Please register or login. There are 0 registered and 1261 anonymous users currently online. Current bandwidth usage: 326.30 kbit/s November 29 - 12:46pm EST 
Hardware Analysis
      
Forums Product Prices
  Contents 
 
 

  Latest Topics 
 

More >>
 

    
 
 

  You Are Here: 
 
/ Forums / Processors /
 

  AMD 65nm Brisbane 
 
 Author 
 Date Written 
 Tools 
Continue Reading on Page: 1, 2, 3
FordGT90Concept Dec 08, 2006, 06:35pm EST Report Abuse
>> Re: AMD 65nm Brisbane
Talk about ignorance. You know most of what you said there is blatantly incorrect. Why should I bother to correct you when you'll just throw more BS at me which, too, is incorrect? For that very reason, I refuse to directly reply as it is a waste of my time.

Oh, by the way, I still don't see a shread of proof showing that Barcelona will even get close to Kentsfield. Why should I even begin to believe any words you type?


Jim does have a point which it did help a lot with NetBurst but, K8 has never been good for high clockspeeds so I am confident it'll have little to no effect. K8 just can't handle high clockspeeds, not due to heat because that is in check, but because the architecture is incapable of it.

________________________
If I remember what I forgot, I have not forgotten it.
Want to enjoy less advertisements and more features? Click here to become a Hardware Analysis registered user.
SuPeR Xp Dec 08, 2006, 06:42pm EST Report Abuse
>> Re: AMD 65nm Brisbane
FordGT90Concept said:
Talk about ignorance. You know most of what you said there is blatantly incorrect. Why should I bother to correct you when you'll just throw more BS at me which, too, is incorrect? For that very reason, I refuse to directly reply as it is a waste of my time.

Oh, by the way, I still don't see a shread of proof showing that Barcelona will even get close to Kentsfield. Why should I even begin to believe any words you type?


I don't care about the Barcelona right now, we are talking about the Native Quad's being released in around Q3 2007, so I will see you then, for now, just twiddle your fingers and see what happends ;) And if my source is full ofs**t, then I will bite myself in the ass & video tape it for you guys. Hey, he is from India, ;)

-------------------------------------------------
Custom AMD HAF 932 Red Dragon GAMING MOD!!!
http://www.techpowerup.com/gallery/2442.html
FordGT90Concept Dec 08, 2006, 06:56pm EST Report Abuse
>> Re: AMD 65nm Brisbane
Super XP said:
I don't care about the Barcelona right now, we are talking about the Native Quad's being released in around Q3 2007, so I will see you then, for now, just twiddle your fingers and see what happends ;) And if my source is full ofs**t, then I will bite myself in the ass & video tape it for you guys. Hey, he is from India, ;)

QFT

By the way, notice how AMD only managed a 200 MHz increase in over a year going from the FX57 (2.8 GHz) to FX74 (3.0 GHz)? Not to mention, they really didn't work on the IPC count either so that 200 MHz is all they were able to achieve. If that isn't a red flag for a plateaued architecture, what is?

________________________
If I remember what I forgot, I have not forgotten it.
SuPeR Xp Dec 08, 2006, 07:11pm EST Report Abuse
>> Re: AMD 65nm Brisbane
FordGT90Concept said:
Super XP said:
I don't care about the Barcelona right now, we are talking about the Native Quad's being released in around Q3 2007, so I will see you then, for now, just twiddle your fingers and see what happends ;) And if my source is full ofs**t, then I will bite myself in the ass & video tape it for you guys. Hey, he is from India, ;)

QFT

By the way, notice how AMD only managed a 200 MHz increase in over a year going from the FX57 (2.8 GHz) to FX74 (3.0 GHz)? Not to mention, they really didn't work on the IPC count either so that 200 MHz is all they were able to achieve. If that isn't a red flag for a plateaued architecture, what is?


GT90, I would have liked for AMD to increase there clock speed, but there was no reason for it at the time. They still kept up in price/performance/power usage etc. So, if AMD can gain there performance crown without upping the core clock, then why not.

But now, they are in trouble IMO, in terms of scalling there core clock. I am sure they will be able to get a good speed boost with the 65nm, maybe around 3.6 GHz to even 4.0 GHz, but that is all IMO, so it all has to do with phisical K8 core enhancements to increase overall performance, without jacking the core clocks up.

This is why they plan on releasing there Native Quad-Core code names Altair @ 2.70 GHz to 2.90 GHz. Notice that they are still not using 3.0 GHz ????

-------------------------------------------------
Custom AMD HAF 932 Red Dragon GAMING MOD!!!
http://www.techpowerup.com/gallery/2442.html
FordGT90Concept Dec 08, 2006, 07:22pm EST Report Abuse
>> Re: AMD 65nm Brisbane
AMD/IBM has had problems with the 65nm process and whenever you make something more complex, it can't clock as high. It is too be expected.

Super XP said:
This is why they plan on releasing there Native Quad-Core code names Altair @ 2.70 GHz to 2.90 GHz. Notice that they are still not using 3.0 GHz ????

The K8 architecture has scaled to it's limits. Also, I doubt those figures are accurate. How is it supposed to get a 2.70 or 2.90 clock speed when using a 200 MHz FSB? I'd expect the first one (FX series) out to be at 2.8 GHz or 2.6 GHz considering the FX74 is at 3.0 GHz and far less complex. I cannot see it be at 3.0 GHz out of the gate unless this is a brand new architecture (which AMD had made clear it isn't).

________________________
If I remember what I forgot, I have not forgotten it.
Bungle Dec 08, 2006, 08:46pm EST Report Abuse
>> Re: AMD 65nm Brisbane

Core i7 4770K @ 4.4GHz | Corsair H110 | Gigabyte GA-Z87X-UD5H | 32GB 1866MHz Corsair Vengeance
2x EVGA GTX 780 SLI | 256GB OCZ Vector SSD | 4TB Hitachi 7K4000
Corsair AX1200 PSU | Corsair 650D | Windows 7 Ultimate x64
FordGT90Concept Dec 08, 2006, 09:13pm EST Report Abuse
>> Re: AMD 65nm Brisbane
The problem with Vapochill is that at 3.6 GHz, it requires an innate amount of energy to keep the thing running and to market 3.6 GHz chips would keep almost every single processor sold coming back just because it wasn't hooked up to phase change. 2.93 GHz Intel's hit over 4 GHz with ease on air clearly showing the architecture has a lot of breathing room.

http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2795&p=18


As stated before AMD/IBM have had problems scaling to 65nm process so I would not expect miracles from it. The K8 architecture was designed for 90nm SOI, not 65nm SOI.

So yes, I stand by statement: K8 has reached it's limits for the simple fact you can't sell 3.6 GHz processors and require people to spend over $1000 just to keep the thing from melting the motherboard. Only processors that can be cooled by price effective means (air) can be offered to consumers. If they didn't, the FCC (or is it FTC?) would get all over them for selling "fire hazards." It simply has run out of breathing room just trying to keep ahead of NetBurst.

________________________
If I remember what I forgot, I have not forgotten it.
G. G. Dec 08, 2006, 09:33pm EST Report Abuse
>> Re: AMD 65nm Brisbane
SuperXP Goofed up Statememt:
But competing with the Core 2 (Dual Core) AMD's Native Quad's should win hands down IMO.



hey ... that was pretty messed up.... I hoped you meant something else here in this statement.... Quad core better hell win hands down to a Dual Core !!!!! there is extra two cores for gods sake... 4 vs 2 ???? if not.... heheheh they got bigger problems than 4x4.

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

-
Bungle Dec 08, 2006, 10:05pm EST Report Abuse
>> Re: AMD 65nm Brisbane
FordGT90Concept said:
The problem with Vapochill is that at 3.6 GHz, it requires an innate amount of energy to keep the thing running and to market 3.6 GHz chips would keep almost every single processor sold coming back just because it wasn't hooked up to phase change. 2.93 GHz Intel's hit over 4 GHz with ease on air clearly showing the architecture has a lot of breathing room.

http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2795&p=18


As stated before AMD/IBM have had problems scaling to 65nm process so I would not expect miracles from it. The K8 architecture was designed for 90nm SOI, not 65nm SOI.

So yes, I stand by statement: K8 has reached it's limits for the simple fact you can't sell 3.6 GHz processors and require people to spend over $1000 just to keep the thing from melting the motherboard. Only processors that can be cooled by price effective means (air) can be offered to consumers. If they didn't, the FCC (or is it FTC?) would get all over them for selling "fire hazards." It simply has run out of breathing room just trying to keep ahead of NetBurst.


Yes netburst hits 4GHz easily because the ARCHETECTURE

Core i7 4770K @ 4.4GHz | Corsair H110 | Gigabyte GA-Z87X-UD5H | 32GB 1866MHz Corsair Vengeance
2x EVGA GTX 780 SLI | 256GB OCZ Vector SSD | 4TB Hitachi 7K4000
Corsair AX1200 PSU | Corsair 650D | Windows 7 Ultimate x64
Bungle Dec 08, 2006, 10:07pm EST Report Abuse
>> Re: AMD 65nm Brisbane
FordGT90Concept said:
The problem with Vapochill is that at 3.6 GHz, it requires an innate amount of energy to keep the thing running and to market 3.6 GHz chips would keep almost every single processor sold coming back just because it wasn't hooked up to phase change. 2.93 GHz Intel's hit over 4 GHz with ease on air clearly showing the architecture has a lot of breathing room.

http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2795&p=18


As stated before AMD/IBM have had problems scaling to 65nm process so I would not expect miracles from it. The K8 architecture was designed for 90nm SOI, not 65nm SOI.

So yes, I stand by statement: K8 has reached it's limits for the simple fact you can't sell 3.6 GHz processors and require people to spend over $1000 just to keep the thing from melting the motherboard. Only processors that can be cooled by price effective means (air) can be offered to consumers. If they didn't, the FCC (or is it FTC?) would get all over them for selling "fire hazards." It simply has run out of breathing room just trying to keep ahead of NetBurst.


Yes netburst hits 4GHz easily because the ARCHETECTURE is capible of around 10GHz before internal busses start getting out of sync. so they could THEORETICALLY scale to 10GHz. However, because of the limits of the manufacturing process they are limited to 4ish GHz. I'm not arguing your points necessarily I'm trying to get you to correct your terminology.

Core i7 4770K @ 4.4GHz | Corsair H110 | Gigabyte GA-Z87X-UD5H | 32GB 1866MHz Corsair Vengeance
2x EVGA GTX 780 SLI | 256GB OCZ Vector SSD | 4TB Hitachi 7K4000
Corsair AX1200 PSU | Corsair 650D | Windows 7 Ultimate x64
FordGT90Concept Dec 08, 2006, 10:10pm EST Report Abuse
>> Re: AMD 65nm Brisbane
I was talking about performance, not architecture with the reference to NetBurst. K8 was designed to oust NetBurst, not Core. NetBurst did run into scaling problems due to heat and leaks (later problems with very high cache miss frequency). That's not necessarily the case with K7 and K8--it's the complexity that keeps it from scaling far. In comparison, the Core architecture is actually quite simple and straight forward but has a lot of brute force to back it up (highest IPC count to date).

________________________
If I remember what I forgot, I have not forgotten it.
SuPeR Xp Dec 08, 2006, 10:22pm EST Report Abuse
>> Re: AMD 65nm Brisbane
Well, the only real advantage I see in the Core architecture & why they are very fast is there cache system. It is tooned very very good. Also, there Dual FSB make a difference too, but it is all about the cache from what I've read.

K8, I still think there is a lot of room for expansion. AMD did say that there K8 is completely scalable when needed to be scalable. And now the way things look, it needs that scalability. They may not scale well on CPU raw speed, but in acual design, as long as they can make this revised K8 more efficient, they will succeed IMO.

-------------------------------------------------
Custom AMD HAF 932 Red Dragon GAMING MOD!!!
http://www.techpowerup.com/gallery/2442.html
FordGT90Concept Dec 08, 2006, 10:29pm EST Report Abuse
>> Re: AMD 65nm Brisbane
A high IPC count (efficiency) is the only way Barcelona will be able to get on top of Kentsfield. They are clearly working on it by stating it can do up to four floating point operations per clock but that needs to be done across the board--not just FPU. If they can, K8 will remain competitive a little bit longer; however, odds are stacked against it. It's like a M1A1 vs a T-72--the T-72 doesn't stand a chance because it's older. The same goes for K8 vs P6.

________________________
If I remember what I forgot, I have not forgotten it.
G. G. Dec 08, 2006, 10:32pm EST Report Abuse
>> Re: AMD 65nm Brisbane
yeahhhhhh baby.... M1A1 Abram running at 45mph across bumpy terrain and shooting a round off at the same time hitting the designated target a mile away...... woot.....

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

-
SuPeR Xp Dec 08, 2006, 10:33pm EST Report Abuse
>> Re: AMD 65nm Brisbane
I see, but I think you are a little too hard on AMD's K8. I think we should give it a chance & see what happends. I mean, that e-mail I got sounds really good, could be too good to be true, but only time will tell.

-------------------------------------------------
Custom AMD HAF 932 Red Dragon GAMING MOD!!!
http://www.techpowerup.com/gallery/2442.html
FordGT90Concept Dec 08, 2006, 10:50pm EST Report Abuse
>> Re: AMD 65nm Brisbane
Yes, only time will tell.

________________________
If I remember what I forgot, I have not forgotten it.

Continue Reading on Page: 1, 2, 3

 

    
 
 

  Topic Tools 
 
RSS UpdatesRSS Updates
 

  Related Articles 
 
 

  Newsletter 
 
A weekly newsletter featuring an editorial and a roundup of the latest articles, news and other interesting topics.

Please enter your email address below and click Subscribe.