classic vs. tbird?

[CTho9305]

currently I have a slotA tbird 700@700, and was wondering how fast of a classic i'd need to match the performance in my applications: mostly 3d studio max and unreal tournament (halflife is smooth on an athlon 500 in my experience, so i'm not worried there ).

thanks

 

[Damascus]

Better safe than sorry, just get an XP2000+.

 

[microAmp]

If I remember right seeing benchmarks with the Classic Athlons vs. Duron vs. TBird vs. XP I think the Duron acutally performed slightly better than the Classic at the same clock speed. If you get the 1 GHz Duron you will be sitting a whole lot better due to more instructions on the 1 GH+ Durons than sub 1 GHz Durons. But then again they're alway the XP.

If I could even remember how I found that site with the Old vs New Athlons I would link it.

 

[AkumaBao]

Yeah, the Duron's definatly surpass the old slot A's. Not as many floating points. I'd get a socket A mobo, and a T-bird. The best Slot A is about a expensive as a XP 1500+.

 

[CTho9305]

sheesh... i ask a specific question and get totally unrelated answers.
I have:
a SLOT tbird 700. it is SLOT A. NOT SOCKET. YES, there is such a thing as a SLOT A thunderbird.
a 550mhz athlon classic that posts at 900mhz and in almost stable at 800mhz iwth the crap heatsink on there... i think with my alpha i can get it up to good speeds.

SOOOOO... i'm asking how fast a classic has to be to match/beat a tbird. if I need only 750classic to beat the 700tbird (at unreal tournament and 3dstudio max), then i'll move to using the classic as my primary CPU. if I need >800mhz, then i'll probably stick with my tbird. I just want to know before I go fight with the classic to get it to whatever speed i might need.

 

[MetroRider]

hehe, thanks for straightening it out and being more specific

 

[MetroRider]

hehe, thanks for straightening it out and being more specific

 

[DDad]

A stupid question- What chipset are you running? If it's a AMD, you should be Ok, as I recal, you may be SOL if it's VIA

 

[TheCorm]

Yeah, I reckon the 750mhz Athlon classic would be on par with Your 700mhz Athlon tbird.

Did they release Slot A Tbirds in Retail or just OEM?, I can't remember.

Corm

p.s: Can you not overclock your slot a tbird?

 

[CTho9305]

DDad - I'm SOL... via KX133. so I can get my tbird to 750 at sock voltage, but 800 is unstable up to 1.85 vcore .
TheCorm - OEM only. I can over clock it... it posts at 900-950mhz but it doesn't get into windows/linux.

http://ctho9305.res.cmu.edu/AthlonPelt.avi <-- shows what I'm planning to do to the poor 550 as soon as I get 4 screws that are the right size . yeah, i know it will be VERY VERY VERY hot on the hot side, but I worked it out and I should still get near freezing on the cold side

 

[DavemanUT]

I'm speaking from experience here. Do this. Take the MB/CPU to eBay. They are still worth a little bit now. Then, go buy a kt133a board and a duron 1.0 (I got mine for $53 shipped from tcwo). My combo is SOO much faster now, and it won't cost you that much.

D.C.

 

[arbntmare]

i think i fried my motherboard.. how much is a new one cost you DavemanUT

 

[L8CS]

<<i'm asking how fast a classic has to be to match/beat a tbird. if I need only 750classic to beat the 700tbird (at unreal tournament and 3dstudio max), then i'll move to using the classic as my primary CPU. >>

According to an old EETimes.com article dated 01/20/2000, AMD's Chairman at the time W.J. Sanders said the performance difference between the Classic and T-Bird should be approximately 10% faster for the T-Bird at the same clock speed. "Intel gained about 10 percent increase in performance, and I would expect AMD to do about the same." I have been unable to find any archived benchmarks to back it up.

I'm still running an OEM version of the Slot A Athlon 1gb Cpu, I bought off a guy last year. The K75-1 gb sucks 65 watts of power at full load, even with the smaller die size. It's not a T-Bird but it is the smaller .18 micron die, which was introduced with the Classic 750. The K7-700 is the .25 die size and the K75 is the .18 die size if I remember correctly. The first T-Birds in the SEC and PGA configurations, were the 650mhz versions.

I guess to answer your question, you would need the K75-800 Classic to beat your 700 T-Bird. But I seem to remember that they cut the cache divider speed on the 750 and higher to 1/3 instead of the 1/2 speed in the 700 Classic which may result in slower performance.

 

[CTho9305]

i guess i'll run some benchmarks myself... the classic has a 650 core and 3.6ns cache... so we'll see. thx.

 

[L8CS]

<i guess i'll run some benchmarks myself...>

I'll keep digging through the old websites to find your info. It sad that alot of the old BB's like BX Boards and Cole 3D are gone now because they always had good info.

 

[Oreo]

In 3d Studio all CPU based on the Athlon core (Slot A, T-bird, Duron, XP) are essentially the same since they all have the same FPU.

 

[L8CS]

<i guess i'll run some benchmarks myself...>

Here's some more figures. I couldn't find any CPUMark Benchmarks for the 700's.


http://www.tech-report.com/

ZDNet Business Winstone99

Athlon 700 Classic Slot A 29.1
Athlon 700 T-Bird Socket A 38.70 (32.9% faster than Classic)

Athlon 750 Classic Slot A 29.2
Athlon 750 T-Bird Socket A 39.60 (35.6% faster than Classic)

Athlon 800 Classic Slot A 29.9
Ahtlon 800 T-Bird Socket A 40.40 (35.1% faster than Classic)



http://www.anandtech.com/showdoc.html?i=1252&p=13


SYSMark 2000 - Win98se

Athlon 700 Class. KX133 chipset 143
Athlon 800 Class. KX133 chipset 153 (6.9% faster than 700 Classic)
Athlon 800 T-Bird KT133 chipset 165 (7.8% faster than 800 Classic)


Quake3 Arena TimeDemo001 640x480x16

Athlon 700 Class. KX133 chipset 118.5
Athlon 800 Class. KX133 chipset 125.9 (6.2% faster than 700 Classic)
Athlon 800 T-Bird KT133 chipset 128.2 (1.8% faster than 800 Classic)


RC5 Long Benchmark - Windows 2000

Athlon 700 Class. KX133 chipset 2.18
Athlon 800 Class. KX133 chipset 2.49 (14.2% faster than 700 Classic)
Athlon 800 T-Bird KT133 chipset 2.49 (no difference)

ALS states about the RC5 Benchmark:
"L2 cache speed clearly does not matter in this purely CPU intensive benchmark."

 

[L8CS]

<<i guess i'll run some benchmarks myself>>


Here's some more info on HardlOCP.


The Classic 750 Athlon on KX133 (K7V) vs. Socket-A TBird 750 on KT133

HardIOCP

 

[Nutzo]

You can't use the Socket thunderbird vs Slot a Classic numbers since the Socket chip set was faster than the older slot chipset. It also depends on what you are running. Some benchmarks like the larger cache, others the faster cache.

As for the Classic vs Slot A thunderbird, The Classic had a 512kb cache vs the Thunderbirds 265kb cache.

At 700 Mhz, not much difference since the 700 classic's cache was at 1/2 speed & larger.

At 750-850 then was about a 10-15% inprovement since the Classic was at 2/5 speed cache

At 900-1 Ghz there was a 15-20% improvement due to the 1/3 speed cache on the classic.

The benchmarks on the 900 classic where not much better than the 850 classic due to the slower cache.

 

[L8CS]

<<You can't use the Socket thunderbird vs Slot a Classic numbers since the Socket chip set was faster than the older slot chipset>>

The pickings are getting slim out there for trying to find benchmarks for older Cpu's. I realize there's a speed difference between chipsets but I don't believe it would be that noticable for CTho9305's purposes. The Hard OCP article appeared to be the best one I could find to answer his questions.

Here's another:
AMD's Thunderbird