1.4 tbird or 1.2 MP?

[Syborg1211]

Which one will perform better? Overclock better? Price isn't the biggest issue here since I don't really want to skimp on the processor, but I'll be placing either one of these beasts of chips into an EPoX EP-8K7A with 256 mb ddr, maybe 384.

 

[johneetrash]

what are you gonna be using it for?

if money is no issue id get 1.2mp just so u can say u have dual processors.. then people you know iwll be like "oh man wut a computer wiz he got 2 cpus in his comp"

 

[MrWhiteUK]

Let me see the 1.4 is cheaper and faster, that's a toughy!

 

[Syborg1211]

Im definitely not going dual cpu for a while, immature chipsets and a little too much money. The 1.2MP can be overclocked to what? and the 1.4 can be overclock to what? I don't really care about stock speeds, just what I can overclock to.

 

[DARRIN]

I'd like to see some benches comparing the 1.2MP to the 1.2 and 1.4 thunderbird on the same mobo.

 

[maxhdwdotcom]

Well the MP is 15-20% faster clock for clock. On the side of overclocking, the MP will probably goto 1.4/1.45GHz happily, whereas the TBird should be capable of 1.65GHz tops. Air cooling.

 

[maxhdwdotcom]

Oh, and a side note. The Epox is a great overclocker

 

[fitzhue]

The MP is 15-20% faster clock for clock? Wow I didn't expect that much of an improvement. Do you have any links to bechs or comparisons? What platform was it running on? At max overclocked speed they should probably be equal in terms of performance, if in fact the MP is 15% faster than the Athlon.

 

[formulav8]

<< Well the MP is 15-20% faster clock for clock. On the side of overclocking, the MP will probably goto 1.4/1.45GHz happily, whereas the TBird should be capable of 1.65GHz tops. Air cooling. >>



The Palomino is only about 1% to 10% faster clock for clock. Maybe a couple benches about 16% faster. Average about 5%.

 

[maxhdwdotcom]

Platform doesnt really matter. Im talking of floating point and calculations. So in real world, even though its about 20% 'quicker' you get around 15% or so. I've tested one on a KT133A and DDR and I was impressed, very impressed. It appears also that AMD are playing down the performance somewhat. I will try and find some benchmarks, I'll be doing a review on my site as soon as possible.

 

[maxhdwdotcom]

<<

The Palomino is only about 1% to 10% faster clock for clock. Maybe a couple benches about 16% faster. Average about 5%. >>



.... well I'm going on tests I've done .... also, I know another guy whos getting about the same sort of results. ALU performance is superb.

 

[maxhdwdotcom]

Heres the benchmarks my associate collected

-----------------------------------------------------------------------
206.2 - Tbird @ 1400
219.1 - Palamino @ 1400

Super PI benchmark 2M test (using crappy WinMe which is really slow for this)

3:52 - Tbird 1400
3:17 - Palamino @ 1400

The Palamino is around 15% faster for Flask under the same conditions


Sisoft Sandra Memory bench

654/861 - Tbird @ 1400
779/865 - Palamino @ 1400 ( A massive increase in the Alu scores)

OCUK SETI Benchmark

4 hours 28 minutes - TBird @ 1400
3 hours 49 minutes - Palamino @ 1400

Full load temps

48.5C - Thunderbird
41.1C - Palamino (honest , this thing is cool)
-----------------------------------------------------------------------

Seems alright to me.

 

[Insane3D]

The 1.2ghz MP should hit 1.4ghz @ default voltage, maybe 1.45ghz with some higher voltage. The 1.4ghz will probably hit around 1.5ghz - 1.6ghz. The MP will have better memory performance due to the hardware prefetch. Also, the MP runs cooler due to the 10% larger surface area. In some things a 1.2 MP @ 1.4 will outperform a Tbird @ 1.4ghz...like Seti. I knocked around a hour off my Seti@Home work unit times with my MP @ 1.4ghz vs my Tbird @ 1.4ghz. If you are only worried about overclocking, the 1.4ghz bird would probably be your best choice. Oh, and if you get the MP, don't tell MrWhiteUK, because according to him, the MP is &quot;a waste of money&quot; and you would just be getting it for &quot;boastage&quot;...

 

[maxhdwdotcom]

There are certain benchmarks that'll give you a score of 20% higher no problem, then are some that will show a negligable or couple of percent increase. Trust me though, its more then 5% on average.

 

[Syborg1211]

Did I mention I have an Alpha PEP66 with Delta 38cfm fan that's gonna cool either one of these chips? Id like to run it with a crappier, but quieter fan but if the speed exchange is high enough ill use the delta.

 

[maxhdwdotcom]

<< Did I mention I have an Alpha PEP66 with Delta 38cfm fan that's gonna cool either one of these chips? Id like to run it with a crappier, but quieter fan but if the speed exchange is high enough ill use the delta. >>



The Alpha should be fine, the MP runs alot cooler

 

[gunf1ghter]

The Palomino is not 15-20% faster in ANY applications. AMD's own benchies show that it is between 3-8% faster in applications.

The amount of mis-information and FUD on this forum is amazing.

 

[Insane3D]

I also noticed the boosted ALU performance. My MP @ 1.4ghz, and a 155 FSB scores a 872/974..that ALU boost is pretty obvious in that score. I actually get Seti work unit times from 3hr, 30 min down to almost 3hrs. flat. Peretty impressive to me...

 

[Insane3D]

&quot;The Palomino is not 15-20% faster in ANY applications. AMD's own benchies show that it is between 3-8% faster in applications.

The amount of mis-information and FUD on this forum is amazing. &quot;

Ok smart guy, what would you attribute the difference in the Seti work unit times? With a tbird @ 1.4ghz and a 165FSB, I get a work unit done in around 4hrs, 15minutes. The Palo @ 1.4ghz with a 155FSB get's one done in around 3hrs, 15minutes. That's 1 hour less...guess that's just FUD right. You seem to have all the answers...are you using one?

 

[maxhdwdotcom]

<< The Palomino is not 15-20% faster in ANY applications. AMD's own benchies show that it is between 3-8% faster in applications.

The amount of mis-information and FUD on this forum is amazing. >>



I'll say again. ME, MYSELF and I achieve around 20% faster results at BEST, NOT everytime. Jeeez.

 

[ST4RCUTTER]

Im definitely not going dual cpu for a while, immature chipsets and a little too much money.

I'm with ya on the price, but I'd say the 760MP is probably as stable a chipset as you will find. AMD does make good chipset, like Intel, they just don't have the fab capacity yet.

I think VIA gets a bad wrap more than they deserve, but the inconsistancies do make one wonder about their validation process. They should at least have tested their southbridge with the most popular PCI-based cards...cough...cough...Soundblaster.

 

[maxhdwdotcom]

Well remember you don't have to run 2 chips on the AMD chipset, so its a feasable, slightly pricey, platform.

 

[gunf1ghter]

You don't need to get hostile about it. Look, you are running two seperate CPU's at totally different bus speeds. What you need to do is run the procs at the same bus speed... for example a 1.2 Tbird vs a 1.2 Pal.

I admit that I am not certain as to the Pal's performance in applications such as RC... however, AMD did a good job of providing us specific information about how the Pal would perform in games, office apps, etc. These are the apps most users care about. I have seen numerous posts of &quot;the Pal is 20% faster&quot;... that type of statement is about as misleading as &quot;you can hit 1600 mhz with that processor no problem&quot;.

I find such statements meaningless and of little value.

 

[Quickfingerz]

If you can push voltage high enough, there is no reason that a 1.2 mp should not hit 1.5 ghz... I'd get the 1.2 MP... especially if you're using DDR memory.

 

[maxhdwdotcom]

Both CPU's at a push 'should' be able to hit 1.6GHz. Both my AXIA 1GHz chips hit 1.6GHz on a VIA KT133A chipset. I'm not saying every chip is the same, some have lower thresholds.