New AMD roadmap sucks

Silent Buddha · May 19, 2004

http://www.anandtech.com/cpu/showdoc.html?i=2056

Most interesting, however, is the small footnote below the roadmap which claims the FX-55 and Athlon 64 >4000+ (presumably 4200+) will require 104 watts/80 amps.

That's as bad as Prescott. Heat consumption won't scale linearly with 90nm...once us oc'ers get our hands on it and try to oc to 3GHz and beyond...at some point temps will skyrocket.

I understand why AMD wants to disable x86-64 on their "Value" line but I don't think that's very smart. Average joe six-pack is going to buy a value processor anyway instead of a Winchester or San Diego. When faced with a Celeron with x86-64 and an AMD value cpu what will he choose?

Gautam · May 19, 2004

I'm not going to come to any conclusions yet. AMD has deviated from their original roadmaps plenty in the past, and usually all for the better.

Captain Newbie · May 19, 2004

If they go to model numbers Intel-style, then I'm going to go found my own microprocessor company.

hitechjb1 · May 19, 2004

hitechjb1 said:
In the past, CPU frequency (MHz) roughly doubled for each generation of technology (180 nm, 130 nm). For 90, 65 nm and beyond, it would be harder and harder to achieve such trend due to the leakage current component which will surpass the active current component, ..., as possibly explained below.

How does leakage current slow down future generations of chips

In the past twenty years, chip manufacturers had a relatively easy time by doubling the CPU frequency every two to three years by shrinking the dimension (feature size) of transistors and wires inside a chip.

There are two main components of electric current inside a chip:

- active current: the "good" component that does logic computations by charging and discharging the internal capacitors of transistors and wires via the internal transistor switches

- leakage current: the "bad" component that is not computation related, and leaks through the transistors from supply voltage to ground, and dissipates as HEAT

E.g. a Tbred B 1700+ at rated 1.5 V, 1.47 GHz, draws about 30 A which is higher than the current of a typical house circuit breaker (which is typically 20-30A). When it is overclocked to 2.5 GHz, 1.9 V, it would draw about 65 A at full load which is more than 2-3 times the current of a house circuit breaker !!!

Historically, active current is the major current in a chip, so when more power is put in, the chip can run faster and does more computation.

From 90 nm, 65 nm and beyond, due to the smaller transistor channel length and thinner transistor oxide thickness, the leakage current increases at a faster rate and will surpass the active current. As a consequence, even when more power is put in, the chip frequency would increase at a slower pace than heat increase, the chip speed would level off due to heat. This is one of the major hurdle for silicon scaling to 90, 65 nm and beyond.

....

For more details:

How does leakage current slow down future generations of chips (page 19)

Captain Newbie · May 19, 2004

Hitech can come too.

Silent Buddha · May 19, 2004

I hope AMD's investment on SOI comes in handy. low-k would be would be nice too.

hitech is indeed very knowledgeable

EDIT: New features in future chips

Chowdy · May 20, 2004

I really hope that AMD and Nvidia work together on a naming system... i'm tired of converting "plusses" to GHz and then it not even being totally accurate anyway. And now, FX-5x's. Gah...

OC Detective · May 20, 2004

Lets look at what we have and what is predicted.
Currently a 2GHz A64 is rated at 89W, a 4200+ A64 (which will be around 3Ghz) is anticpated to be 104W.
So we have a 50% increase in frequency and only a 17% increase in heat output. So I dont think there is too much of an issue - expecting a reduction at such frequencies to present wattage is fanciful even with the size reduction.
Remember Prescotts produce more heat than P4C's at the SAME frequency that is not the case here.

hitechjb1 · May 20, 2004

OC Detective said:
Lets look at what we have and what is predicted.
Currently a 2GHz A64 is rated at 89W, a 4200+ A64 (which will be around 3Ghz) is anticpated to be 104W.
So we have a 50% increase in frequency and only a 17% increase in heat output. So I dont think there is too much of an issue - expecting a reduction at such frequencies to present wattage is fanciful even with the size reduction.
Remember Prescotts produce more heat than P4C's at the SAME frequency that is not the case here.

How do you get the 104 W power number for the 4200+ at 3 GHz?

OC Detective · May 20, 2004

from article
the FX-55 and Athlon 64 >4000+ (presumably 4200+) will require 104 watts/80 amps
How I get to the 3Ghz is that I expect the A64's to be 512KB L2 cache so currently the 3000+ is 2.0Ghz (with increments upward of 200MHz = 200+) extrapolating up to 4200+ gives 3GHz when the one-off socket 939 effect (say an extra 200+) is factored in.

CandymanCan · May 20, 2004

How is 105 as bad as prescott, from i heard prescot is 120 watt's.

Besides the 105 watt on the amd cpu is only running @ 1.3v that isnt bad at all, stop whinning

Silent Buddha · May 20, 2004

Because if you look at the power leakage when going to 90nm or 65nm the power consumption does not scale linearly with clock speed somewhere along the line it just becomes exponential.

FX-55 will be 2.6GHz and 104 watts. I don't see where 3GHz comes from.

OC Detective · May 20, 2004

Silent Buddha said:
I don't see where 3GHz comes from.

Didnt you read my post?

The Coolest · May 21, 2004

CandymanCan said:
How is 105 as bad as prescott, from i heard prescot is 120 watt's.

Besides the 105 watt on the amd cpu is only running @ 1.3v that isnt bad at all, stop whinning

Actually it is bad. the higher the current draw of a CPU the hotter the MOSFETs will get, not all motherboards will be able to handle that. and we'll have the same story we had w/ prescotts.
Usually after a process shrink the power draw of a processor dramatically decreases, even at higher clock speeds.
The 90nm is fundamentally flawed somehow, somewhere, as all manufacturers are reporting pretty much the same thing. The transistors start leaking current at higher clock speeds, which leads to higher power consumption, more heat, and the need of more voltage to stay stable at the specific speed, which makes matters even worse.
I'm pretty sure that this heat issue will not be as bad as it is with prescotts, but it certainly isn't good. And we overclockers are gonna need some better cooling solutions if we want to push these things anywhere noteworthy

xtrmeocr · May 21, 2004

actually its not as bad as prescott, because you forget intel and AMD report thermal specs differently. intel gives a normal operating thermal power and AMD gives a maximum theoretical power for the entire line.

so while the prescott is normally around 104 watts the athlon64 never even approaches 104 watts.

The Coolest · May 21, 2004

yes, but the current draw will still be way higher than 130nm A64s

Silent Buddha · May 21, 2004

OC Detective said:
Didnt you read my post?

Your post is wrong. That's not how the Socket 939 PR system works. Socket 939 A64 3500+ is clocked @ 2.2GHz and the Socket 939 A64 3800+ is clocked @ 2.4GHz. Now tell me how a Socket 939 A64 4200+ extrapolates to 3GHz? And why would that make sense? A64 FX-55 by itself requires 104 watts/80 amps. The FX-55 is clocked @ 2.6GHz and needs 104 watts so why would a 3GHz A64 only require 104 watts?

xtrmeocr · May 21, 2004

A64 FX-55 by itself requires 104 watts/80 amps. The FX-55 is clocked @ 2.6GHz and needs 104 watts so why would a 3GHz A64 only require 104 watts?

because like i said above, AMD's spec is for the ENTIRE line of chips(ones based on that core). they dont give a per-chip wattage, they give a theoretical maximum for the whole line.

Silent Buddha · May 21, 2004

Entire line? And how far does that whole line extend? Would that go to 3GHz? Intel is having trouble making 3.4GHz Prescotts. Can Intel keep the same power consumption for Prescotts up to 4GHz?

Who will be the first one to test a FX-55 past 3GHz and see how much power it consumes?

xtrmeocr · May 21, 2004

uhg do you read or what?

Silent Buddha said:
Entire line? And how far does that whole line extend? Would that go to 3GHz?

the entire line would be all the chips based on that chip revision. we are at 2.4ghz now and i think there will be a 2.6ghz 130nm version, i think its safe to say 90nm will bring us to atleast 3ghz.

Intel is having trouble making 3.4GHz Prescotts.

good for them how does that have any effect on amd?

Can Intel keep the same power consumption for Prescotts up to 4GHz?

what are you trying to say here?

New AMD roadmap sucks

Member

Senior Benchmark Addict

Senior Django-loving Member

Senior Member

Senior Django-loving Member

Member

Member

Member

Senior Member

Member

Disabled

Member

Member

Senior Member

Member

Senior Member

Member

Member

Member

Member

Similar threads