Will AMD Phenom beat CD2???

[Regs]

I think AMD should take all the time they need with shanghai. Maybe they can figure out how to even lower the L1 and L2 cache latency and further improve their out-of-order loads.

If they go with another "just another die shrink", it will be one of their worst mistakes. I still don't see a benifit of L3 cache at all.

 

[James007Bond]

I don't understand the question. Beat it at what?

It looks like it may per clock 4 clock about on par with the Conroe but it won't have the same speed bins.

So performance / Watt maybe it will be better. I'm sure AMD will market it at Price/Performance stand point within a few pounds of Conroe.

If you want the quickest system get Conroe. If your happy with a mid range CPU, you will probably have a choice of Conroe or Phenom.

 

[T2k]

Originally posted by: Viditor
Considering that AMD has already demonstrated a 3 GHz Phenom, I think your estimate of Q2 may be a bit pessimistic...

Did they...?

Remember that Shanghai is due in Q2/Q3 2008 (45nm Barcelona),

Is it...?

PS: I think your statements need some links...

 

[T2k]

Originally posted by: James007Bond

I don't understand the question. Beat it at what?

Err in terms of prformance, it's more than obvious, I think. You know, you get what you pay for - not clock for clock (never understood why it's important for anybody outside of the bench-boyz) but dollar-for-dollar.

I'd say it'll be very-very hard to beat the ~$280 Q6600 and I don't even want to get into the OC performance...

It looks like it may per clock 4 clock about on par with the Conroe but it won't have the same speed bins.

So performance / Watt maybe it will be better. I'm sure AMD will market it at Price/Performance stand point within a few pounds of Conroe.

If you want the quickest system get Conroe. If your happy with a mid range CPU, you will probably have a choice of Conroe or Phenom.

Dunno but I'm not sure about that clock-for-clock matchup...

 

[jones377]

Originally posted by: Viditor

Originally posted by: jones377

Originally posted by: zsdersw

Originally posted by: Viditor
Remember that Shanghai is due in Q2/Q3 2008 (45nm Barcelona)

If that's going to occur, we should've seen the 45nm tape-out by now.. or will have to see it very soon.

Tape-out to release is usually a year or more, especially when it involves a new microarchitecture or a new process. I wouldn't count on a Q2 release unless someone can show me a statement from AMD that says Shanghai has already taped out.

9 months to a year is correct (1 year being more common)...and I haven't seen a tape-out announcement either. That said, it is still in their roadmap (though I certainly don't disagree with the skepticism here...).

Edit: BTW, I believe that Penryn will be a 9 month lead from tape-out for their first chip...

Penryn taped out in late November last year. First samples arrived in early January.

Tapeout
First samples

Since it's expected to be released towards the end of the year that makes it pretty much 12 months from tape-out to shipping.

 

[T2k]

Originally posted by: jones377

Originally posted by: Viditor

Originally posted by: jones377

Originally posted by: zsdersw

Originally posted by: Viditor
Remember that Shanghai is due in Q2/Q3 2008 (45nm Barcelona)

If that's going to occur, we should've seen the 45nm tape-out by now.. or will have to see it very soon.

Tape-out to release is usually a year or more, especially when it involves a new microarchitecture or a new process. I wouldn't count on a Q2 release unless someone can show me a statement from AMD that says Shanghai has already taped out.

9 months to a year is correct (1 year being more common)...and I haven't seen a tape-out announcement either. That said, it is still in their roadmap (though I certainly don't disagree with the skepticism here...).

Edit: BTW, I believe that Penryn will be a 9 month lead from tape-out for their first chip...

Penryn taped out in late November last year. First samples arrived in early January.

Tapeout
First samples

Since it's expected to be released towards the end of the year that makes it pretty much 12 months from tape-out to shipping.

Right but we have't seen anything like that about Shanghai from AMD...

 

[Nemesis 1]

Originally posted by: Viditor

Originally posted by: jones377

Originally posted by: zsdersw

Originally posted by: Viditor
Remember that Shanghai is due in Q2/Q3 2008 (45nm Barcelona)

If that's going to occur, we should've seen the 45nm tape-out by now.. or will have to see it very soon.

Tape-out to release is usually a year or more, especially when it involves a new microarchitecture or a new process. I wouldn't count on a Q2 release unless someone can show me a statement from AMD that says Shanghai has already taped out.

9 months to a year is correct (1 year being more common)...and I haven't seen a tape-out announcement either. That said, it is still in their roadmap (though I certainly don't disagree with the skepticism here...).

Edit: BTW, I believe that Penryn will be a 9 month lead from tape-out for their first chip...

Viditor I understand your position perfectly. If I may I would like to point out a few things to ya. Maybe it will help.

I understand your investment position. Its really not that bad . But several things have to occur . Fusion has got to > Nehalem and it is possiable. Read some of the news thats posted at the Inquirer today. If AMD can pull off a few of the things in one of those articles you will kmow which one!
But I must interject on the AMD 45nm process in 2qt of 08 . It isn't going to happen . Why?
AMD was planning on SOI on the 45nm process if amd tries this it will be disasterious. Soi @ 65 is a headache for AMD @ 45 nm OH my!
IBM has announced that it is getting ready with High K and metal gates. IBM has also stated that it process will be much better than Intels.
I really don't usually believe IBM's hype but in this case I do. I believe IBM is looking at (excuss the spelling as I won't google for the proper wording) Finfit on the 45nm Higk K and dual metal gates. Which should > Intels 45nm process.

So AMD has to make a decision. TRY soi which I believe will be horriable for AMD . Or get IBM to sell license 45nm High K /metal gate process for use in thier fabs.

SOI at 45nm will be wrenched with delays and its performance will be desending compared to 65nm. Better for AMD to go with the future now. AMD needs to forget about Z-ram for L3 .

As far as K10 fab rejects. Simple logic tells us your going to get a higher rate of rejects with a native 4core prossor. So yes yields are going to be much lower than intels. Than factor that Intel will be on 45nm process and a cheaper process at that Vs. soi. Intels margins will be much much higher. So much infact that Intel could price AMD out of the market if they desired. But they won't they need AMD.

As you are aware Intel is going in a differant direction than AMD. Penryn is already showing us Intel is moving towards Ray tracing on the desk top . We won't see it on penryn . (lack of power) But Nehalem will be the next big step in that direction . But it still won't be ready for prime time. (softwear) . So AMD does have a window of oppertunity If you read the Inq articles I was referring to you can see what I am saying.
Now Intel has an excellant plan BUT they never ever believed AMD would buy ATI. SO they didn't plan for it. AMD has a window of opertunity here. They just have to deliver.

But once we get to Geshner its end game . Real time raytracing will be a fact. Softwear will be ready and Intel plans will be in place. Once we have real time ray tracing that it . Good by NV goodbye ATI.

I am going to follow fusion very closely as I see great oppertunity here as an investor . Its a long shot and a ways off . But this really could be the KING for 2 more years If AMD exacutes perfectly.

 

[Viditor]

Originally posted by: jones377

Originally posted by: Viditor

Originally posted by: jones377

Originally posted by: zsdersw

Originally posted by: Viditor
Remember that Shanghai is due in Q2/Q3 2008 (45nm Barcelona)

If that's going to occur, we should've seen the 45nm tape-out by now.. or will have to see it very soon.

Tape-out to release is usually a year or more, especially when it involves a new microarchitecture or a new process. I wouldn't count on a Q2 release unless someone can show me a statement from AMD that says Shanghai has already taped out.

9 months to a year is correct (1 year being more common)...and I haven't seen a tape-out announcement either. That said, it is still in their roadmap (though I certainly don't disagree with the skepticism here...).

Edit: BTW, I believe that Penryn will be a 9 month lead from tape-out for their first chip...

Penryn taped out in late November last year. First samples arrived in early January.

Tapeout
First samples

Since it's expected to be released towards the end of the year that makes it pretty much 12 months from tape-out to shipping.

You got that backwards...first samples in preperation for tape-out were in November, tape-out was in Dec/January.
Daily Tech

"Sources inside Intel have confirmed the company recently received the taped-out 45nm Penryn processor"
"Penryn, the 45nm optical-shrink of the Core architecture, was prepped for tape-out in late November, and returned to the Intel development team just a few weeks ago"

Penryn is to launch November 11th. Shipping should begin the month before (unless they are doing a paper launch).
DailyTech

So 9 months, maybe 10...

 

[sdsdv10]

Originally posted by: T2k

Originally posted by: jones377
Penryn taped out in late November last year. First samples arrived in early January.

Tapeout
First samples

Since it's expected to be released towards the end of the year that makes it pretty much 12 months from tape-out to shipping.

Right but we have't seen anything like that about Shanghai from AMD...

A question...

Was there ever an official AMD announcement of the tapeout for Barcelona? I don't remember one, but that doesn't mean there wasn't one.

 

[Phynaz]

Originally posted by: Viditor

Originally posted by: jones377

Originally posted by: Viditor

Originally posted by: jones377

Originally posted by: zsdersw

Originally posted by: Viditor
Remember that Shanghai is due in Q2/Q3 2008 (45nm Barcelona)

If that's going to occur, we should've seen the 45nm tape-out by now.. or will have to see it very soon.

Tape-out to release is usually a year or more, especially when it involves a new microarchitecture or a new process. I wouldn't count on a Q2 release unless someone can show me a statement from AMD that says Shanghai has already taped out.

9 months to a year is correct (1 year being more common)...and I haven't seen a tape-out announcement either. That said, it is still in their roadmap (though I certainly don't disagree with the skepticism here...).

Edit: BTW, I believe that Penryn will be a 9 month lead from tape-out for their first chip...

Penryn taped out in late November last year. First samples arrived in early January.

Tapeout
First samples

Since it's expected to be released towards the end of the year that makes it pretty much 12 months from tape-out to shipping.

You got that backwards...first samples in preperation for tape-out were in November, tape-out was in Dec/January.
Daily Tech

"Sources inside Intel have confirmed the company recently received the taped-out 45nm Penryn processor"
"Penryn, the 45nm optical-shrink of the Core architecture, was prepped for tape-out in late November, and returned to the Intel development team just a few weeks ago"

Penryn is to launch November 11th. Shipping should begin the month before (unless they are doing a paper launch).
DailyTech

So 9 months, maybe 10...

You have either confused yourself, or you don't know what tape-out is.

Either way, you can't get chips before tape-out, as you appear to claim above.

The term tape-out comes from when magnetic tapes were used to transfer data. When the design of a chip was complete it would be output to tape (tape-out), and then the tape sent to the fab.

As you can see, there's no way that a chip could be produced in Nov, and tape-out to occur in Dec/Jan. That's backwards from the way the process works.

 

[jones377]

Originally posted by: Viditor

Originally posted by: jones377

Originally posted by: Viditor

Originally posted by: jones377

Originally posted by: zsdersw

Originally posted by: Viditor
Remember that Shanghai is due in Q2/Q3 2008 (45nm Barcelona)

If that's going to occur, we should've seen the 45nm tape-out by now.. or will have to see it very soon.

Tape-out to release is usually a year or more, especially when it involves a new microarchitecture or a new process. I wouldn't count on a Q2 release unless someone can show me a statement from AMD that says Shanghai has already taped out.

9 months to a year is correct (1 year being more common)...and I haven't seen a tape-out announcement either. That said, it is still in their roadmap (though I certainly don't disagree with the skepticism here...).

Edit: BTW, I believe that Penryn will be a 9 month lead from tape-out for their first chip...

Penryn taped out in late November last year. First samples arrived in early January.

Tapeout
First samples

Since it's expected to be released towards the end of the year that makes it pretty much 12 months from tape-out to shipping.

You got that backwards...first samples in preperation for tape-out were in November, tape-out was in Dec/January.
Daily Tech

"Sources inside Intel have confirmed the company recently received the taped-out 45nm Penryn processor"
"Penryn, the 45nm optical-shrink of the Core architecture, was prepped for tape-out in late November, and returned to the Intel development team just a few weeks ago"

Penryn is to launch November 11th. Shipping should begin the month before (unless they are doing a paper launch).
DailyTech

So 9 months, maybe 10...

lol... someone else already replied so I don't have to.

 

[Conroe]

Knowing that the first AM2 Phenom will be a 2.2 or 2.4ghz FX-80 it doesn't look like it will beat a Q6600 in price or performance. Maybe they should rename it a Duron-80 and sell it cheap.

 

[Conroe]

double post

 

[zsdersw]

Originally posted by: Conroe
Knowing that the first AM2 Phenom will be a 2.2 or 2.4ghz FX-80 it doesn't look like it will beat a Q6600 in price or performance. Maybe they should rename it a Duron-80 and sell it cheap.

It's a pretty large amount of silicon.. I doubt it can be sold for what we'd consider "cheap" in this context.

 

[SerpentRoyal]

Phenom's competition with come from the 45nm Intel Quads. I expect an overclocked 45nm quad to blow past 4.0GHz with a capable MB. Again, AMD's best quad will still run 15% slower than Intel's best quad.

 

[T2k]

Originally posted by: Viditor

Originally posted by: jones377

Originally posted by: Viditor

Originally posted by: jones377

Originally posted by: zsdersw

Originally posted by: Viditor
Remember that Shanghai is due in Q2/Q3 2008 (45nm Barcelona)

If that's going to occur, we should've seen the 45nm tape-out by now.. or will have to see it very soon.

Tape-out to release is usually a year or more, especially when it involves a new microarchitecture or a new process. I wouldn't count on a Q2 release unless someone can show me a statement from AMD that says Shanghai has already taped out.

9 months to a year is correct (1 year being more common)...and I haven't seen a tape-out announcement either. That said, it is still in their roadmap (though I certainly don't disagree with the skepticism here...).

Edit: BTW, I believe that Penryn will be a 9 month lead from tape-out for their first chip...

Penryn taped out in late November last year. First samples arrived in early January.

Tapeout
First samples

Since it's expected to be released towards the end of the year that makes it pretty much 12 months from tape-out to shipping.

You got that backwards...first samples in preperation for tape-out were in November, tape-out was in Dec/January.
Daily Tech

"Sources inside Intel have confirmed the company recently received the taped-out 45nm Penryn processor"
"Penryn, the 45nm optical-shrink of the Core architecture, was prepped for tape-out in late November, and returned to the Intel development team just a few weeks ago"

Penryn is to launch November 11th. Shipping should begin the month before (unless they are doing a paper launch).
DailyTech

So 9 months, maybe 10...

Viditor, you have no clue what tape-out means, it's obvious... why don't you read about on it before you post *anything*, seriously?

 

[Regs]

There are many ways to make a point and there are many ways to earn respect. To be derogatory is not one of them.

 

[Idontcare]

Originally posted by: Viditor

Originally posted by: jones377
Penryn taped out in late November last year. First samples arrived in early January.

Tapeout
First samples

Since it's expected to be released towards the end of the year that makes it pretty much 12 months from tape-out to shipping.

You got that backwards...first samples in preperation for tape-out were in November, tape-out was in Dec/January.
Daily Tech

"Sources inside Intel have confirmed the company recently received the taped-out 45nm Penryn processor"
"Penryn, the 45nm optical-shrink of the Core architecture, was prepped for tape-out in late November, and returned to the Intel development team just a few weeks ago"

Penryn is to launch November 11th. Shipping should begin the month before (unless they are doing a paper launch).
DailyTech

So 9 months, maybe 10...

For some generally accepted definitions and example uses of the term "tape-out" see:

Tape-out is usually a cause for celebration by everyone who worked on the project, followed by eager anticipation of an actual product returning from the manufacturing facility.
http://en.wikipedia.org/wiki/Tape-out

Tape-out refers to the release of the layout design package to the manufacturing fab for the creation of photolithographic masks. Practically it means that the chip developer is confident in the released design.
http://www.xbitlabs.com/news/c...ay/20061019222838.html

 

[imported_mackintire]

Guys.... its as simple as this.... Phemon is 30% faster in floating point at 2.0Ghz than the current Core 2 Duo at 2.0Ghz
But the Core 2 Duo is 30% faster at interger performance at 2.0Ghz.

AMD's Phemon has better scaling per clock than Core 2 Duo at 65nm. But no one knows how it will do against the 45nm chip.

With a faster FSB, hypertransport and memory aka 1066 PC8500 I am sure that the 2.4-3.0Ghz Phemon will be competitive to Intel's 45mn offering at the same clock speed.

I also believe that AMD's gradulated voltage and clocking will be better at managing real life power usage than Intels ALL or NOTHING version of speed step.

As long as the yields are decent AMD will do ok. What I am unsure of, is if financially ok is good enough.

I also seem to remember reading something that 65nm K10 could scale to 3.4Ghz without a rework of the pipeline. So with refridgerent Phemon should be able to get to 3.4-3.6Ghz

If Shangai is tweaked for clock speeds we could see 2.8Ghz-3.8Ghz shangai systems but I think reality will be 2.6Ghz-3.6Ghz

 

[Regs]

My worry mackintire is that I wonder how many consumer applications are actually more FP intensive than integer intensive.

I also would not mind if AMD sacrificed some of it's power saving features on a Phenom to increase the clock cycle.

 

[imported_mackintire]

I forget who posted it. But I agree AMD will have their butts handed to them on a plate on overclocking performance. aka Q6600 at 3.2Ghz would probably be a little faster in most ways then a 3.0Ghz Phemon FX.

That said, the idle power of the Phemon FX at 3.0Ghz is most likely 100 watts less then the overclocked Q6600. (Speed step disabled for overclocking)

Also under full load it probably eats some 70-90 watts less.

Under partial load it probably consumes 75-100 watts less.


Considering my Qx6700 based machine raises my electric bill near $10.00 a month, efficiency IS important to me.

My Qx6700 running 3.0Ghz IS fast enough. I would consider a 3.0Ghz Phemon an upgrade if my electric bill goes down. Also considering that Gigabyte can not get Standby to work on my GA-P35-DS3P with my QX6700. I now have to turn my machine off, something I used to never have to do.

Still I would like a Phemon 3.0 Ghz system with 2 of the upcomming 55nm version(s) of the X2900XT

 

[Blacklash]

Overclocking does change the picture quite a bit. Look at what folks are already doing with Q6600 G-0s on average:

http://www.xtremesystems.org/f...?p=2410961&postcount=2

Now if the 45nm chips are 200MHz better than this we could soon see 4.0GHz with Quad core on air with a decent after market cooler. I know folks are already booting up to 4.0GHz on air and I am talking Quad core and Prime stable.

 

[firewolfsm]

Originally posted by: Regs
My worry mackintire is that I wonder how many consumer applications are actually more FP intensive than integer intensive.

I also would not mind if AMD sacrificed some of it's power saving features on a Phenom to increase the clock cycle.

Games are usually very FP intensive, and I think the gains showed there. It's just that, while most people are saying that the memory was limiting the chip, I think it's actually the onboard memory controller and L3 cahche that are holding back performance. Both of these increase with clockspeed (that's why Gary Key hinted at a possible better than linear gain with higher clocks). Also, the new chipsets place the memory controller on a separate power plane than the cores, which means it can be overclocked with a very small change in power consumption (I think AMD hinted that they will support this, like Penryn supports overclocking one core without the others now)

Originally posted by: mackintireThat said, the idle power of the Phemon FX at 3.0Ghz is most likely 100 watts less then the overclocked Q6600. (Speed step disabled for overclocking)

Also under full load it probably eats some 70-90 watts less.

Under partial load it probably consumes 75-100 watts less.

Don't just spew random power figures, especially when the estimates are way off. Overclocks usually don't increase power consumption by more 50%. Your numbers don't make any sense. Q6600s only eat up 95W at stock.

And I have speedstep enabled on my OCed Q6600. It just drops to 1.9GHz instead of 1.6. Also, with my low voltage setting I think my power consumption is the same, if not lower, than stock.

 

[aigomorla]

WAIT, penryn will come out this november in the form of QX6950.

So whats all this talka bout it coming out 10 months from now?

IF you can afford it, you can pick one up in 2 months. I know i will be picking one up and having massive amounts of fun with it.


And i still think AMD is a lost cause.

Originally posted by: Blacklash
Overclocking does change the picture quite a bit. Look at what folks are already doing with Q6600 G-0s on average:

http://www.xtremesystems.org/f...?p=2410961&postcount=2

Now if the 45nm chips are 200MHz better than this we could soon see 4.0GHz with Quad core on air with a decent after market cooler. I know folks are already booting up to 4.0GHz on air and I am talking Quad core and Prime stable.

SHOW ME 1 computer whose booting up priming @ 4.0ghz Stable on air on ANY Quad B3 G0, ES even.

My ES is on a UBER water setup, and thats how i kept her stable. The board is on water, cpu on water, and im not talking about entry lvl, or even mid level water. Im talking about full blown PA120.3 with dual DDC-2's on a Copper top ApogeeGTX bowed.

Infact, even for water 4.0ghz prime stable is VERY VERY difficult, and will definitely shortn the life of your CPU + Board.

So what are u smoking? Care to share?

 

[Phynaz]

Originally posted by: firewolfsm

Originally posted by: Regs
My worry mackintire is that I wonder how many consumer applications are actually more FP intensive than integer intensive.

I also would not mind if AMD sacrificed some of it's power saving features on a Phenom to increase the clock cycle.

Games are usually very FP intensive, and I think the gains showed there. It's just that, while most people are saying that the memory was limiting the chip, I think it's actually the onboard memory controller and L3 cahche that are holding back performance. Both of these increase with clockspeed (that's why Gary Key hinted at a possible better than linear gain with higher clocks). Also, the new chipsets place the memory controller on a separate power plane than the cores, which means it can be overclocked with a very small change in power consumption (I think AMD hinted that they will support this, like Penryn supports overclocking one core without the others now)

Originally posted by: mackintireThat said, the idle power of the Phemon FX at 3.0Ghz is most likely 100 watts less then the overclocked Q6600. (Speed step disabled for overclocking)

Also under full load it probably eats some 70-90 watts less.

Under partial load it probably consumes 75-100 watts less.

Don't just spew random power figures, especially when the estimates are way off. Overclocks usually don't increase power consumption by more 50%. Your numbers don't make any sense. Q6600s only eat up 95W at stock.

And I have speedstep enabled on my OCed Q6600. It just drops to 1.9GHz instead of 1.6. Also, with my low voltage setting I think my power consumption is the same, if not lower, than stock.

When you say games, I assume you are talking about an FPS.
I once read that in even FP intensive code (outside of specialty applications) only about 1 out of 6 instructions is FP, the rest is integer. This explains why both AMD and Intel have stuck with what are essentially single FP units (although in some cases both cpu's can execute more than 1 FP operation per clock).