote]
It's already begun, I believe. Look at the 8800 "Ultra". An overclock and branded as a new card? And the 8600? Shittyness on silicon. Nvidia has already decided it doesn't need to do anything until when (more likely if) Ati manage to catch up, maybe even overtake Nvidia.
Intel, however looks like it has learned its lesson, and isn't slacking, for now (thank god).
It's already begun, I believe. Look at the 8800 "Ultra". An overclock and branded as a new card? And the 8600? Shittyness on silicon. Nvidia has already decided it doesn't need to do anything until when (more likely if) Ati manage to catch up, maybe even overtake Nvidia.
Intel, however looks like it has learned its lesson, and isn't slacking, for now (thank god).
BaronMatrix
Splendid
- Dec 14, 2005
- 6,655
- 0
- 25,790
I think there's more needed to be done in order to give K10 a speedboost. IMO, AMD's current 65nm process may be hard to scale compared to Intel's offering, is mainly due to the process, not the memory controller. As we can see from K8, the highest stock speed released by AMD so far is 3.0Ghz. Enthusiasts generally cannot clock K8 higher than 3.6 without the use of exotic cooling (Phase, LN2, etc).
As a result, IMO, AMD still needs to improve their process before they can benefit from this memory controller upgrade.
It looks like it says the B0 stepping FIXED the IMC bottleneck, not upgraded it. Tell Jack to find the latest numbers for it. I'm sure CTI works and they ARE improving the process. It didn't take long for 90nm revs to do big things and I expect a similar thing with 65nm.
And also you have to remember that Brisbane is a shrink while Barcelona was designed for 65nm. I expect that to make a noticeable difference.
There needs to be a sticky about posting stuff from The Inq. :roll:
Anyways, what have we actually heard from AMD concerning the release speeds of K10?[/b]
What % of inq reports regarding AMD has proven true or false so far?
If it's 100% false I may agree with you.
ritesh_laud
Distinguished
- Nov 16, 2001
- 456
- 1
- 18,780
Unlike The Inquirer's misguided hype about AMD reverse hyperthreading last year, I think there may actually be a grain of truth in this one. Let's hope it is, it'll make 2008 a lot more interesting. I think all of 2007 is likely to be dominated by Intel, since AMD likely won't be able to make enough Barcelonas to compete.
In 2008 Intel floods the market with cheap quad-core Penryns, against AMD's high IPC Barcelona that might have the performance crown but can't be priced competitively due to the large SOI die.
BaronMatrix
Splendid
- Dec 14, 2005
- 6,655
- 0
- 25,790
Unlike The Inquirer's misguided hype about AMD reverse hyperthreading last year, I think there may actually be a grain of truth in this one. Let's hope it is, it'll make 2008 a lot more interesting. I think all of 2007 is likely to be dominated by Intel, since AMD likely won't be able to make enough Barcelonas to compete.
In 2008 Intel floods the market with cheap quad-core Penryns, against AMD's high IPC Barcelona that might have the performance crown but can't be priced competitively due to the large SOI die.
Opteron dual is 223mm^2
Barcelona quad is 283mm^2
That's not that big a difference. Besides, Barc is server and Kuma will be the volume SKU. Kuma should be no bigger than the current Windsor.
If IPC is above 20% then it will command a premium for the high clocks while the low clocks can be on price parity with C2Q.
ritesh_laud
Distinguished
- Nov 16, 2001
- 456
- 1
- 18,780
223 mm^2 is a large die. This is a big reason AMD lost so money much last quarter, when their gross margins tanked to 28.1%. ~$80 ASP simply isn't enough to pay for that large a die with sufficient margins to keep the business afloat.
283 mm^2 will only make the problem worse. You can bet Intel will continue the pricing pressure; already in Q3 they are slated to offer quad cores at $266. You are right that AMD may be able to leverage their performance advantage to raise ASPs somewhat, let's hope it's enough. But if Intel clocks Penryn to the moon, bye bye AMD.
turpit
Splendid
- Feb 12, 2006
- 6,373
- 0
- 25,780
indeed
But as to charlie having penned it and claimed he came up with the k8L name :? Charlies credibility is as good as Bush's WMD in Iraq or Nixon at watergate.Charlie is to news what clinton is to the press about lewinsky during the first week of denial. Hows that cigar charlie? :roll:
Lets tear apart the fact that he doesnt nail down the architectural changes that could support his anticipation that the clock speeds will even remotely be competitive.
AMD and Intel have shown that clock speeds mean diddly if the architechture cannot utilize it efficiently. Happy people in austin could just be listening to SXSW cd's at work and having a common local favorite hitting a great stride towards national stardom.
Even if AMD can compete with penryn on this chip, it remains to be seen wether or not it will continue to scale appropriately afterwards or wether they will leave us hanging again due to underestimating Intel.
I dunno. If its true, as published, its absolutely fantastic news and right along the lines of what AMD needs. If its another of their ".....well, in this instance of simulations in FP we experianced an equivelent operating frequnecy of X.XGhz based on Spec2004" BS style metrics, then they should all burn.
Dont ask me why, but for some reason, even though its from the Inq,( :roll: ) I have a gut feeling this is true.
indeed
But as to charlie having penned it and claimed he came up with the k8L name :? Charlies credibility is as good as Bush's WMD in Iraq or Nixon at watergate.Charlie is to news what clinton is to the press about lewinsky during the first week of denial. Hows that cigar charlie? :roll:
Lets tear apart the fact that he doesnt nail down the architectural changes that could support his anticipation that the clock speeds will even remotely be competitive.
AMD and Intel have shown that clock speeds mean diddly if the architechture cannot utilize it efficiently. Happy people in austin could just be listening to SXSW cd's at work and having a common local favorite hitting a great stride towards national stardom.
Even if AMD can compete with penryn on this chip, it remains to be seen wether or not it will continue to scale appropriately afterwards or wether they will leave us hanging again due to underestimating Intel.
I dunno. If its true, as published, its absolutely fantastic news and right along the lines of what AMD needs. If its another of their ".....well, in this instance of simulations in FP we experianced an equivelent operating frequnecy of X.XGhz based on Spec2004" BS style metrics, then they should all burn.
Dont ask me why, but for some reason, even though its from the Inq,( :roll: ) I have a gut feeling this is true.
your gut is telling u that because AMD is due for some good luck....
its an ILL WIND that blows one way
erocker
Distinguished
- Jul 2, 2006
- 276
- 0
- 18,780
Wow, any of you who do not want AMD's new processors to be faster are just plain ignorant. I don't give a crap what company it is, faster processors = better competition, better prices, more products. So, what is it? Why do some of you want to see a company like AMD or Intel fail?
Major_Spittle
Distinguished
- Nov 17, 2006
- 459
- 0
- 18,780
I certainly hope this is true. 500mhz would mean sooooo much to AMD's K10, 20% increase in performance. Hopefully their benchmarks, when released, will show K10 will compete with higher clocked Penryn and this will give AMD some thunder back before Nahelam smites them.
Legenic
Distinguished
- Mar 9, 2006
- 148
- 0
- 18,680
the inquirer also posted things like conroe being released at 3.33ghz, and that intel's 45nm manufacturing FIXED (not just improved) leakage.
considering AMD just announced clock speeds of their upcoming chips I wouldn't believe this.. but you never know.
considering AMD just announced clock speeds of their upcoming chips I wouldn't believe this.. but you never know.
turpit
Splendid
- Feb 12, 2006
- 6,373
- 0
- 25,780
Nobody wants to see AMD fail. Everybody wants to see AMD succeed. The problem is the realists see the facts, the majority of which, havent supported AMD succeeding, while the fanatics and dreamers hang on every word from Dirk Meyers, Henri Richards and Hector Ruizs mouths as if they were Moses, and disregard the facts.
Either company failing would be extremely bad as there is no other significant manufacturer in the x86 CPU market. No competition = monopoly = prices hikes + technological stagnation. But the reality is AMD has been floundering, and the gas their executives has passed has not been very heartening. Now, if the Uarch they have revealed for K10 is the actual Uarch it will use, then it looks very, very good. Time will tell.
Nahelam is more of I'll believe it when it's proven, for me, because I'm guessing the easy arch improvements have been done now. That's just a sensible guess. On the other hand, if Intel has some geniuses at work on arch, then....we'll see. I'm skeptical. And I believed everything about CoreDuo when the first faint buzz rumors were out there. After all, it was only evolving the PentM after plenty of time, so improvements were very likely to be something then.
TRENDING THREADS
-
Question User RX 580 8gb red devil powercolor, asking original BIOS- Started by PizzaInGame
- Replies: 5
-
-
-
News Microsoft updates Windows 11 24H2 requirements, CPU must support SSE4.2 or the OS will not boot- Started by Admin
- Replies: 15
Facebook
Twitter
Instagram