Phenom II info leaks out, AMD hints at something

[Sparky]

But its also 125W at 3Ghz and servers stopped at 2.7Ghz with 105W TDP or so.

I wouldnt expect high clocks as stock anyday soon. Not even with 32nm.

I didn't say I'd expect it to happen, just said it is possible, however slight that possibility might be. The TDP may come down over time too.

[Donnie27]

I didn't say I'd expect it to happen, just said it is possible, however slight that possibility might be. The TDP may come down over time too.

Sure it does. Usually later steppings improve and the TDP can be adjusted downward. That's the case for AMD and Intel. You can also add that Errata is fixed on those later steppings as well. Both AMD and Intel fixed bugs on Phenom and Penryn (E0 updates) respectively.

I'm looking at Phenom II simply because I didn't like the Bugs in my E8400 as compared to the fixed ones. I had to Flash the BOIS to even try the *fixed stepping. I want to buy a Phenom II because I feel that was a y move by Intel. It's the kind of stuff that will get worse if AMD falters. So yes I hope those Phenom II's are affordable, not as hot and can overclock much better than their counterparts. I don't need to reach 4GHz a stable reliable 3.4GHz would be geat IMHO!

[Eson]

But its also 125W at 3Ghz and servers stopped at 2.7Ghz with 105W TDP or so.

I wouldnt expect high clocks as stock anyday soon. Not even with 32nm.

i7 has 130W @ 3.2GHz. Whats your point?

[Helmore]

i7 has 130W @ 3.2GHz. Whats your point?

His point is that the moon is made out of cheese....
The 2,8 GHz. is running at 95 W. and thus a 3,2 GHz. stock clocked part isn't that far out of reach.

[Shintai]

i7 has 130W @ 3.2GHz. Whats your point?

And if you would tell me there would be a stock 3.6-4Ghz i7 I would laugh and call you an idiot.

However, it seems saying the same for a 3.6-4Ghz Phenom II is something you get lynched for.

If you want that speed at stock. Then play with netburst or maybe a dualcore.

Maybe 32nm with alot of luck can do it. But surely not 45nm.

[Rammsteiner]

Im quite surprised to see that slide of Athlon X2's just going on, I thought they quit it. But if an Athlon X2 based on Deneb with 6MB L3 cache hits the market, that thing is going to fly

Saw some interesting thoughts about FX-line. I dont know whether AMD is in the position to release an expensive product or not. Keep in mind that not all FX/EE CPU owners are OC'ers, they see those CPU's as the best and want it. If AMD manages to release a FX CPU clocked at say 3.4~3.6Ghz for say ~750Euro with good performance Im sure people are going to buy it, although probably only non-OC'ers with too much money. However, as Ive said before, FX does have to come with something special or BE has to be quit.

[Jazzman]

Special THANKS to flippin waffles for this news SOURCE.

AMD decided to postpone 45nm desktop chip (Deneb, e.g. Phenom II) for hard launch on January 8, 2009 - first day of CES 2009. This way, complete 45nm production in 2008 will be branded as Opterons and target higher ASPs.
But don’t think for a second that AMD decided to drop the towel on desktop market: Deneb and Deneb FX are turning to be something really special. Unlike current batch of 65nm processors, Phenom II will come to market with 50% overclocking headroom, if you obey the essential rules of overclocking. In fact, Phenom II will probably be the most overclockable processor AMD released to market in the past eight years.

According to our sources, AMD worked with some selected enthusiasts to enable LN2 cooling as well and the CPUs should not suffer from “cold bug”. Does this means we’ll start seeing 5GHz+ clocks on AMD processors as well?

For now, all that we can do is wait and see can AMD get back in the game or not. Judging by our sources, ATI guys are making great changes inside the company and changing the product line-up in ways that old AMD would never think of.

[Sparky]

And if you would tell me there would be a stock 3.6-4Ghz i7 I would laugh and call you an idiot.

However, it seems saying the same for a 3.6-4Ghz Phenom II is something you get lynched for.

Who is lynching you for saying that 3.6-4GHz stock is laughable? If you are referring to me, I'm not. At this point it is. But all I am saying is it may be possible to have 3.4-3.6GHz stock, MAYBE, as in slight chance (cherry picked chips for the FX for example). Me saying that is in no way lynching you

[eric66]

'ATI guys are making great changes inside the company and changing the product line-up in ways that old AMD would never think of.' wonder what does that mean

[Eson]

I don't think we will be seeing any future processors above 3.4-3.6GHz stock because we hit the wall. And with more cores, higher clocks becomes less important.

[Shintai]

Special THANKS to flippin waffles for this news SOURCE.

You do know who Theo is?

Theo’s Bright Side Of IT

Here is a reminder....

http://www.theinquirer.net/en/inquir...30000-3dmark06

One thing is sure, he aint bright!

[Hornet331]

You do know who Theo is?



Here is a reminder....

http://www.theinquirer.net/en/inquir...30000-3dmark06

One thing is sure, he aint bright!

ack, i would like to see this by another source.

[Zucker2k]

There's a lot of optimism in here, I like it. I'll love to see AMD make a strong move this holiday season. It could mean that I pick up a good deal on a new platform or a dirt cheap upgrade on core i7.

[BrowncoatGR]

Imo Phenom II 940 will cost $400 at release as it's direct competitor is the non-XE 9650 and not i920. Intel in it's effort to not cannibalize 775 sales has done AMD a great favor here and priced the entire i7 platform in the extreme high-end category, and i fully expect AMD to take advantage of that. I also think that AMD will release at least a 3.4GHz cpu by the end of the year.

You mean the EEtimes article where it states that amds (ibms) 45nm has a 62% lower pmos driver current(660 vs 1070 µA/µm) then intels 45nm? (higher is better btw.)

Amd/ibm has a long way to catch intel on transistor performance. But transistor performance alone means nothing when the design cant keep up.

i guess you missed this part

The transistor drive current for AMD's 45-nm devices is much lower than that of the Intel HKMG transistors. But power consumption is quickly becoming a high priority for server chips. AMD's transistors exhibit very low channel leakage. Our transistor benchmarks indicates that leakage current is less than one-third of the value measured on AMD's 65-nm process. It's also significantly lower than the Intel 45-nm HKMG process. In fact the Ion/Ioff ratio for AMD's PFET is nearly 10 times better than that for the Intel PFET.

source: eetimes

ack, i would like to see this by another source.

Well i really don't think AMD would release slides mentioning overclocking unless they had something to show. In fact the early slides on ACC were quite accurate in what they promised which was 200mhz more on average.

[Helmore]

Imo Phenom II 940 will cost $400 at release as it's direct competitor is the non-XE 9650 and not i920. Intel in it's effort to not cannibalize 775 sales has done AMD a great favor here and priced the entire i7 platform in the extreme high-end category, and i fully expect AMD to take advantage of that. I also think that AMD will release at least a 3.4GHz cpu by the end of next year.

Slight correction there, although I think you meant to say that. IMHO, I think that would be feasible to do for AMD, it will probably be the limit of what their Phenom II is capable of.

Well i really don't think AMD would release slides mentioning overclocking unless they had something to show. In fact the early slides on ACC were quite accurate in what they promised which was 200mhz more on average.

I think Hornet is indirectly responding to Jazzman's post....

[Rammsteiner]

Whoa, awesome they actually worked on the coldbug I mean, this isn't something for AMD, nor Intel btw, to fix. Good they do really care about the enthusiasts by doing this

[BrowncoatGR]

Whoa, awesome they actually worked on the coldbug I mean, this isn't something for AMD, nor Intel btw, to fix. Good they do really care about the enthusiasts by doing this

That's just Theo talking. I won't believe that until i see benches

[Sparky]

That's just Theo talking. I won't believe that until i see benches

x2
though part of me doesn't care about coldbug, I'm only on water.

[Helmore]

Whoa, awesome they actually worked on the coldbug I mean, this isn't something for AMD, nor Intel btw, to fix. Good they do really care about the enthusiasts by doing this

Could you tell me how the current Angena's react under some LN2? This is what biohead said:

the coldbug with phenom doesn't seem to be related to HTT reference clock speeds anymore. and chips that boot under cold at stock settings will overclock under that cold. this means some major changes happened along the road, but we haven't kept track, as all the cold "fun" is at intel's feet.

Is there any more proof of this?

[Cooper]

Could you tell me how the current Angena's react under some LN2? This is what biohead said:


Is there any more proof of this?

they moved clock gen to south bridge on SB750. though I haven't seen any tests prooving this.

[Donnie27]

they moved clock gen to south bridge on SB750. though I haven't seen any tests prooving this.

That's what I was told as well! I'd be surprised if it isn't true. Sometimes on die isn't always the best

[biohead]

Could you tell me how the current Angena's react under some LN2? This is what biohead said:


Is there any more proof of this?

just search the AMD section for people running their 9950 under cold. the ones that boot at 200 htt, will also boot at way higher htt.

and take the 525 HTT results from toppc for example. that isn't even phenom but still. usually such a high htt means you can use lots of cold. but toppc can't even post that 500+ htt chip below 0 degrees celcius.

[mAJORD]

And if you would tell me there would be a stock 3.6-4Ghz i7 I would laugh and call you an idiot.

However, it seems saying the same for a 3.6-4Ghz Phenom II is something you get lynched for.

If you want that speed at stock. Then play with netburst or maybe a dualcore.

Maybe 32nm with alot of luck can do it. But surely not 45nm.

January 2006:
FX-60 2.6ghz 1.40vid 110watt

May 2006: AM2 launch.
FX-62: 2.8Ghz 1.40v VID 125w

Aug 2007
6000+ 3.0 Ghz 1.30-1.35vid 89watt
6400+ 3.2Ghz 1.40vid 125Watt

All 90nm

People said K8 would never make it past 3ghz, hell no one could even clock one past 3ghz reliably before late 2006 even with loads of voltage.

Don't get me wrong I think a 4ghz stock chip is DREAMLAND also, but don't discount the possiblity of 3.4 - 3.6ghz parts in a couple of steppings, on 45nm.

I've clocked endless AM2 K8s and the improvements over the last 2 yrs have been great, but there's a very obvious wall as far as the architecture goes, that in my opinion would stop them ever making it past 3.4 at a pinch, BUT K10 is not K8 and it's very obvious that K10 scales higher. Fact is even current Phenoms on problimatic 65nm scale better with voltage than these K8's

bottom line is, it's been shown before how quickly AMD can tweak their process when under pressure.. I don't think anyone would have predicted a 3.2ghz K8 X2 on 90nm back in 2005, that's netburst DC level clocks on an arch that's in complete contrast when it comes to reaching high freq.

We really don't know if 45nm will turn out to be as good as 90nm was, so lets all just wait and see

[stangracin2]

Fact is even current Phenoms on problimatic 65nm scale better with voltage than these K8's

i wouldn't say that because my x2 4000 had a vid of 1.35 and did 3.245 GHz @ 1.46

[[XC] Oj101]

According to our sources, AMD worked with some selected enthusiasts to enable LN2 cooling as well and the CPUs should not suffer from “cold bug”. Does this means we’ll start seeing 5GHz+ clocks on AMD processors as well?

Yeah, we all know that enthusiasts are world class engineers