Q6600 isn't real quad?

[keithlm]

That's because, as you said, the newer BIOS increased Vcore for Phenom from 1.2V to 1.3V, and sometimes up to 1.35V. And it does seem that AMD upped the voltage on Phenom.

http://forum.msi.com.tw/index.php?topic=118051.0

So let's see, in reality Phenom is actually factory overvolted.

EDIT: To further illustrate my point,
http://www.misco.ie/productinformation/~138546~WW~all~/AMD%20Phenom%209850%20quad%20core%20socket%20AM2%20%20processor.htm

Apparently this seller still pegs stock Vcore at 1.2/1.25V

Default VCORE for the 9850 and 9950 is 1.3V on most motherboards. The vcore is set lower for the other Phenom models in the same motherboards. I've never seen a vcore of 1.35V on any motherboard. But I'll accept your claim without going into a diatribe about you not linking proof like some other posters are known to do. Although I will still doubt your facts until I see one.

Your point is not further illustrated by linking to an obscure e-tailer.

Regardless: what would you like me to say about any of this?

Because it won't matter anyway... since I don't agree with you or several other Intel fans on this forum about many issues this post will become hidden fast. Apparently some people do NOT want any discussion about anything except what they approve of on these forums.

It says a lot about their maturity level when several of these people are the worst trolls and/or "button pushers" on these forums.

 

[keithlm]

Sorry if you felt like that, Keithlm, but that wasn't my intention. I DO agree with you regarding the E8400 vs X3 8750 comparison, Keithlm. Actually, I think it's awesome for AMD. However, at the same time I agree with Jimmy, since the E8400 can overclock like there's no tomorrow and Intel is ahead with 45nm. So, we have 2 perspectives here. Both are awesome performers in certain conditions. AMD wins some, Intel wins some and that's it. Probably, what makes people mad is that you usually seem to defend AMD at any costs - or at least that's the impression that everyone has. Sometimes I like some of your posts, just as sometimes I strongly disagree with you, however, I'm not bashing either you or AMD. If we are discussing like this then the purpose of the forum is doing great. It *MUST* happen. It's a good thing until it becomes agressive or fanboyish.

Anyway, we must be aware that just because someone says something great about Intel or AMD it doesn't mean that he or she is a fanboy. Intel is doing a lot better nowadays, but AMD is still kicking in the server environment and budget computers.

(PS: I don't agree with people rating you - or anybody else - down to hide your posts. You *have* to say what you got to say if it's fair and reasonable)

Sorry. You just got caught in the crossfire.

Your last line of your last post set me off:

The problem is that Keithlm will brag about it forever and say that's the rule for AMD: ALWAYS delivering more for less $. As you noticed, he will make the comparisons that will make him happy as a fanboy.

As a general rule I NEVER claim AMD ever delivers more for less $. That would be somebody else that claims that. What I claim is that they deliver about the same for about the same. The problem is that people will have a different definition of the word "better". My definition doesn't include overclocking and it highly discounts any single threaded application benchmark. And when I look at benchmarks... I don't look for a "winner" or a "loser" I just look for parity and verify that the results are about the same. That's all that is needed. To treat them as anything more important than that is not logical.

My biggest "problem" is that somebody will jump in and start focusing on something that really makes no difference to the conversation and will lead it into a direction that will usually end with sarcasm and trading veiled insults.

EDIT: Of course then adding the topic of architecture... things can really go crazy.

 

[yomamafor1]

Obscure e-tailer?

How about newegg?
http://www.newegg.com/Product/Product.aspx?Item=N82E16819103249

or CPU (Computer Power User)
http://www.computerpoweruser.com/editorial/article.asp?article=articles%2Farchive%2Fc0806%2F10c06%2F10c06.asp

No. The Default Vcore seems to have great variance, from down to 1.2V to up to 1.3V. So my guess is that AMD stock over-volted them, or the motherboard OEMs choose to increase the stock Voltage for them.

You can try harder.

 

[keithlm]

Obscure e-tailer?

How about newegg?
http://www.newegg.com/Product/Product.aspx?Item=N82E16819103249

or CPU (Computer Power User)
http://www.computerpoweruser.com/editorial/article.asp?article=articles%2Farchive%2Fc0806%2F10c06%2F10c06.asp

No. The Default Vcore seems to have great variance, from down to 1.2V to up to 1.3V. So my guess is that AMD stock over-volted them, or the motherboard OEMs choose to increase the stock Voltage for them.

You can try harder.

What point are your attempting to make that somehow makes it IMPORTANT?

The VCORE of the 9850 and 9950 is 1.3V on most motherboards. It is lower for the lower models. Period. End of story.

The 9850 and 9950 have a voltage rating of 1.05V to 1.3V.... so I don't get your point.... it is within specifications.

 

[dattimr]

Most enthusiast mobos overvolt the CPU and the Northbridge a bit. That's also the case of my Maximus Extreme. It even overvolts my memory by 0.6+. You'll probably find quite a difference when comparing different models from different manufacturers (and also when comparing different BIOS releases). Anyway, it's no big deal.

 

[MarkG]

I DO agree with you regarding the E8400 vs X3 8750 comparison, Keithlm. Actually, I think it's awesome for AMD.

It's awesome so long as AMD have plenty of broken quad-cores that they can sell as triple-cores and make some money rather than throwing them away. But it's not so awesome if they have to start taking working quad-cores and selling them as triple-cores because there's not enough of a market for the quads.

I once worked for a company doing something similar, and it was a great idea in theory, but the actual number of chips that were broken but not so broken as to be unusuable turned out to be quite small; I believe we did end up selling fully working chips with units disabled to meet demand.

AMD have a tricky path to follow between setting the price at a level where they'll sell all the broken quads as tri-cores and not selling so many that they have to start crippling lots of perfectly working quads and selling them at a lower price.

 

[jimmysmitty]

It's awesome so long as AMD have plenty of broken quad-cores that they can sell as triple-cores and make some money rather than throwing them away. But it's not so awesome if they have to start taking working quad-cores and selling them as triple-cores because there's not enough of a market for the quads.

I once worked for a company doing something similar, and it was a great idea in theory, but the actual number of chips that were broken but not so broken as to be unusuable turned out to be quite small; I believe we did end up selling fully working chips with units disabled to meet demand.

AMD have a tricky path to follow between setting the price at a level where they'll sell all the broken quads as tri-cores and not selling so many that they have to start crippling lots of perfectly working quads and selling them at a lower price.

Thats my question too. What happens when the process matures? What happens when they do not have enough faulty quads to meet demands? They currently have no dual core succesor to the Athlon X2 (Phenom X2). And Nehalem is going to scale from dual to octo core.

It probably started out as a great idea but then they too realized it may end up bad and they are asking themselves what do they do next.

 

[TechnologyCoordinator]

The tri-cores are a great reponse to the quad-core manufacturing issue. I don't think AMD is worried about running out of tri-core candidates, in fact, I bet AMD is hoping one day they won't have enough to sell the tri-cores and hopefully can fetch more money for the processors as quad cores.

Yes, the tri-cores are just gimped quad-cores, but this is actually a good move to sell the gimped quads as tri's that clock higher than it would have as a quad (or perhaps the fourth core isn't even functional).

 

[MU_Engineer]

Thats my question too. What happens when the process matures? What happens when they do not have enough faulty quads to meet demands? They currently have no dual core succesor to the Athlon X2 (Phenom X2). And Nehalem is going to scale from dual to octo core.

AMD is not expecting to be making the current 65 nm Phenoms for very much longer. This is evidenced by a couple of AMD's actions:

1. They only have one mask set for the K10 CPUs at the moment- the 65 nm quad core, 285 mm^2 die mask. If AMD was not envisioning a rapid transition to 45 nm, they most likely would have wanted to sell dual-core K10s as they could get more money for them in the marketplace. They would have been based on a native dual-core die as it would be too expensive to make all Phenoms from the big quad-core die. Note that AMD did shell out for a new mask set for the "Griffin" Turion Ultra but nothing with the Phenoms.

2. The numbering of the Phenom X4 line tops out at 9950, which is the current top-dog Phenom. AMD thus doesn't expect the current Phenom generation to go any faster than the current 2.6 GHz but they are not just going to sit around for a long time with the Phenom X4 9950 as the fastest chip. Maybe AMD will resurrect the FX line and bring in 2.7+ GHz Phenoms, but I doubt it since AMD has stated they wanted to reserve the FX moniker for chips they can sell for a lot of money ($700+). They would have to goose the Phenom X4 to a little over 3 GHz to command anything near $700/CPU and while some might do it, the thermals will be unreal (over 200 watts!) and the yields will be pretty low.

AMD has been rumored to have sent out 45 nm processors for validation to motherboard manufacturers but there have been no good concrete stories on AMD's 45 nm development. This is not surprising as AMD is notoriously tight-lipped on actual details as opposed to overall plans (e.g. Fusion) about new products until very shortly before launch, particularly in the last couple of years. I bet that we'll start to see actual information (clock speeds, prices) about 45 nm AMD products maybe a month or a few weeks before launch day, just like with the ATi Radeon HD4000 launch.

It probably started out as a great idea but then they too realized it may end up bad and they are asking themselves what do they do next.

The current execution of the 65 nm K10s is less than ideal. But if you really look at it, it is a test run and little more. There is an old saying about ICs: do not do both a die shrink and a new microarchitecture at the same time or you are asking for trouble. Thus AMD made one K10 mask on an established 65 nm process to get the bugs worked out in making Phenoms. This is exactly the same logic behind the 180 nm P4 Willamette. The real meat of the K10 will be coming in the 45 nm generation just as the real meat of the P4 line came not with the first 180 nm Willamettes but with the 130 nm Northwoods. The Willamette was also a clock-speed-limited processor that was warm, got beaten up on by PIII Tualatins and Athlon XPs, and was a very large die with an insufficient amount of cache. The 65 nm Agena Phenoms are pretty much the same way and I think that the 45 nm die shrink will do to the Phenom what the 130 nm die shrink did for the P4.

 

[TechnologyCoordinator]

I was looking for this and now I finally found it:

This pretty much summarizes the whole thread.

 

[amdfangirl]

the thermals will be unreal (over 200 watts!) and the yields will be pretty low.

Well look at the Barcy based Opterons! 75W TDP for the same clockspeeds and avalible for socket AM2. A 3GHZ Phenom at a reasonable TDP is not unreasonable!

 

[epsilon84]

It's probably 75W ACP, not TDP. 😉

 

[amdfangirl]

^ Sorry, you were right! =P

 

[harna]

Thats my question too. What happens when the process matures? What happens when they do not have enough faulty quads to meet demands? They currently have no dual core succesor to the Athlon X2 (Phenom X2). And Nehalem is going to scale from dual to octo core.

It probably started out as a great idea but then they too realized it may end up bad and they are asking themselves what do they do next.

Easy. Higher bin, higher price, greater return. Won't matter a rats tossbag if low end quad or high end tripple.

 

[jimmysmitty]

^ You completely lost me buddy.

 

[MU_Engineer]

Well look at the Barcy based Opterons! 75W TDP for the same clockspeeds and avalible for socket AM2. A 3GHZ Phenom at a reasonable TDP is not unreasonable!

The socket AM2 Opteron 1300s are the same processors as the Phenoms but since they are third-generation Opterons, they have an average power consumption figure in addition to the TDP.

Opteron 13xx Phenom X4
1352 (2.1 GHz) 75W ACP/95W TDP <none>
1354 (2.2 GHz) 75 W ACP/95W TDP 9550 (2.2 GHz) 95 W TDP
1356 (2.3 GHz) 75 W ACP/95 W TDP 9600 (2.3 GHz) 95 W TDP
<none> 9750 (2.4 GHz) 125 W TDP
<none> 9850BE (2.5 GHz) 125 W TDP
<none> 9950BE (2.6 GHz) 140 W TDP

AMD has socket F Opteron 23xx/83xx operating at 2.4 and 2.4 GHz (x358SE/x360SE) rated with an ACP of 105 watts, which correlates to a 125-watt TDP.

On an interesting note, the socket AM2 Opteron 1300s all ship with the northbridge running at the full 2000 MHz, not the 1600 or 1800 MHz that the non-BE Phenom X4s do.

 

[Wisecracker]

I just want to bump this to 12 pages so ...

I went out to eat last night. Our waitress had been married 3 times and was still a virgin.

I asked her about her first husband. She said he was a psychiatrist and all he wanted to do was talk about it. She then said her second husband was an OB/GYN and all he wanted to do was look at it. When I asked her about her third husband she said ...

He was an AMD fan and all he could say was, "Wait until next year."



(Lighten up people)

 

[jimmysmitty]

12 pages WOOOO!!!!!!!!

 

[TechnologyCoordinator]

I just want to bump this to 12 pages so ...

I went out to eat last night. Our waitress had been married 3 times and was still a virgin.

I asked her about her first husband. She said he was a psychiatrist and all he wanted to do was talk about it. She then said her second husband was an OB/GYN and all he wanted to do was look at it. When I asked her about her third husband she said ...

He was an AMD fan and all he could say was, "Wait until next year."



(Lighten up people)

ROFL.

 

[harna]

As would many. And then we'd all get slammed for being AMD fanboys. :lol:

The best trick of all would be if you could get that CPU for only $200.00 reckon it would be closer to $1000.00 and maybe then some.