AT Ocing Review of Intel's 65nm line...impressive!!

[Duvie]

http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2578...

4.5ghz on a single core effortlessly....WOW....


How about 4.25ghz on the presler....


AMD better bring something big to the table when these things arrive or the enthusiast may jump on these...especially if the price is right....

 

[n7]

Hmm, interesting.

Really though, i want to see Netburst die sooner, rather than see yet another last gasp rehash :roll:

Still no reason to switch from an X2 as far as i am concerned.

 

[Viditor]

4.5 GHz with the voltage pumped, 4.25 without. I wonder how it will compare to the 5000+?

 

[AkumaX]

hmmm... I'd like to put a Conroe against an X2...

 

[Viditor]

Originally posted by: AkumaX
hmmm... I'd like to put a Conroe against an X2...

Do you think it will still be an X2 when Conroe finally arrives?

 

[bunnyfubbles]

Originally posted by: AkumaX
hmmm... I'd like to put a Conroe against an X2...

Sounds like you'll have to wait a year

However this does look like its going to be good news, as it does actually make the P-D seem like a viable option as opposed to severly restricted niche senarios. However Intel is going to have to release a P-D higher than 3.2GHz to really put pressure on the X2, even if they can hit 4.2+ without much trouble. While they'd still be enthusiast alternatives, it won't help to drop the price of the X2 as quickly.

 

[AkumaX]

Originally posted by: Viditor

Originally posted by: AkumaX
hmmm... I'd like to put a Conroe against an X2...

Do you think it will still be an X2 when Conroe finally arrives?

haha, yeah, isn't that the sad part? that we'd have to wait an entire year to compare it to a part that's available now... i know AMD's got something up its sleeve, so in a year, we'll have our sub $200 dual proc X2 6000+ P)

 

[Leper Messiah]

meh. Big Freakin' Deal. 4.25GHz P-Ds are like what? 2.6 Ghz X2s? So when F revision or whatever X2s come out and we can hit 3 GHz on air, what then?

 

[Lonyo]

Hopefully Intel will keep its prices low and the 2.8GHz dual core will be cheap and overclock well, and somehow Intel may end up with the budget overclockers dream (although it'll still end up fairly hot compared to an overclocked 3800+, but not as hot as it used to be, and cheaper)

 

[stevty2889]

A Presler at 4.25ghz could actualy give an AMD dual core a run for it's money in encoding if it's not excessivly overpriced and doesn't require excessive cooling to keep from throttling. My 830@stock is ~14% behind my X2@2.618, so the extra 1.25ghz could give it a decent lead and the extra cache might even be benificial in this case..

 

[biostud]

the dual cores seems most impressive to me, while the p4 are just a p4 "done right".....finally.

 

[Pabster]

How about 4.25GHz x 2 cores?

Intel was spanked down, but you knew they wouldn't stay there.

Perhaps 2006 won't be as dismal as I suspected. :laugh:

Edit: Hmmmmm...Presler 4.25GHz x 2, each die with 2MB L2, I'd like to see some performance figures

 

[Leper Messiah]

performance numbers would be nice, but I thought the equation was something like 1.6*AMDCPU=INTELCPU which would make my X2 like a 4 GHz dualie. The extra cache may prove to be benefical, as long as its not higher latency...

EDIT: And you gotta think cherry picked. remember PCPER's review of the venice's that did 2.8 on stock volts? How many times did they actually do that...

But hey, if I can pick up a cheap one and clock it good, Dual 4 GHz running QMD would sure be nice for F@H.

 

[Elcs]

Its a step in the right direction for Intel I think.

Heat improvements, higher clock speeds and probably better performance could see them pull up to or near AMD's level currently.

If Intels' procs can reach that high at this stage, this could get very heated between AMD and Intel overclockers.

 

[Pabster]

Originally posted by: Leper Messiah
performance numbers would be nice, but I thought the equation was something like 1.6*AMDCPU=INTELCPU which would make my X2 like a 4 GHz dualie. The extra cache may prove to be benefical, as long as its not higher latency...

Yeah, that has to be considered. We know that while 4.25GHz sounds big, AMD offers similiar performance around 2.8ish. So if AMD releases an X2 5200+ (?) it would be a wash.

But hey, if I can pick up a cheap one and clock it good, Dual 4 GHz running QMD would sure be nice for F@H.

Perhaps they'll do like the Pentium D series and release them real cheap. We'll see.

 

[Vee]

Hold your horses.
We have no sustained benchmark loads to evaluate, yet.

The Cedar Mill/Pressler puts a lot of Watts on a very small area. While the lower overall power means total cooling capacity requirements are more modest, it looks as if CedarMill/Pressler heat/area is worse than Smithfield/Prescott at stock clocks.

On strategic hotspots on the die are integrated heat sensors that will control different types of throttling. Intel have diligently pursued this technology to get heat buildup in hand. So the clockrate may well be illusory. You really have to benchmark it. Load it for an hour and look at the 'work' output. Achievable nominal clockrate doesn't say much.

 

[Markfw]

Still power hogs compared to X2, and as pointed out above, and X2 at 2.6 may beat them at 4.5 ! benchmarks !!!!

 

[Markfw]

Originally posted by: stevty2889
A Presler at 4.25ghz could actualy give an AMD dual core a run for it's money in encoding if it's not excessivly overpriced and doesn't require excessive cooling to keep from throttling. My 830@stock is ~14% behind my X2@2.618, so the extra 1.25ghz could give it a decent lead and the extra cache might even be benificial in this case..

Mt 820@3.4 is way behind my X2@2550 (as much as 50% in encoding), so 4.5 MIGHT even the field. By then the X2 can do what ?

 

[Pabster]

Yeah, we need some 65nm X2 samples to do a fair comparison

One thing I noted in Anand's article is that AMD is 6 months behind Intel on 65nm production -- which I stated months ago -- and many here denied vehemently.

 

[Leper Messiah]

Originally posted by: Pabster
Yeah, we need some 65nm X2 samples to do a fair comparison

One thing I noted in Anand's article is that AMD is 6 months behind Intel on 65nm production -- which I stated months ago -- and many here denied vehemently.

Yes, but you must qualify that with the fact that AMD's 90mm Procs run cooler than Intel's 65mm Procs. Smaller arc is the be all end all.

 

[Lithan]

Uhh a good prescott can do 4.25+ stably. Unless these chips improve noticible (as they might, but aren't likely to) before release, it's no big deal. By the time I'd expect these to start clocking where it'd be worth it, Intels p3-decendants should render it obsolete.

 

[Vegitto]

Well, if these aren't cherrypicked and have similair performance per clock to Prescott/Northwood, I might get an Intel. But, you have to remember, these next-gen Intels shouldn't be compared to the current-gen AMD's, or vice versa.

You know what I hate about AMD? The fact that we know nothing about new products until like 2 weeks before launch.

 

[stevty2889]

Originally posted by: Markfw900

Originally posted by: stevty2889
A Presler at 4.25ghz could actualy give an AMD dual core a run for it's money in encoding if it's not excessivly overpriced and doesn't require excessive cooling to keep from throttling. My 830@stock is ~14% behind my X2@2.618, so the extra 1.25ghz could give it a decent lead and the extra cache might even be benificial in this case..

Mt 820@3.4 is way behind my X2@2550 (as much as 50% in encoding), so 4.5 MIGHT even the field. By then the X2 can do what ?

I'm only refering to AutoGK. I get 1h 21m with my 830 at stock, and 1h10m with my X2 @2.618ghz. The X2 also has a newer hard drive, and faster memory as well. I am also basing the fact, on Techno's 4.4ghz prescott being ~24% behind, so with some margin for error, a dual core at the same speed should get ~45-50% improvement over hyperthreading, which would put it about 20-25% faster than my X2. Of course until somebody manages to get one, it's still speculation, and it will still have huge power consumption and massive heat.

 

[lifeguard1999]

It is an interesting article, but there were no benchmarks. It is too early to decide right now if these will match or beat the AMD CPUs when they arrive. (And yes, 4.5GHz does get my attention!)

 

[TraumaRN]

Not to sound like a fanboy or a skeptic but I just have one problem with all this, throttling and heat....what are the temps on these new Intel procs...are you going to be frying eggs still at 4.5 Ghz? Thats what I want to know