• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Athlon 64 3400 vs. 3500

Originally posted by: Zoinks
What kind of CPU would you recommend for a new nforce build? I was thinking something in the $200-300 range to build a good gaming machine.
If you prefer to be able to upgrade your CPU to a nicer one in the future without getting a new motherboard, or if you are go to overclock, get a 90nm socket 393 CPU, like the 3500+. Otherwise, the socket 754 A64's, like the 3400+, are probably better for the money.
 
Im sure theres a 2.4 Ghx 3400+, but what I dont get is, add dual channel, and an extra 512k Cache and you have a "4000+"...


Hmm 600 Pr points for all that ?....
 
"The other point is that the Athlon 64 3400+?s P-Rating isn?t really important. Changing from 3400+ to 4000+ only improved performances by 4.9%. This result is clearly smaller than the 7.4% gain obtained in going form a 3200+ to a 3400+. The source of this difference comes from processor improvements. Increased frequency significantly changes performance while the double memory channel or the 1 MB of cache L2 aren?t always used by applications and shouldn?t strongly influence the P-Rating.

Even if these processors provide excellent results, we feel that the Athlon 64 3500+, 3700+, 3800+ and 4000+ don?t deserve their P-Rating. We think that the 4000+ sales name should be 3600+, for example."


source : http://www.behardware.com/articles/525/page10.html
 
Originally posted by: clarkey01
Im sure theres a 2.4 Ghx 3400+, but what I dont get is, add dual channel, and an extra 512k Cache and you have a "4000+"...


Hmm 600 Pr points for all that ?....

Yes that why a 3400 NC is a 3600 as the benchmarks prove.

It's simple really

A64 * 1.50 = PR
Add 5% for dual channel (aka 939)
Add 5% for 1 mb lvl 2.

So what's a 3400 NC? = 2400* 1.5 = 3600

Whats a 3400 clawhammer? = 2200 * 1.5 * 1.05 = 3465

Whats a 3500 NC? = 2200 * 1.5 * 1.05 = 3465

What's a 4000? = 2400 * 1.5 * 1.05 *1.05 = 3969

The only chips that are terribly labeled are the 3200 and 3400 754 NC's which should be 3300 and 3600 repectivly.
 
Zebo, work for AMD on the marketing side, you'd do a better job then the guy who came up of the idea of calling the 2.2 Ghz AXP a "3200+"
 
Hey thanks. i drink english newcastle exclusivly too.😛


Seriously AMD's PR ratings are a bit messed up. But they can't call a chip 3765 right? So they call it 3800.
 
Originally posted by: clarkey01
Zebo, work for AMD on the marketing side, you'd do a better job then the guy who came up of the idea of calling the 2.2 Ghz AXP a "3200+"

YA I call it a 2800

AXP * 1.25 = PR
Add 2% for 512

2200 * 1.25 *1.02 = 2805 and the benchmarks prove it. Compare to a 2.8 P4C. Course thier "logic" was it was meant to compare with orginal tbird (ya right) but it just made them look stupid when a 3200 mhz P4 was whipping the floor with them.



 
Originally posted by: bigal40
Originally posted by: Zebo
???

http://www.behardware.com/articles/531/page7.html

130 beats 125.

only one bench 3500's win, UT2004 all rest 3400 wins.

I don't know if we are looking at the same article or not Zebo. THe only benches the 3400 wins in is the Pacific Fighters and the Overall test you mentioned above.

you maybe arn't reading benches correctly. Tell me exactly which bench you see 3500 winning over 3400 and I'll try and explain.
 
Originally posted by: Zebo
Originally posted by: bigal40
Originally posted by: Zebo
???

http://www.behardware.com/articles/531/page7.html

130 beats 125.

only one bench 3500's win, UT2004 all rest 3400 wins.

I don't know if we are looking at the same article or not Zebo. THe only benches the 3400 wins in is the Pacific Fighters and the Overall test you mentioned above.

you maybe arn't reading benches correctly. Tell me exactly which bench you see 3500 winning over 3400 and I'll try and explain.

Sorry about thatZebo i went back and took a closer look and it turns out i didn't read what it was benching. It was showing time in seconds. I was just looking for the higher numberwhen i looked through them relly quick. My bad.
 
Originally posted by: drpootums
this is great news for me! I just ordered the DFI nF3 board and the 3400+ newcastle.

Cant wait!

Ya that is great, rumor is the newer 3400 uses same SOI as FX-55. 2700-2900Mhz overclocker.
 
Originally posted by: drpootums
this is great news for me! I just ordered the DFI nF3 board and the 3400+ newcastle.

Cant wait!




Ya that is great, rumor is the newer 3400 uses same SOI as FX-55. 2700-2900Mhz overclocker.


Anyword on what online retailers have these in stock. My current 3400+ overclocks poorly.
 
Although at stock speeds the 3400 beats the 3500 what about when you factor in oc'ing.

I'm trying to decide between a 3500 939 (130nm) and the 3400. I read through the articles comparing the two processors but I am wondering how they will compare with overclocking. Is the 3400 still going to hold an edge over the 3500 under a standard air cooled setup. I mean is there any reason to go with a 3500 other than the benefit of being able to upgrade later on since you can use a 939 socket mb. Most people seem so high on the 3500/939 combo and after reading the benchmark comparisons I am wondering why.
 
Back
Top