AMD Silently Intros 95W Phenom II X4 945

[btarunr]

AMD quietly decked up shelves with a new processor, this time, a low TDP variant of the Phenom II X4 945 quad-core processor. Carrying the model identifier "HDX945WFK4DGI", the Phenom II X4 945 95W comes with a significantly lower TDP rating compared to its 125W predecessor. It first surfaced on a CPU-support list by MSI, that leaked details of some unreleased processors. Based on the 45 nm Deneb core, in the AM3 package, the Phenom II X4 945 has an operating frequency of 3.00 GHz, 3.6 GT/s HyperTransport 3.0 system interface, 512 KB L2 cache per core, and a 6 MB shared L3 cache. It supports DDR2 and DDR3 memory standards. It has a bus multiplier of 15.0x which is upwards-locked since this is not a Black Edition SKU. It will have the same suggested retail price as its 125W version, at US $225.



View at TechPowerUp Main Site

 

[lococol]

should these offer better overclocks?

 

[btarunr]

should these offer better overclocks?

Looking at AMD's past record with Phenom 9750 125W and 95W, I'm skeptical.

 

[Winterwind]

should these offer better overclocks?

maybe little.. but they should oc better with low vcore

 

[TheMailMan78]

should these offer better overclocks?

I doubt it with a locked multi.

 

[btarunr]

I doubt it with a locked multi.

The 125W one has a locked multiplier too.

 

[Darknova]

I think AMD missed out by not making this a BE, a 95W quad would be nice with an unlocked multi.

 

[TheMailMan78]

The 125W one has a locked multiplier too.

I was thinking of the 955

 

[mdm-adph]

Well, regardless of how it overclocks, that's a nice TDP. For people with boards that have shitty MOSFETS like me, it's great.

 

[btarunr]

I was thinking of the 955

He was asking about how it compares to the 125W version.

 

[Cheeseball]

Now all they have to do is make 65W versions (that aren't E parts) and we should have efficient processors. (Or at least to catch up to the awesomeness of the Q6600.)

 

[mdm-adph]

Now all they have to do is make 65W versions (that aren't E parts) and we should have efficient processors. (Or at least to catch up to the awesomeness of the Q6600.)

What's wrong with the E parts? Got the same amount of cache and everything.

 

[newtekie1]

Hopefully this means that whole issue with the 140w barrier isn't really an issue. Maybe the first batch of 965s will be 140w, but hopefully they can release a second batch that is 125w, and something higher than the 965 that pushs back up to 140w.

What's wrong with the E parts? Got the same amount of cache and everything.

They are rediculously expensive, and have low clock speeds and multipliers...

 

[1Kurgan1]

Now all they have to do is make 65W versions (that aren't E parts) and we should have efficient processors. (Or at least to catch up to the awesomeness of the Q6600.)

Then you need to have Intel make chipsets that don't consume so much power. The final results for AMD or Intel are pretty close in Wattage, one requires more to the proc and one requires more to the chipset.

Hopefully this means that whole issue with the 140w barrier isn't really an issue. Maybe the first batch of 965s will be 140w, but hopefully they can release a second batch that is 125w, and something higher than the 965 that pushs back up to 140w.

Is the 965 for sure a 140W model? I haven't looked into it a ton, but I could see it being released as a 125W model seeing as they just made this move.

 

[newtekie1]

Is the 965 for sure a 140W model? I haven't looked into it a ton, but I could see it being released as a 125W model seeing as they just made this move.

Since it hasn't been released yet, we aren't 100% for sure, at least I'm not. However, there is some evidence to suggest it. Main with motherboard manufacturers listing it in their support lists as 140w. Which is usually a pretty good indicator that it will be 140w.

 

[mdm-adph]

They are rediculously expensive, and have low clock speeds and multipliers...

Yeah, and that's why no one in their right mind should be buying E parts for their clock speeds. That's not the point.

 

[newtekie1]

Yeah, and that's why no one in their right mind should be buying E parts for their clock speeds. That's not the point.

Actually, I think that was exactly his point.

 

[Bjorn_Of_Iceland]

nice

 

[mdm-adph]

Actually, I think that was exactly his point.

I don't think I see any point, sir.

Why don't Ferraris cost the same as Fords?

Ferrari seriously needs to make a car that costs the same as a Ford.

 

[Bjorn_Of_Iceland]

Now all they have to do is make 65W versions (that aren't E parts) and we should have efficient processors. (Or at least to catch up to the awesomeness of the Q6600.)

Well I do wish you have the numbers to prove that. If so, thats great.. AMD users are now experiencing performance on what intel users have been experiencing ever since three or more years ago

 

[newtekie1]

I don't think I see any point, sir.

Why don't Ferraris cost the same as Fords?

Ferrari seriously needs to make a car that costs the same as a Ford.

I didn't say I agreed with is point, I was just explaining what he meant.

He wants the best of all worlds: Cheap, Low Power, and Fast.

The normal Phenoms are Cheap and Fast, and the Es are Low Power. Personally, I'd at least like to see Low Power and Fast, or Low Power and Cheap. I don't see why the low power chips are suddenly being priced so high, they should at least be close to the price point of similar performance non-low power chips, a little bit more expensive is understandable, but the markup on them now is just insane. Intel and AMD are both guilty of this...

You car example is flawed. A Ferrari is a completely different car from a ford, for the most part at least...

A Ferrari is fast and expensive, a Ford cheap(relatively). Of course, this ignores the Super Snake, which at ~$75,000 is very close to being as fast as a Ferrari for a fraction of the price.

And ignoring all that, you are comparing to different companies in two completely different markets. Here we are comparing a single company, producing products for the same market.

A better anology would be Toyota trying to sell the Camery Hybrid for $10,000 more than the XLE. As it is now, the Hybrid is slighly more expensive, but not outragously so, only about $250.

 

[johnnyfiive]

BE or not, they all seem to overclock about the same. My 920 reached the same clocks as a 940 would. The 945's are a hell of a bargain IMO. Cheaper than a 955 but should reach similar clocks. Plus, you will be HT oc'ing which yields better performance across the board, litearlly.

 

[TheMailMan78]

You guys are talking about Mustangs. I must be involved. Whos side am I on? newtekie1 or mdm-adph?

 

[Kitkat]

yay there it is finaly lol

Hopefully this means that whole issue with the 140w barrier isn't really an issue. Maybe the first batch of 965s will be 140w, but hopefully they can release a second batch that is 125w, and something higher than the 965 that pushs back up to 140w.

That 965 had alot of incorect info to begin with most of that is mobo manus fault and the web going apesh!t over early info. They anounced they got it down to 95w/3.0 a while ago. (2 months?) I think they have been testing it all this time. 975 should be awesome. The first batch of 965 WILL be 125w and the second.. And yes its 100% sure lol

You guys are talking about Mustangs. I must be involved. Whos side am I on? newtekie1 or mdm-adph?[/url]

rofl rofl

 

[mdm-adph]

I didn't say I agreed with is point, I was just explaining what he meant.

He wants the best of all worlds: Cheap, Low Power, and Fast.

The normal Phenoms are Cheap and Fast, and the Es are Low Power. Personally, I'd at least like to see Low Power and Fast, or Low Power and Cheap. I don't see why the low power chips are suddenly being priced so high, they should at least be close to the price point of similar performance non-low power chips, a little bit more expensive is understandable, but the markup on them now is just insane. Intel and AMD are both guilty of this...

You car example is flawed. A Ferrari is a completely different car from a ford, for the most part at least...

A Ferrari is fast and expensive, a Ford cheap(relatively). Of course, this ignores the Super Snake, which at ~$75,000 is very close to being as fast as a Ferrari for a fraction of the price.

And ignoring all that, you are comparing to different companies in two completely different markets. Here we are comparing a single company, producing products for the same market.

A better anology would be Toyota trying to sell the Camery Hybrid for $10,000 more than the XLE. As it is now, the Hybrid is slighly more expensive, but not outragously so, only about $250.

Wasn't necessarily comparing Ford and Ferarri cars -- I was comparing Ford and Ferarri buyers (which are completely different). My point is that enthusiasts (like we have here) shouldn't be complaining about the price of E parts, just like I don't think E-part buyers complain about the wattage of enthusiast parts. They're for different markets. My problem isn't anything to do with E-parts -- it's that the poster seemed to imply that the exact same E-part chip could be made better simply by not making it an E-part chip.

And the Super Snake's not a Ford -- it's a Shelby.

You guys are talking about Mustangs. I must be involved. Whos side am I on? newtekie1 or mdm-adph?

http://www.justanothergamingblog.com/wp-content/uploads/2009/07/875694-nerd_rage_super.jpg

You will NOT take my side. I don't even care if you agree with me.