XP2100 to a XP2700? Worthwhile in your opinion?

[gizbug]

Wondering if upgrading my cpu to this is worthwhile?
Athlon XP 2700+ (2.167GHz)
$256 is the current price.


Any comments, or anyone who has recently upgraded?

 

[wampa]

If you have the money why not

But I guess if you are doing anything CPU intensive it would benefit from the increase. Personaly I would just get a cpu and OC it to acheive that speed.

 

[Link]

 

[Link]

Why not get a 2100+ Tbred and overclock it to 2700+ ?

 

[Mangler]

Yup..what the guy above me said!

 

[gizbug]

Seems my mb only supports 2600xp.

Followup question:
Does anyone know a place with benchmarks? I am curious what the performance increase would be from my 2100xp now...going intel 2.33ghz. That would cause me to buy a new mb, not sure if my pc2100 ram will work or not.

Anyone gone this route, or offer any expertise on this upgrade?

 

[MrEgo]

If you flashed your motherboard BIOS, it should be able to support the latest AMD processors if you say it already supports the 2600+. Anandtech is full of benchmarks for CPUs, it should be very easy to find benchmarks for any CPU you've ever heard of. However, I've gone through the upgrade from 2100+ to 2700+ first hand. I use the computer mostly for games, and this upgrade was extremely useful for it.. but if you get a 2600+ or 2700+, get the 333 MHz FSB version. I made my own benchmarks for 3D Mark 2001 SE using different CPU speeds.

I have a Radeon 9700 PRO on an Epox 8RDA+ motherboard, with 512MB PC3200 Corsair DDR.

With the 2100+ at stock speeds, I got a score of ~11,500.
WIth the 2100+ at 333 MHz FSB, I got a score of ~12,100.
With the 2100+ overclocked to 2700+ (2.167 GHz w/ 333MHz FSB) I got a score of ~13,600.
With the 2100+ overclocked to 2.2 GHz w/ 200MHz FSB I got a score of ~14,000.
With the 2100+ overclocked to 2.3 GHz w/ 200MHz FSB I got a score of ~15,300.
With the 2100+ overclocked to 2.3 GHz w/ 200MHz FSB and the Radeon 9700 Pro overclocked to 350 MHz / 333 MHz, I got a score of 16,230. (had to throw this in)

So if you're familiar with 3D mark, this is probably the same pattern you'd see if you upgraded. I only made this upgrade because I sold my Dual Athlon MP rig for $375, and I actually nearly saved money by buying the 2100+ Tbred B for $93, the 8RDA+ Mobo for $105, the SLK-800 HS + Fan for $40, and the memory for $160... otherwise I probably would have waited another half a year or so until Athlon 64 comes out to lower prices of everything else.

 

[MTDEW]

MrEgo's numbers are right on the money!

I have a Radeon 9700 PRO on an Epox 8RDA+ motherboard, with 512MB PC3200 Corsair XMS DDR.
Im using a 2400xp Tbred b, fsb & DDR in sync, Dual channel.

And my scores at different speeds/fsb mirror MrEgo's almost EXACTLY!

You may notice Im running watercooled, but I got same results on air with just slightly higher temps.

2100+ Tbred B for $93, the 8RDA+ Mobo for $105, the SLK-800 HS + Fan for $40, and the memory for $160

Seems like a darn good suggestion to me!

At these prices theres no excuse for have a slow rig anymore!

EDIT: MrEgo I just updated to the 3305 bios and saw a nice jump in 3dmarks scores.
2.3ghz / 11.5 x 200fsb / Dual Channel / DDR 200mhz in SYNC / 6 2 2 2 NOTNING else OC'ed , 9700 at 324/310
Scoring 16,080 - 16,150 now.. GiVE THAT BIOS A TRY!

 

[bgeh]

Originally posted by: Link
Why not get a 2100+ Tbred and overclock it to 2700+ ?

yup, the XP 2100+ T-Bred B's seem to oc really well

 

[bgeh]

oops, double post

 

[MangoX]

With the 2100+ overclocked to 2.2 GHz w/ 200MHz FSB I got a score of ~14,000.
With the 2100+ overclocked to 2.3 GHz w/ 200MHz FSB I got a score of ~15,300

I doubt that increasing the cpu clock by just 100mhz will give you an increase of 1300 marks. If the 2.2ghz was at 166fsb and the 2.3 at 200fsb you could probably gain that much marks.
My guess is that you didn't actually run every single one of those tests and that it's just an estimate from your experiences.

Chan-Chan

 

[ChampionAtTufshop]

no its not imo
wait for next gen before an upgrade imo

 

[Shimmishim]

bah!

i think benchmarks are crap... when you play a game, do you notice the difference between 200 fps and 230 fps? honestly, are your eyes that sensitive?

 

[ChampionAtTufshop]

even if they were sensitive, i think its impossible to differentiate between anything over like 100fps or so for the human eye

maybe it was around 30-40fps,....

in any case, you probably wont notice any diff b/w them

but if you wanted to upgrade, id probably go for xp2500+ (barton) and oc that

ohterwise, like i said before, wait for athlon 64 (i think its still called that)

 

[bgeh]

Originally posted by: ChampionAtTufshop
even if they were sensitive, i think its impossible to differentiate between anything over like 100fps or so for the human eye

maybe it was around 30-40fps,....

in any case, you probably wont notice any diff b/w them

but if you wanted to upgrade, id probably go for xp2500+ (barton) and oc that

ohterwise, like i said before, wait for athlon 64 (i think its still called that)

i agree. most monitors can't even display over 100-130Hz

 

[MTDEW]

I would disagree a bit that benchmarks are crap.

They give you an 'idea" of the performance of your setup and come in real handy when tweaking setups, to see if changes
you make make a difference or not.

With my P4 rig (2.4 @ 3.0ghz / Epox 4BDA+ mobo) I got roughly roughly 12,500 3dmarks in 3dmark 2001.
(it was a 400/p4 btw, not a 533)

With my 2400xp @ 2.3ghz / Epox 8RDA+ combo I get low to mid 16,000 3d marks.

Do I care about that score?
ABsolutely not, I dont play 3dmark 2001.

But I do care about Papyrus's Nascar 2003 performance.

With my p4 setup I ran at 1280x960x32bit color with 4x AA & 4x Aniso & averaged 30fps with the game options cranked.

Now with my Nforce2 setup, I run the game at 1600x1200x32bit color with 4x AA & 8x Aniso & average 30fps with game options maxed.

So yes, looking at what 3dmarks others were getting with Nforce2 setups gave me a pretty good idea what kinda performance increase i would get upgrading my p4 setup to an Nforce2 setup.

So I dont believe anyone really cares about an actual score or the actual fps in benchmarks, cuz no-one plays games at 200+ fps as stated earlier.

Benchmarks are just a way of judging performance to help make intelligent tweaks and upgrades to run the games you like to play at
the visual quality settings you feel are important to YOU.

Or to help avoid upgrades that may not give much of a performance boost 4 that matter also.

Without benchmarks....
1) It would be really hard to judge what hardware to buy as an upgrade, cuz you wouldnt know what performance to expect.
2) What the best price/performance "bang for buck" hardware is for people looking for that kinda upgrade.
3) You wouldnt know whether those bios and OS tweaks were helping or hurting your rig.

So I agree "in a way" , that benchmarks are worthless if your looking at a score and using it as "bragging rights" so to speak.

But if you use benchmarks as an intelligent way to chose hardware upgrade paths and choosing the right settings (tweaks) to get maximum performance from your hardware, then to me you're understanding exactly what benchmarking is all about.