Sign in with
Sign up | Sign in
Your question

Finally, AMD made a good choice; FX-76 Cancelled

Last response: in CPUs
Share
March 24, 2007 12:20:11 AM

This is what would have happened...
Related resources
March 24, 2007 12:28:17 AM

Does that pic mean nothing, nice calm night, or fire in the sky, end of world?

:lol: 
March 24, 2007 12:28:50 AM

Quote:
This is what would have happened...


hahah
that reminds me of Contra III
March 24, 2007 12:44:10 AM

Very smart of them. Last thing they need are more 90 nm processors floating around when they are already overstocked on 90 nm.

I'm looking forward to barcelona..Atm I'm on a icky p4 prescott....Going to build a new rig within a year. If barcelona takes the crown from the c2d going to nab one for my rig ^_^...
March 24, 2007 1:07:38 AM

True Dat!!! For AMD to release a chip, the stock speed must be stable for 24/7 use, and to do this a margin below the actual max operating speed is held, and it seems to be in the 15 to 20% range, whereas Intel has been more liberal with their margins in the past, as a result they overclock higher percentage wise...
March 24, 2007 1:19:32 AM

Quote:
Does that pic mean nothing, nice calm night, or fire in the sky, end of world?

:lol: 
It's a blackout...
March 24, 2007 1:55:02 AM

Quote:
Very smart of them. Last thing they need are more 90 nm processors floating around when they are already overstocked on 90 nm.

I'm looking forward to barcelona..Atm I'm on a icky p4 prescott....Going to build a new rig within a year. If barcelona takes the crown from the c2d going to nab one for my rig ^_^...


The last thing they can do is get 3.0 GHz out of 90 nm.... they cancelled because they can't do it.

8O jack i'm surprised... isn't the x2 6000+ 3.0ghz. :?: .. iirc i read in a pc something magazine that the 6000+ is th fx 7? for am2... but i could be wrong... x2 6000+
March 24, 2007 2:18:26 AM

Quote:
Very smart of them. Last thing they need are more 90 nm processors floating around when they are already overstocked on 90 nm.

I'm looking forward to barcelona..Atm I'm on a icky p4 prescott....Going to build a new rig within a year. If barcelona takes the crown from the c2d going to nab one for my rig ^_^...


The last thing they can do is get 3.0 GHz out of 90 nm.... they cancelled because they can't do it.

8O jack i'm surprised... isn't the x2 6000+ 3.0ghz. :?: .. iirc i read in a pc something magazine that the 6000+ is th fx 7? for am2... but i could be wrong... x2 6000+

That is why I posted 3.0 GHz is the last that could do, the next bin is unreachable.

My suspicioun is they are also getting this last top bin out with an aggressive poly line width tweak, which means yields are likely depressed.

This is typically how both AMD and Intel push a top bin, Extreme or FX edition CPU and why they cost so darn much.... (and consequently, why Intel still charges > 900 bucks for 965EE).

ok sorry been a long day my brain interpreted your post incorrectly. i read they could not do 3ghz, not, 3ghz is the best they can do with 90nm.
my bad.
March 24, 2007 2:29:22 AM

The last few FX's are all so close in speed and are still 90nm, so even if you had AM2, upgrading to one made absolutely no sense. At least someone is thinking.
March 24, 2007 2:38:48 AM

Was the FX-76 ever on AMD's official roadmap? I've only seen it speculated about.

Just wondering since the same author who announced the FX-76 to the world (Link) is the same one saying it's been cancelled.

Not that I don't trust the Inq or it's staff.
March 24, 2007 2:39:31 AM

Quote:
Very smart of them. Last thing they need are more 90 nm processors floating around when they are already overstocked on 90 nm.

I'm looking forward to barcelona..Atm I'm on a icky p4 prescott....Going to build a new rig within a year. If barcelona takes the crown from the c2d going to nab one for my rig ^_^...


The last thing they can do is get 3.0 GHz out of 90 nm.... they cancelled because they can't do it.

8O jack i'm surprised... isn't the x2 6000+ 3.0ghz. :?: .. iirc i read in a pc something magazine that the 6000+ is th fx 7? for am2... but i could be wrong... x2 6000+

That is why I posted 3.0 GHz is the last that could do, the next bin is unreachable.

My suspicioun is they are also getting this last top bin out with an aggressive poly line width tweak, which means yields are likely depressed.

This is typically how both AMD and Intel push a top bin, Extreme or FX edition CPU and why they cost so darn much.... (and consequently, why Intel still charges > 900 bucks for 965EE).Baron will chime in that it's Intels fault for causing this price-war, and that's why they cancelled it. He's gotta pass-the-buck, whenever he can. :wink:
a b à CPUs
March 24, 2007 2:46:10 AM

nothing unusual, kinda like the 4ghz P4 being canned - with a new core around the corner, no point in embarrassing them selves like intel did with the 1.13ghz Coppermine P3.
March 24, 2007 3:09:55 AM

Quote:
nothing unusual, kinda like the 4ghz P4 being canned - with a new core around the corner, no point in embarrassing them selves like intel did with the 1.13ghz Coppermine P3.
The difference being, that Intel could have released a 4GHz P4...no problems...not so for AMD's FX-76.
March 24, 2007 4:10:22 AM

Quote:
nothing unusual, kinda like the 4ghz P4 being canned - with a new core around the corner, no point in embarrassing them selves like intel did with the 1.13ghz Coppermine P3.
The difference being, that Intel could have released a 4GHz P4...no problems...not so for AMD's FX-76.
Not that 4ghz meant much on p4.

Good one.@ Vern & JJ: Do you couple'a cowboys want me to get all huffy now, like the Baron? :D  :D 

Bite Me!! :( 

Both your comments are beside the point that i was trying to make.... Hmmmm. :wink:
a b à CPUs
March 24, 2007 8:10:48 AM

Quote:
nothing unusual, kinda like the 4ghz P4 being canned - with a new core around the corner, no point in embarrassing them selves like intel did with the 1.13ghz Coppermine P3.
The difference being, that Intel could have released a 4GHz P4...no problems...not so for AMD's FX-76.

they could have released an 8ghz but the fire retardant case would have cost more then the cpu, and i dont think intel could find any material that could survive 12,000+ *c either 8O
March 24, 2007 12:28:30 PM

Quote:
Baron will chime in that it's Intels fault for causing this price-war, and that's why they cancelled it. He's gotta pass-the-buck, whenever he can.



Get off my jock little girl.
March 24, 2007 12:31:04 PM

Quote:
Do you couple'a cowboys want me to get all huffy now, like the Baron?


Stay off my jock little girl, unless you have something you need to get off your chest... in which case BedStuy is where you need to be not anonymously talking crap on a Forum.
:twisted:
March 24, 2007 1:01:14 PM

Quote:
is where you need to be not anonymously talking crap on a Forum.
:twisted:


:wink:
March 24, 2007 1:11:31 PM

Quote:
is where you need to be not anonymously talking crap on a Forum.
:twisted:


:wink:

Check my location.
March 24, 2007 1:38:19 PM

This is great news, it shows that AMD is starting to make some better decisions instead of launching new processors on the outdated line they have. Although you have to give it to them that the althlon processor lasted so long. How long have they been running on the same/similar architecture?? its been a while. That just proves that AMD knows what its doing in the long run.

They seem pretty confident that they will come back so hopefully AMD's confidence along with some great price drops :)  :)  :)  :)  can boost everyone's confidence.

The only thing they need to do with their older lineup is to stop making processors of every variation. Sure if they still have stock then they should continue to sell them but this whole thing about 2.2ghz 512kb and 2.2ghz 1mb is stupid. Get rid of the 1mb variants in the low end to make the low end cheaper and keep the mid-high end just 1mb. I bet that would save money

Its just like video card lineups when they become outdated. Manufacturers stop making certain models and keep making the best sellers.
March 24, 2007 1:56:39 PM

They even wanted to release 3.4 GHz on 90nm...
Come on... even Prescott broke 100W above 3.4 GHz :roll:
March 24, 2007 2:51:52 PM

I'm lovin' that URL 8)
March 24, 2007 2:54:43 PM

Quote:
Baron will chime in that it's Intels fault for causing this price-war, and that's why they cancelled it. He's gotta pass-the-buck, whenever he can.



Get off my .....<deleted>.... .

Baron, a word of warning.... this is one of those statements that many people find repulsive, the thought of just using it to initiate an insult reflects on your moral character.


Reported.

Jack

Yeah, mocking people is much more tasteful.
March 24, 2007 2:59:43 PM

Quote:

Actually, you did make a good point.... there was still clocking room and the P4... but as Otellini said... you can't put a 150 watt CPU into a desktop box. Power is what limited it.... and neither AMD (even with SOI) nor Intel are able to breach that barrier at the moment.

Wait for high-K...
Yeah, they were still on 90nm IIRC when they canned the 4GHz part. But if you look at overclocks with Cedar-Mills(631-661), 4GHz with default vCore and cooling is a minimum, with 4.4GHz being more the norm. Heck, even the new Cedar-Mill based Celeron's are clocking between 4.5GHz-5GHz on air. Too bad they had to increase the pipeline to 31 stages to achieve this. Had they got these kind of clocks with a 20 stage pipeline(ala Northwood), and the 2MB L2(and all the other "good" Prescott attributes), they would have been a pretty good performer. :x Oh well, hindsight is 20/20. :wink:
March 24, 2007 3:50:34 PM

Quote:
Very smart of them. Last thing they need are more 90 nm processors floating around when they are already overstocked on 90 nm.

I'm looking forward to barcelona..Atm I'm on a icky p4 prescott....Going to build a new rig within a year. If barcelona takes the crown from the c2d going to nab one for my rig ^_^...


The last thing they can do is get 3.0 GHz out of 90 nm.... they cancelled because they can't do it.

Not with decent yields, anyway.
March 24, 2007 4:01:39 PM

Quote:
Very smart of them. Last thing they need are more 90 nm processors floating around when they are already overstocked on 90 nm.

I'm looking forward to barcelona..Atm I'm on a icky p4 prescott....Going to build a new rig within a year. If barcelona takes the crown from the c2d going to nab one for my rig ^_^...


The last thing they can do is get 3.0 GHz out of 90 nm.... they cancelled because they can't do it.

Not with decent yields, anyway.

Good point, they could push it to the ultimate but would likely get just a handful of good die...They may have realized that it would be really pushing the limits of the architecture, and they would have to either raise the vCore for the stability, and possibly install a better HS/F.....which at this point in K8's life would not be worth the added expense and trouble. Also, the last thing they need(considering all their missteps this past year) is..as Apache_lives suggested, a P3 1.13GHz-type embarrassment.
March 24, 2007 4:12:24 PM

Quote:

Yea the other fx 7x's were enough of an embarrassment as it is.


While I have done my part to poke a fun at the Fx 7x saga, it is unfortunately AMD incarnated the 4x4 dual socket DT platform the way they did.... this failed not because of the idea, but because they tried pushing ahead with an architecture that needs a great deal of revision to compete with the competitor revision.

Ultimately, this platform will prove fruitful, but for now it has this black mark simply because of an intense desire to have an answer. I would have prefered they conceived the product but pair up a killer revision of the architecture before giving it to the public.

Jack

Yeah, the idea of whatever QuadFX flavor comes with dual quad-core K10s and dual R600s makes my mouth water, in terms of F@H potential alone, :p 
March 24, 2007 4:13:05 PM

Quote:
Very smart of them. Last thing they need are more 90 nm processors floating around when they are already overstocked on 90 nm.

I'm looking forward to barcelona..Atm I'm on a icky p4 prescott....Going to build a new rig within a year. If barcelona takes the crown from the c2d going to nab one for my rig ^_^...


The last thing they can do is get 3.0 GHz out of 90 nm.... they cancelled because they can't do it.

Not with decent yields, anyway.

Good point, they could push it to the ultimate but would likely get just a handful of good die...They may have realized that it would be really pushing the limits of the architecture, and they would have to either raise the vCore for the stability, and possibly install a better HS/F.....which at this point in K8's life would not be worth the added expense and trouble. Also, the last thing they need(considering all their missteps this past year) is..as Apache_lives suggested, a P3 1.13GHz-type embarrassment.

This is a good point.... the 1.13 GHz P3 debacle is another.... and shows the difference between the two companies in PR....

Rather than address the problem quickly and efficiently, Intel let it stew.

However, recall the 'remove the HSF' video on AMD showing faulty thermal diode.... within a day, Tom's had a update where AMD nipped that right in the bud....

Intel needs to work on the PR and addressing issues like this when they arise.True, and they(AMD) didn't take long to recall the "potentially" faulty Opterons ~1- 1-1/2 yrs ago. Customers like to see that kind of quick response... for sure.
a c 102 à CPUs
March 24, 2007 8:36:13 PM

Thermal dissipation of two different chips will only be similar at the same clock speeds for similar transistor sizes, construction, number, and voltage. The Windsor and Prescott are both 90 nm chips. But the similarities end there. The 3.4 GHz Prescotts (3.4E/F, 550, 550J, 551, 650) run at about 1.4 V Vcore and have a TDP of 115W. The 1 MB L2 versions had 125 million transistors and the 2 MB L2 versions had 169 million transistors. The Windsor FX series runs at 1.35-1.4 volts Vcore.

So at similar clock speeds and voltages, the full-throttle power dissipations of the FX-series Windsors should exceed that of the P4 Prescotts because of a larger transistor count. It's a dual core versus a single core...
March 25, 2007 3:07:21 AM

Quote:
Thermal dissipation of two different chips will only be similar at the same clock speeds for similar transistor sizes, construction, number, and voltage. The Windsor and Prescott are both 90 nm chips. But the similarities end there. The 3.4 GHz Prescotts (3.4E/F, 550, 550J, 551, 650) run at about 1.4 V Vcore and have a TDP of 115W. The 1 MB L2 versions had 125 million transistors and the 2 MB L2 versions had 169 million transistors. The Windsor FX series runs at 1.35-1.4 volts Vcore.

So at similar clock speeds and voltages, the full-throttle power dissipations of the FX-series Windsors should exceed that of the P4 Prescotts because of a larger transistor count. It's a dual core versus a single core...


LOL, you DO realize you're talking to ycon, and he put his anti amd spin on his statement on purpose :wink:

ycon... :roll:
!