3400+ Socket 939 for sale on da Egg

SparkedFire

[H]ard|Gawd
Joined
Jun 27, 2003
Messages
1,109
Socket 939 3400+ appears on da Egg. Seems to be a surprise becuause no roadmap mentioned this. It cant be an error because the laser engravings on the lid says it's a 939 part.

Link on the Egg. It goes for $288.

3400_939.jpeg
 
So... is that a 939 part with single channel memory? I don't get it, its clocked just like a 3500.
 
james__bean said:
So... is that a 939 part with single channel memory? I don't get it, its clocked just like a 3500.

90nm core, is my guess... but wouldnt that put it higher than the 3500? because of the improvments to the core?

it does seem to be 90nm tho, all the 130nm parts ive seen on newegg have it listed, and all the 90nm's dont

edit: yes it is 90nm, newegg search with AMD and 90nm comes up with this chip alone. never heard of a 90nm 3400+ clocked the same as the 3500 tho... :confused:
 
socket 939 so it is dual channel.

you know, thinking about it. maby AMD will kill the current 3500+ and replace it with the winchester core moving the newcastle 3500+ to the 3400+ rating. the 939 3400+ and 3500+ have the same clock speed. that will help eleviate the rating confusion issues.

90nm core, is my guess... but wouldnt that put it higher than the 3500? because of the improvments to the core?

it is a 130nm part. look at what i put above.

edit: newcastle is 130nm. it doesn't make sense for it to be 90nm.
 
the voltage is wrong for a 90nm part too... but.. i dunno, it just doesn't make sense.
for now, i'll assume it's a 90nm chip running at 2.2ghz
 
newegg just updated the processor specifications. here is what it now is.

Model: AMD Athlon 64 3400+
Core: Newcastle
Operating Frequency: 2.4GHz
FSB: Integrated into Chip
Cache: L1/64K+64K; L2/ 512KB
Voltage: 1.5V
Process: 0.13Micron
Socket: Socket 939
Multimedia Instruction: MMX, SSE, SSE2, 3DNOW!, 3DNOW!+
Packaging: OEM(Processor Only)

new pic:
3400_939.jpeg
 
OK... so now its clocked higher than a 3500. I realize its a different core, but there is no way a 3500 clocked at 2.2 can beat a 3400 clocked at 2.4 if they both run dual channel and both have the same cache. Remember that there are 130nm 3500's so its not like they're saying its a 3400 because its 130nm.
 
just strange. also AMD has confirmed a 3300+ Athlon 64 on the AMD Forums. Based on the Paris core with the 64bit enabled. interesting weeks ahead of us.
 
hahaha, i love it when amd just arbitarily releases new cpus without really telling anyone. it rocks
 
yeah i didnt notice the newcastle thing.. that sure is odd. the only thing i can think of is that AMD is completly phasing out 754? moving the 3400+ to 939 just doesnt make any sence
 
lithium726 said:
yeah i didnt notice the newcastle thing.. that sure is odd. the only thing i can think of is that AMD is completly phasing out 754? moving the 3400+ to 939 just doesnt make any sence

AMD will discontinue the Athlon 64 line on socket 754 in a matter of days unless you already want to say it has happened. No new Athlon 64 socket 754 chips are expected to be released nor are planned. Socket 754 will be purely Sempron based. A kick in the a** to early adopters if you ask me. at the bare minimum there thould be another 754 release like a 3900+.
 
SparkedFire said:
AMD will discontinue the Athlon 64 line on socket 754 in a matter of days unless you already want to say it has happened. No new Athlon 64 socket 754 chips are expected to be released nor are planned. Socket 754 will be purely Sempron based. A kick in the a** to early adopters if you ask me. at the bare minimum there thould be another 754 release like a 3900+.

exactly my thoughts. not contuing s754 and athlons is just like nVidia not putting an AGP slot of thier NF4 ultra boards. kick in the ass too all the early adopters of the 6800 series(me). i think its kinda dumb to not have any more CPU releases for 754 when such great boards are out for it, while s939 has a bunch of crap boards.. just dont make sence to me..

oh, and do you have a source for this?
 
lithium726 said:
exactly my thoughts. not contuing s754 and athlons is just like nVidia not putting an AGP slot of thier NF4 ultra boards. kick in the ass too all the early adopters of the 6800 series(me). i think its kinda dumb to not have any more CPU releases for 754 when such great boards are out for it, while s939 has a bunch of crap boards.. just dont make sence to me..

oh, and do you have a source for this?

i agree. although keep in mind there will be future Sempron additions to the socket 754. It isn't compleatly dead yet :p. source is from good old anandtech.

here is my source for the discontinuance of the Athlon 64 on the 754 - here.

here is my source for the 3300+ Athlon 64 - here
 
It is looking like my next big upgrade will be socket 939 more and more... :)

BTW: Hey sparkedfire! I'm from CPU too! Nice to see ya! Ben is also on these forums a little too... :D
 
bobsaget said:
BTW: Hey sparkedfire! I'm from CPU too! Nice to see ya! Ben is also on these forums a little too... :D

i figured that when i saw your name :p . I was like hmm... then i read the post. i agree s939 is the way to go now.
 
SparkedFire said:
i agree. although keep in mind there will be future Sempron additions to the socket 754. It isn't compleatly dead yet :p. source is from good old anandtech.

yeah, but who the hell would buy a DFI board with a.. a... SEMPRON!!??

if s754 is this close to its end as an enthusiast part, i wonder why DFI didnt just skip s754 instead of skipping nf3 on 939

edit: i noticed on anand, on the amd roadmap, that there was no 3400 for s939 listed at all. i think newegg has screwed up
 
lithium726 said:
yeah, but who the hell would buy a DFI board with a.. a... SEMPRON!!??

lol, your right about that.

lithium726 said:
edit: i noticed on anand, on the amd roadmap, that there was no 3400 for s939 listed at all. i think newegg has screwed up

Newegg didn't screw up because i ran the part number on the cpu lid under their product pics and it is a 939 part as i stated earlier.

look at this picture and you see that "D" next to the 3400 in the ADA3400DEP4AZ. that signifies a 939 pin chip.
 
This is great ive been waiting for this I could have bought a 754 in june I had the cash and it was in front of me @ Frys (My friends tried to get me to do it) but I was holding out for a cheap 939. Can I get a w00t from the Audience?
 
SparkedFire said:
Newegg didn't screw up because i ran the part number on the cpu lid under their product pics and it is a 939 part as i stated earlier.

look at this picture and you see that "D" next to the 3400 in the ADA3400DEP4AZ. that signifies a 939 pin chip.

wierd...


...and a w00t
 
Well, the 3800+ 939 runs @ 2.4, how weird unless they halved the cache in it for some reason.
 
I think that it's a 90nm part. Use the newegg CPU search doohicky. Set the interface to 939 and the micron to 90nm. What comes up?

The 3400+ that you linked.
 
3000+ 939 1.8Ghz 512mb newcastle
3200+ 939 2.0Ghz 512mb newcastle
3500+ 939 2.2Ghz 512mb newcastle

There has to be a typo somewhere.

3400+ 939 2.4Ghz 512mb newcastle

the 3400+ doesn't make sense especially at a lower price point than the 3500+ ... faster and cheaper? can't be right :rolleyes:
 
haroldmeyer said:
Newegg updated it and put it at 2.2ghz, still a good deal to be had!

yeah, just saw that and came to post back. it has a 1600MHz bi-directional HyperTransport bus, not the 2000Mhz bi-directional HyperTransport bus on all other socket 939 chips.
 
See, this is why the damned PR ratings suck.

different L2 cache, different HT speeds, different core speeds, same damn rating.

This is confusing and I do this for a living, every damn day.

How the hell can I explain this to my customers when I have a hard time keeping up?

Too many chips in the same spot in the same market, they're setting themselves up for a flop.

Not enough cars on the lot, no sales, doesn't look serious. Too many cars on the lot, no sales because the customer is overwhelmed.
 
0ldman said:
How the hell can I explain this to my customers when I have a hard time keeping up?

I show my customers the benchmark graphs. They couldn't care less why a product has a certain name.
 
Back
Top