Anand uses FX-51 for roundup.

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Sep 15, 2003
139
0
0
I take it Smashp is another poster but I havent seen anything from him yet. Maybe a search will help me learn more of this poster. I take it he is a die hard AMD fan.

I dont want to be labeled as an AMD cheerleader. If you would have asked me several months ago which chip to buy I would first ask you what your budget is.

Less than 100.00 get an Athlon XP 2500+ Barton Core and overclock it.
Around 200.00 get a P4 2.4C Proc. and again Overclock it.

If you asked me about servers I would have gone for a dual XEON. If asked about AMD MP processors I would have told you NO WAY. AMD makes a great inexpensive desktop but not a good server CPU.
-------------
Now that Opteron is Here.
I have had demonstrations from IBM and spoken with a few vendors. I can honestly say wait a few months and get yourself an Opteron when the prices drop and when AMD has full production going. Especially around Week 12, 32, and 50 is the time to look for good overclockers. Like I said before if the Opteron can stay within 10% of the performance of Intel's Best 32 bit CPU's you have to buy the Opteron because of its future capabilities in the 64 bit arena.

On the server end dual and quad Xeons are a waste of money now.

As to Prescott. I know that come 1Q 2004 the 3.8 & maybe even the 3.6 will be more of a paper launch if announced and that the actual chips will not be available to anyone looking to buy one. I suspect Intel could produce a Prescott at 3.4 in decent quantity. But talking to a few people they still dont believe Intel can get much higher than that and have the processor be stable. Hot Spots, Current Leakage, and something I never heard of are all suspected to be the problem. All pointing back to strained silicon issues and design flaws. Yup Intel can make a mistake its called a learning curve as they attempt to move toward newer technologies.

As to Opteron 2.6 on air there is very few of them being sold as 2.2 FX's. So Few that AMD will not announce a 2.6 Opteron or 2.4 because they couldnt possibly provide enough of them. Expect to see AMD get the processor yeilds easily to 2.6 levels by the time a prescott 3.8 paper launches. But Windows XP 64 will be available making more sense to buy an Opteron.

AS TO THE PEOPLE WHO THINK 106 WATTS IS RAISING THIER ELECTRIC BILLS HAS NO CONCEPT THAT PEOPLE ARE TALKING ABOUT A HEAT DISSIPATION PROBLEM OF THE CHIPS!!! DO YOU KNOW HOW HARD IT WILL BE TO KEEP PRESCOTT COOL WITH HEAT DISSIPATION THAT HIGH? ALMOST NO ONE WILL BE ABLE TO SELL A PRESCOTT IN A STANDARD PC CASE WITHOUT EXOTIC COOLING.
 

So how long do you think it will take for AT to ban you for posting flamebait threads like this?
I hope sooner than later.
 
Sep 15, 2003
139
0
0
Truth Hurts? Anyone boasting his rig shows his real side. Anyone opposing your decision must be a fool?

ASUS P4P800
P4 2.8/533 512k


 
Sep 15, 2003
139
0
0
-------------
gosh the price difference of the athlon 64 3200 and Athlon FX is really big.

whats the ultimate difference between these two processors?
-------------

A.) EGO
B.) Bragging Rights.
C.) Compensating for a Small Winkie.

Top dog commands Top Dog Prices. Demand is higher than quantity.
 

Dustswirl

Senior member
May 30, 2002
282
0
0
Has anyone seen the new 20$ bill? :D very nice!!!


:p cool it ppl! /me offers everybody a round o' beer :beer::D:beer:
 

Wingznut

Elite Member
Dec 28, 1999
16,968
2
0
Originally posted by: Ticktanium2038
Especially around Week 12, 32, and 50 is the time to look for good overclockers.
How do you justify that? What is different about the process during those weeks?
ALMOST NO ONE WILL BE ABLE TO SELL A PRESCOTT IN A STANDARD PC CASE WITHOUT EXOTIC COOLING.
WHERE DO YOU COME UP WITH THIS STUFF???



 
Sep 15, 2003
139
0
0
Wingznut (I work for Psychic Friends.)

I read, take notes, Make meetings, and understand what goes into technology instead of just assuming Intel will never make a mistake and that everything AMD does is wrong.

So For all you Intel geniuses you can stop critisizing me and tell the Intel Engineers how to cool 106Watts dissipation on a standard PC case so Prescott can be released today Since you have all the answers. Maybe clear up those problem they cant seem to figure out. Im sure Intel would love to hear the electricity bill theory. But then Thats who is buying into 32 bit prescott cpu's. LOL


On to the Weeks. 12, 32, and 50 are just weeks. This is not 100% pure science and its just my observation. The week numbers will vary but I think are pretty close to being around the week of good CPU makes. For instance week 48 was very good for some CPU's.
----------------------------
My Comment
Especially around Week 12, 32, and 50 is the time to look for good overclockers.
----------------------------

Here is Why.
The maufacturing process yeilds just get better with age. But in reality 12 weeks is 3 months into a production. This far into production the engineers have tweeked the production and the materials that produce the chips are finely tuned and usually yeild better chips. Usually the next speed increase is announced because the improved manufacturing is now producing chips that are a speed grade above previous samples. As time goes on this only gets better and better.

Its Week 1.
So lets say in the middle of the wafer your getting your 2 gig parts and on the outer rim your getting 1.5gig parts.

Weeks go by As the engineers break in the manufacturing and do some calibrations to the equipment and now.

Week 13.
The same process in the middle of the wafer is producing 2.4gig parts while the outer rim your producing 1.8 gig parts.

AMD would then announce 2.4 gig parts available.

Its safe to say that week 12's parts should be capable of 2.4Gig. Take into consideration they have to allow for poor air flow cases and a week12 CPU marked as 2.0 might actually do 2.6 gig.

Now 30 weeks or so into Production.
Now what happens is the 1.5-2.0 parts dry up as everyone is buying the faster 2.2-2.6 parts. So lets say its week 32. The manufacturing is so good they cant make a chip that slow but manufacturers are demanding 1.8 gig parts. AMD/INTEL then remark the 2.2 chips as 1.8 gig parts.

Much like the AMD 2500 is probably a 3200 with a 166 FSB. And the Intel 2.4C is really a 3.2 part just remarked. The manufacturing process is so refined they cant make bad or slower chips and it doesnt pay for them to do so.
 

Duvie

Elite Member
Feb 5, 2001
16,215
0
71
Do you work in a fab, or are familiar with fabbing a cpu?? If not this is a lot of assummming. You couldn't have possibly wathced enough product lines to come up with this as anything scientific. It is a half-baked theory. There is an obvious variable here....That being product cycle. If intel is not in a rush to push the product up I don't think we see those speed upgrades in the sequqnces you suggest...

 
Sep 15, 2003
139
0
0
Its a rough estimate. But its pretty close to what happens. Try looking at good overclocking CPU's and notice the week produced but also take into considertation demand.
 

Wingznut

Elite Member
Dec 28, 1999
16,968
2
0
Originally posted by: Ticktanium2038
Its a rough estimate. But its pretty close to what happens.
No, not really.

Here's the problem, Tick... You write these long posts, making a ton of assumptions with really no basis. And then you present them as sheer fact, which they aren't facts at all.

I don't feel like addressing your every point. But I will address a couple.

Let's take the following quote:
So For all you Intel geniuses you can stop critisizing me and tell the Intel Engineers how to cool 106Watts dissipation on a standard PC case so Prescott can be released today...
(First off, let's presume that 106w is an accurate number.) Do you think that this would be the first 106w cpu ever made? Well, it's not. Not only that, but there are higher power consuming cpu's that live in even smaller cases, that don't use "exotic" cooling.

Also, your whole "middle of the wafer your getting your 2 gig parts and on the outer rim your getting 1.5gig parts" theory is way off base. How wafers yield and bin is not nearly that cut and dry.

But that's your MO. You take some theory that you heard somewhere, apply it as if it were fact, and then many times extrapolate it even further. But when the base is inaccurate, you cannot consider the end result fact.
 

Sahakiel

Golden Member
Oct 19, 2001
1,746
0
71
1. Lots of fanaticism here... I'm seeing hella crap thrown about and I'm not even qualified to start working, yet.

2. I don't see much into Intel having trouble getting Prescott past 2.8GHz. I seem to remember Opteron having problems getting past 800MHz a while a go.

3. 64-bit might really kill Prescott. Of course, this is assuming 64-bit takes of as well as 32-bit did. Me, I'm hoping the stores get some decent deals during midnight sales of Longhorn.

4.
Originally posted by: Wingznut

Let's take the following quote:
So For all you Intel geniuses you can stop critisizing me and tell the Intel Engineers how to cool 106Watts dissipation on a standard PC case so Prescott can be released today...
(First off, let's presume that 106w is an accurate number.) Do you think that this would be the first 106w cpu ever made? Well, it's not. Not only that, but there are higher power consuming cpu's that live in even smaller cases, that don't use "exotic" cooling.
I think I remember reading somewhere that DEC Alphas used to dissipate about 100W or so during normal use.

 

rgreen83

Senior member
Feb 5, 2003
766
0
0
Wow, didnt mean to start an all out flame war, changed title to reflect. (It might have been a little provocative for the big fanboys of each side ;) )

The point I was trying to get across is not which is the chip to buy, intel or amd. Just that AMD makes good chips, this might not save them from the deep pockets of intel in the end though. It is good to see AMD getting see credit, not too long ago ATI wasnt getting any credit either but few would argue that they make some great GPUs now.

Some folks need to chill out, youd think their lives depend on who wins the chip war, I love discussing this stuff and grinding out the facts but I dont think it needs to get to the point of personal insults against each other. There are a lot of smart people in hear with differing opinions so remember that and try not to be so close-minded or you might keep yourself from learning something new. So, behave!

Carry on.
 

BD231

Lifer
Feb 26, 2001
10,568
138
106
I thought watts is a cooler form of power than volts. I'm probably off in lala land, but I could have sworn I read an artical stating computer technology manufactures were leaning towards using more watts over volts to fight the heat output problem.

I know this is the case with ATI.
 

Cerb

Elite Member
Aug 26, 2000
17,484
33
86
Originally posted by: BD231
I thought watts is a cooler form of power than volts. I'm probably off in lala land, but I could have sworn I read an artical stating computer technology manufactures were leaning towards using more watts over volts to fight the heat output problem.

I know this is the case with ATI.

Watts are power...heat. One way to lower the watts used is to lower voltage. But then you raise clock speed, and it evens out, so you still gradually raise the heat being used, save for Athlons, since the original ones got nicknames like "space heaters" and then got cooler from there :).
 

Evdawg

Senior member
Aug 23, 2003
979
0
0
i think its sweet amd is finally coming up with a lead (in my opinion) over intel... its like ATI is finally raping nvidia, well for at least 7 or 8 months they have. but nvidia was the leader for countless years. Now ati is taking the lead and not letting go. Its cool that the underdogs are now the leaders. My machine in amd/ati.. and i love it ;D
 

dguy6789

Diamond Member
Dec 9, 2002
8,558
3
76
here is a road map for amd so far, hope it helps you guys,

athlon 64FX will be at 2.8ghz sometime in q1 2004, thats impressive

http://www.pbase.com/image/17079307/original

and good news for you socket 754 users, amd wont be dumping that socket for a while you will also get to enjoy 2.8ghz processing heh.
 
Sep 15, 2003
139
0
0
That seems to support my week 12 theory. LOL.

AMD has been stockpiling 2.6 but in VERY VERY limited quantities.

I meet with AMD next week and will try to get the scoop on the current manufacturing and how its going. Maybe they are getting 2.6 clean as 2.4 seems pretty easy now for overclockers.

The 2.8 is a minor trouble spot but I suspect they can get 2.8 out of the current process.

3.0 might be a problem that might need some die tweekage. Here is where the IBM/AMD engineers will show what they can do together.

So get those magical Prescott 5 Gig procs ready. Your going to need them.

Whats also interesting is they shows 9 micron in Q1 for AMD.
I wonder if they will beat prescott on 9 micron out the door?
 

Wingznut

Elite Member
Dec 28, 1999
16,968
2
0
Originally posted by: Ticktanium2038
WINGNUTZ.

Enlighten me. Name one 105+ Watt dissipation processor released in standard PC's?
See... There you go again with your ambiguity. What exactly do you mean by "standard" PC? And why are we limiting our discussion to a narrow range of systems, other than to try and prove your point (even though it would still prove nothing.)

The fact of the matter is that a cpu which is rated at 105w+ would not require "exotic cooling."

 
Sep 15, 2003
139
0
0
I see your avoiding the question and not answering with anything usefull.

Why dont you spend another 4 years here posting 10000 your OPINION instead of something truthfull.
 
Sep 15, 2003
139
0
0
Im calling you out wingnutz

Name one 105+ Watt dissipation processor released in standard PC's sold to consumers?

You claim there are many. Tell us. Move me with your wisdom.
 

Wingznut

Elite Member
Dec 28, 1999
16,968
2
0
Well, since you won't define "standard PC's", I'll presume you mean desktop PC's. Now, show me where I said there "are many" desktop PC's that use 105w+ cpu's... Because I never limited the scope of my discussion to the desktop. So if you'd like to continue this (or any other) discussion, then do not twist my words around.

Now, what I did say was: "there are higher power consuming cpu's that live in even smaller cases, that don't use "exotic" cooling."... And that is very much a fact.
 
Sep 15, 2003
139
0
0
---
Now, what I did say was: "there are higher power consuming cpu's that live in even smaller cases, that don't use "exotic" cooling."... And that is very much a fact.
---

Again. Name one!

You continually call me out and harrass my posts with your gospel talk. Now I am asking you to name one without posting 20 additional messages which still dont answer anything.

If you want to play moderator at least be able to back yourself up.