P-4 puts Thunderbird to shame !!!!!!!!!!!

[PliotronX]

<< You seem to really like using Quake 3 as an example as much as Intel does! >>



Yeah, didn't you know? Quake 3 is the only game in existence

 

[DarkMajiq]

NOS, you really are an idiot. The outcome of a Q3 match between the 2 of you would have absolutely nothing to do with the systems, only to do with who is a better Q3 player.

And it's funny that you should accuse me of having a small pecker, because do you know who normall accuses people of that? People with small peckers!

And it just goes to show, once again, that you're resorting to idiotic name-calling and insults when presented with clear and logical arguments.

 

[Zephyr]

<< Unlike you I learn from my mistakes. >>


What mistakes would that be? I have had absolutely no problems with any of my AMD setups.


<< Also I just benched with the LAmerz Sisoft and the CPU bench is a Tie he wins ALU I win the FPU mine was 2081 on the FPU and 3760 ALU . Now lets talk about the rest of the Sisoft Benchmarks how about memory Bench I score 1796 ALU/ram and 1869 FPU/ram and with the Multl media Interger is 6740 and Flonting Point is 8363 >>


What good is it you can feed your cpu from the meemory if 1) that is not what apps need, and 2) your cpu is generally weak so memory isn't really the bottleneck?

 

[backWERD]

P4 dont have a great memmory subsystem I will agree with you there its just the rest of the processor that sucks. including the arm and leg you have to give to pay for it.:Q:Q:disgust::frown:

 

[backWERD]

I am now a SENIOR MEMBER

 

[MrGrim]

Hey NOS440 it's me again. Can you answer my question now? How come a TB at 1,46GHz outperforms a P4 at 1,5GHz?

You are making a fool of yourself by saying that &quot;It runs smoothly and that's enough to say that P4 owns&quot; when ppl present you facts and benchmarks that clearly show that TBs are better than P4s.

You can ignore me all you want, but if you keep on replying on this thread I'll keep on asking you this very same thing.

Have fun.

 

[frustrated2]

Mr Grim is your heating bill going down with that thing running??

 

[frustrated2]

Also I think he answered your questions a few posts prior to these.

 

[NOS440]

MrGrim Um I thought I just posted the answer to your post and I haven't heard what kind of cooling your using to run that Juice sucking furnace to get it that high and also you seem to only want to post 1 benchmark is that the only one it will run before it gives ya a BSOD get real man if were going to compare then lets really compare. Not just one benchmakr that is a Tie.

 

[Zephyr]

<< Mr Grim is your heating bill going down with that thing running?? >>


Scared of a little heat are we? That's kinda weak sisn't it? we're supposed to be enthusiasts here or did I miss something?

 

[dougjnn]

My two cents is this.

Right now Tbirds are clearly faster for the vast majority of apps. Essentially everything that wasn't written to use SSE. And TBirds are vastly better in bang for the buck. To compare here one has to add the price of 1) cpu; 2) mobo and 3) same amount of Ram. I'd use 256meg of ram as the standard, because I wouldn't run at less than that. (Actually I just got into the Crucial price reduction to up my W2k ram to 640megs.)

However, in about 9 months, the chances are the p4's will be the fastest in real world applications. They still won't have that much stuff written for them. Especially they won't have games written for them. (Rember the x-Box is using a P-III at about 733mhz, so that is what games will write for, or perhaps that and a bit faster, same general architecture. Good for the speedy TBirds.) But in 9 months there may well be 2ghz P4's out there. Palominos probably max out at about 1.7ghz. Perhaps at those two respective speeds the Tbird Palominos will still be somewhat faster. In any event, they are almost certain to be the vastly better price/performance buy.

I suspect that P4's when produced using .13 trace lines may actually hit even 3ghz. They were designed to ramp way up in speed. But we are talking 2003 here. By then Palominos will be left in the dust. Probably even on a .13 process. It is not clear that AMD will have something really competitive by then. Maybe, maybe not. I bet I switch back to Intel in 2003, using P4, while meanwhile Intel will be making noise about some 64 bit CPU at that time -- for which precious little software will yet have been optimized for.

So my upgrade path is about like this. I'm building a KT133a system w/ a Tbird 1.0@1.2 next week. (Or ordering the final parts then, after intel's price drop maybe helps a tad.) I expect to buy about a 1.5 Palomino in about 9months for about $150, which I'll run at 1.7-1.8. Then sometime in 2003 I'll maybe go to a 2.5-3.5ghz P4 .13 process, with a DDR (or 4DR) mobo -- (while intel is trying to get us to buy their 64bit job). Oh, I probably get the nv20 video card about the time I get the Palomino. About next Xmas, before or after. Yeah, the NV20 is supposed to be out by the end of March. But I won't buy it then. It won't make a huge diff until X-box games come out written for it, and easy converted to PCs. Meaning by late fall - Xmas.

My musings on this holy war.

 

[Zephyr]

<< ...before it gives ya a BSOD get real man... >>


And you base the assumption that Mr Grim's system is unstable on what?
(other than the obvious anti AMD bias you posses)

 

[MrGrim]

OMG, by the time I posted my message there were another 6 replies!

BTW it's NOT my system, I have a Celeron2 566@850 soon to be Duron 800@1GHz. This is were I got the benchmarks from.

NOS you really are an idiot. I'm saying that your claim of P4 owning a TB is BULLSHIT and I prove it with these benchmarks and you come back and say, it's propably not air cooled. If you read the thread you will see that it IS air cooled, but what difference does it make?

Oh, and why don't you show us those benchmarks? We already know that it's running smoothly ...

 

[NOS440]

Zephyr wake up dude you know as well as I that wattage used and heat produced is a major issue and has been since the day's of the Classic Athlon and a unit like that is producing tons of heat and running at at least 80 watts of usage also it is on a very high FSB which gives it a edge but also will most likely cause instability. I'm really starting to wonder about you I have always respected that you usually will not post something that is not well thought out and has meaning. That last statement is not up to your normal excellence !

 

[NOS440]

MrGrim Now the facts are straight you are a winnie hiding behind some elses OCed rig LOL!!!!!!!

 

[MrGrim]

You want more benchmarks? Why? Look at the title of your thread ... if one CPU owns another that means that it owns it in every aspect, in every single benchmark and by FAR. BTW I linked 2 benchmarks and I gave you the link to the thread I got them from. Go and ask emsixteen for more benchmarks.

 

[NOS440]

This is my point this is from the owner of the rigs post



<< Cooling - generic compusa case with fop32, cardcooler xts, the case is open and situated towards an open window, the cooling that is generated is amazing! Ambient 63F !!! >>





Laughing my A$$ off I love it when I'm right !!

 

[MrGrim]

<< MrGrim Now the facts are straight you are a winnie hiding behind some elses OCed rig LOL!!!!!!! >>



EH? And you are supposed to have kids? If I didn't know that I would have thought that you are 9 years old. You claim that P4 > TB, I know it doesn't and I prove it to you. Can you do the same? Can you do anything else other than say it runs smoothly?

 

[MrGrim]

<< Laughing my A$$ off I love it when I'm right !! >>



You are not right, it's air cooled! How thick are you? What difference would it make even if it was water cooled? The fact remains, P4 costs double what the TB does and the TB outperforms a P4 at the same clock speed.

Why is it so had to accept that? Face it, your system is not the fastest in the world. Learn to live with that, cause with the money you just spent on that you propably can't afford to get a TB.

 

[dougjnn]

NOS440

I just took a look at your Rig link. That's some pretty impressive speed you got out of your coppermine for air cooling. What HSF are you using?

Also, do you have advice on where to buy cC0 stepping intels? I want to be sure of getting cC0. I don't really want to pay for specific overclocking guarantees. They tend to price those at close to the price of the Intel specified processor.

I'm planning on upgrading a celeron 500 fpga chip to a 566@850++ fcpga (or maybe a 600@900, if that will work) for some kids I'm close to. (They are sort of my step kids - the kids of my girl friend who just recently died at a tragically young age.) I just got them a Geforce2 MX video card. I'd like to up the processor to go with it. The mobo (a DFI via Apollo job) does support 100 and 133fsb, and the fcpga voltages. I THINK but am not yet positive that it allows voltage and fsb changes in small steps from the bios as well. That seems to be something the manual (at the DFI web site) alludes to, although it sure doesn't high light it.

But the thing is, I've got to find some source where I can be sure of getting a cC0 stepped Celeron. I don't know of any local shops here in NYC which are good for that.

Any advice?

 

[NOS440]

MrGrim I did all you have to do is READ below and lets see some real Benchmark on that Unit that has to have fresh outside winter air to cool it not just Sisoft winnie Benchmark But something that will push the machine I bet it won't run them at all.

 

[MrGrim]

Sorry, what you did was state your scores. Frankly I don't believe you. I want to see some pictures of a P4 OWNING this TB@1,46GHz. If it was my system I would be more than happy to run more benchmarks but that is not the case. Just post the pictures of the scores that you mentioned.

Remember, your scores will have to put it to shame!!!!!!!!!!!

 

[NOS440]

dougjnn there are 2 guy's that specialize in them in the For Sale and Trade forum that you should track down for Cc0's they are Compuwiz 1 and Bonkers325 I was are you ready for this using a Golden Orb for thoise speeds with a closed case and 1.85 volt with a temp of 102 max. I still have that basic setup running my server but at stock clock !

 

[NOS440]

MrGrim I stated below that I would Email them to ya if you like I have no server to put them on it this time. Just sent them to ya But really please take the gun from the side of your head it isn't worth it dude !

 

[Zephyr]

<< Zephyr wake up dude you know as well as I that wattage used and heat produced is a major issue and has been since the day's of the Classic >>


Nope.. only an issue untill mobo manufacturer figured out how to desing mobos capable of handling these power requirements, and untill builders figures out that a 145W psu doesn't cut it anymore.


<< unit like that is producing tons of heat and running at at least 80 watts of usage also it is on a very high FSB which gives it a edge but also will most likely cause instability. >>


no not really... 80watts is only a problem if you have insufficient power and cooling for it.
Very high FSB.. well not really.. northbride technology keeps on improving .. the KT133A is capeable to handle these high FSB settings, so I wouldn't say it's most likely unstable.


<< I'm really starting to wonder about you I have always respected that you usually will not post something that is not well thought out and has meaning. That last statement is not up to your normal excellence ! >>


Well thank you NOs for those kind words, however I fail to see how this was not well thought through.. I stand by my statement... I don't find it surprising that the Tbird can do 1400-1500... on .18 micron .. remember the old K7 .25micron? it was believed that the loft for the .25micron process would be 650MHz, yet the Athlon hit 750 on stock speed, and oc/'ed well above 800. It's all a result of a superiour architechture, that ramps well without sacrificing clock-for-clock preformance.
The next step is SOI.. even though Intel moves to .13micron I believe SOI will keep AMD very competitative even on .18micron (so they can wait longer before shelling out $$$ for a retool) if the preliminary reports about SOI are correct.