Sign in with
Sign up | Sign in
Your question

2nd Gen or 4th gen CPU?

Last response: in CPUs
Share
July 4, 2013 8:38:19 PM

Hi - I am looking to get 3930K processor because it supports 64gb of ram. But I am now wondering if its mistake not to get the 4th generation processors? Is there a huge difference in performance between 2nd and 4th generation processors from intel?

More about : 2nd gen 4th gen cpu

a b à CPUs
July 4, 2013 8:41:19 PM

What applications do you run that require that much memory? What will this computer be doing?
m
0
l
a b à CPUs
July 4, 2013 8:43:34 PM

It is not the difference between second and fourth gen, it is the difference between Extreme-Edition and a regular CPU. A 3930k has six cores, and is a much better performer than the 4770k. However, in 99.9% of programs, you won't be able to tell the difference from it to a 4770k. Also, 64GB of RAM is most likely very excessive for your needs.
m
0
l
Related resources
July 5, 2013 6:01:05 AM

dannyboy2233 said:
It is not the difference between second and fourth gen, it is the difference between Extreme-Edition and a regular CPU. A 3930k has six cores, and is a much better performer than the 4770k. However, in 99.9% of programs, you won't be able to tell the difference from it to a 4770k. Also, 64GB of RAM is most likely very excessive for your needs.


I need 64 gigs to run an Oracle Fusion application. Two VMs taking up at a minimum of 24 to 28 gb RAM each.
m
0
l
a b à CPUs
July 5, 2013 6:13:15 AM

Intel is releasing their 6-core variant of Ivy Bridge this fall. If you need a Intel 6 core, go for that.
m
0
l
July 5, 2013 6:19:29 AM

He is probably not willing to wait till fall for that. It seems the OP will be better with the 3930k.
m
0
l
a b à CPUs
July 5, 2013 7:36:33 AM

its doubtful the performance difference will be "huge"
m
0
l
a b à CPUs
July 5, 2013 10:02:44 AM

dirtyferret said:
its doubtful the performance difference will be "huge"


Agreed. And plus, so many of these applications are not properly threaded for hexa-core as opposed to quad-core...
m
0
l
a b à CPUs
July 5, 2013 12:20:30 PM

The only reason I can think of for waiting for Ivy Bridge-E is PCIe 3.0 support. And that's only going to matter if you're using many powerful GPUs/other PCIe cards.
m
0
l
July 6, 2013 8:19:45 AM

ihog said:
The only reason I can think of for waiting for Ivy Bridge-E is PCIe 3.0 support. And that's only going to matter if you're using many powerful GPUs/other PCIe cards.


3930k has PCIE 3.0
m
0
l
a b à CPUs
July 6, 2013 10:52:54 AM

Tests have already been ran on the 4930K 6 core Ivy-E chip and its 5-7% faster then the 3930K
m
0
l
a b à CPUs
July 6, 2013 11:22:45 AM

Intel God said:
Tests have already been ran on the 4930K 6 core Ivy-E chip and its 5-7% faster then the 3930K


Can you please link us to these tests? Not denying anything (really, I'm not), I'm just curious. :) 
m
0
l
a b à CPUs
July 6, 2013 11:25:27 AM

dannyboy2233 said:
Intel God said:
Tests have already been ran on the 4930K 6 core Ivy-E chip and its 5-7% faster then the 3930K


Can you please link us to these tests? Not denying anything (really, I'm not), I'm just curious. :) 


No problem

http://www.pcgameshardware.de/Ivy-Bridge-Codename-23881...
m
0
l
a b à CPUs
July 6, 2013 11:28:04 AM

Intel God said:
dannyboy2233 said:
Intel God said:
Tests have already been ran on the 4930K 6 core Ivy-E chip and its 5-7% faster then the 3930K


Can you please link us to these tests? Not denying anything (really, I'm not), I'm just curious. :) 


No problem

http://www.pcgameshardware.de/Ivy-Bridge-Codename-23881...


It seems to me as though this only discusses the 4960X, the successor to the 3970X... Am I mistaken?
Thanks for the link :) 
m
0
l
a b à CPUs
July 6, 2013 11:29:56 AM

dannyboy2233 said:
Intel God said:
dannyboy2233 said:
Intel God said:
Tests have already been ran on the 4930K 6 core Ivy-E chip and its 5-7% faster then the 3930K


Can you please link us to these tests? Not denying anything (really, I'm not), I'm just curious. :) 


No problem

http://www.pcgameshardware.de/Ivy-Bridge-Codename-23881...


It seems to me as though this only discusses the 4960X, the successor to the 3970X... Am I mistaken?
Thanks for the link :) 


I could have sworn it was the 3930/4930K he tested but i was mistaken when i googled the article. Expect the same type of gains with the lower chips as well
m
0
l
a b à CPUs
July 6, 2013 11:43:46 AM

Intel God said:
dannyboy2233 said:
Intel God said:
dannyboy2233 said:
Intel God said:
Tests have already been ran on the 4930K 6 core Ivy-E chip and its 5-7% faster then the 3930K


Can you please link us to these tests? Not denying anything (really, I'm not), I'm just curious. :) 


No problem

http://www.pcgameshardware.de/Ivy-Bridge-Codename-23881...


It seems to me as though this only discusses the 4960X, the successor to the 3970X... Am I mistaken?
Thanks for the link :) 


I could have sworn it was the 3930/4930K he tested but i was mistaken when i googled the article. Expect the same type of gains with the lower chips as well


That may be true; however, this is the world of computers, and assumptions do tend to be wrong...
m
0
l
a b à CPUs
July 6, 2013 12:02:14 PM

dannyboy2233 said:
Intel God said:
dannyboy2233 said:
Intel God said:
dannyboy2233 said:
Intel God said:
Tests have already been ran on the 4930K 6 core Ivy-E chip and its 5-7% faster then the 3930K


Can you please link us to these tests? Not denying anything (really, I'm not), I'm just curious. :) 


No problem

http://www.pcgameshardware.de/Ivy-Bridge-Codename-23881...


It seems to me as though this only discusses the 4960X, the successor to the 3970X... Am I mistaken?
Thanks for the link :) 


I could have sworn it was the 3930/4930K he tested but i was mistaken when i googled the article. Expect the same type of gains with the lower chips as well


That may be true; however, this is the world of computers, and assumptions do tend to be wrong...


Whatever gains you got from a 2600K to a 3770K expect the same with SB-E to IB-E
m
0
l
a b à CPUs
July 6, 2013 1:36:50 PM

csf60 said:
ihog said:
The only reason I can think of for waiting for Ivy Bridge-E is PCIe 3.0 support. And that's only going to matter if you're using many powerful GPUs/other PCIe cards.


3930k has PCIE 3.0


Not according to Intel's site.
m
0
l
a b à CPUs
July 6, 2013 1:39:11 PM

ihog said:
csf60 said:
ihog said:
The only reason I can think of for waiting for Ivy Bridge-E is PCIe 3.0 support. And that's only going to matter if you're using many powerful GPUs/other PCIe cards.


3930k has PCIE 3.0


Not according to Intel's site.


SB-E has pcie 3.0 for Amd and you need to enable a hack for it to work with Nvidia
m
0
l
a b à CPUs
July 6, 2013 2:50:19 PM

Intel God said:
dannyboy2233 said:
Intel God said:
dannyboy2233 said:
Intel God said:
dannyboy2233 said:
Intel God said:
Tests have already been ran on the 4930K 6 core Ivy-E chip and its 5-7% faster then the 3930K


Can you please link us to these tests? Not denying anything (really, I'm not), I'm just curious. :) 


No problem

http://www.pcgameshardware.de/Ivy-Bridge-Codename-23881...


It seems to me as though this only discusses the 4960X, the successor to the 3970X... Am I mistaken?
Thanks for the link :) 


I could have sworn it was the 3930/4930K he tested but i was mistaken when i googled the article. Expect the same type of gains with the lower chips as well


That may be true; however, this is the world of computers, and assumptions do tend to be wrong...


Whatever gains you got from a 2600K to a 3770K expect the same with SB-E to IB-E


That isn't necessarily true at all.
m
0
l
a b à CPUs
July 6, 2013 3:07:08 PM

dannyboy2233 said:
Intel God said:
dannyboy2233 said:
Intel God said:
dannyboy2233 said:
Intel God said:
dannyboy2233 said:
Intel God said:
Tests have already been ran on the 4930K 6 core Ivy-E chip and its 5-7% faster then the 3930K


Can you please link us to these tests? Not denying anything (really, I'm not), I'm just curious. :) 


No problem

http://www.pcgameshardware.de/Ivy-Bridge-Codename-23881...


It seems to me as though this only discusses the 4960X, the successor to the 3970X... Am I mistaken?
Thanks for the link :) 


I could have sworn it was the 3930/4930K he tested but i was mistaken when i googled the article. Expect the same type of gains with the lower chips as well


That may be true; however, this is the world of computers, and assumptions do tend to be wrong...


Whatever gains you got from a 2600K to a 3770K expect the same with SB-E to IB-E


That isn't necessarily true at all.


Ok why not? What do you expect to happen? Thats exactly what is going to happen and the numbers for the 3970X to 4970X show that
m
0
l
a c 535 à CPUs
July 6, 2013 3:14:37 PM

Tests done with pre-production don't mean a lot, and no one has any idea what the final product will be, generally though, CPUs that are tested and written about are hand picked to to be among the best available - they don't want to send out a dweeb CPU that doesn't run for cr**
m
0
l
a b à CPUs
July 6, 2013 3:16:25 PM

Tradesman1 said:
Tests done with pre-production don't mean a lot, and no one has any idea what the final product will be, generally though, CPUs that are tested and written about are hand picked to to be among the best available - they don't want to send out a dweeb CPU that doesn't run for cr**


But expecting more then what the 3770K over the 2600K is foolish. This is the same type of progression.
m
0
l
a b à CPUs
July 6, 2013 3:20:04 PM

Intel God said:
dannyboy2233 said:
Intel God said:
dannyboy2233 said:
Intel God said:
dannyboy2233 said:
Intel God said:
dannyboy2233 said:
Intel God said:
Tests have already been ran on the 4930K 6 core Ivy-E chip and its 5-7% faster then the 3930K


Can you please link us to these tests? Not denying anything (really, I'm not), I'm just curious. :) 


No problem

http://www.pcgameshardware.de/Ivy-Bridge-Codename-23881...


It seems to me as though this only discusses the 4960X, the successor to the 3970X... Am I mistaken?
Thanks for the link :) 


I could have sworn it was the 3930/4930K he tested but i was mistaken when i googled the article. Expect the same type of gains with the lower chips as well


That may be true; however, this is the world of computers, and assumptions do tend to be wrong...


Whatever gains you got from a 2600K to a 3770K expect the same with SB-E to IB-E


That isn't necessarily true at all.


Ok why not? What do you expect to happen? Thats exactly what is going to happen and the numbers for the 3970X to 4970X show that


Dude... You're not magical. I get that you are trying to sense patterns, but those patterns don't always hold true. Just accept that you don't know what's going to happen with the 4930k, that you do know what's going to happen (maybe) with the 4960x, and that those two may or may not have any correlation whatsoever. Okay? Thank you.
m
0
l
a b à CPUs
July 6, 2013 3:21:01 PM

Tradesman1 said:
Tests done with pre-production don't mean a lot, and no one has any idea what the final product will be, generally though, CPUs that are tested and written about are hand picked to to be among the best available - they don't want to send out a dweeb CPU that doesn't run for cr**


I completely agree. These things can't really be truly seen until much closer to the release of the CPU itself, sometimes not until the CPU is released and people can test chips that are not binned for sending to companies.
m
0
l
a c 535 à CPUs
July 6, 2013 3:23:36 PM

That's assuming, adn you know what they say about assuming ;)  There's just no way to tell until it's out and then what you get is luck of the draw, might get a good one, might get a bad one, reviews are the same, might see a crappy review and it was written by an idiot, might see a review that was written by someone who knew what they were doing and got a crappy CPU......too many IFs involved
m
0
l
a b à CPUs
July 6, 2013 3:28:28 PM

Tradesman1 said:
That's assuming, adn you know what they say about assuming ;)  There's just no way to tell until it's out and then what you get is luck of the draw, might get a good one, might get a bad one, reviews are the same, might see a crappy review and it was written by an idiot, might see a review that was written by someone who knew what they were doing and got a crappy CPU......too many IFs involved


Precisely. You really cannot know until you own the chip yourself. Only then can you truly give out information about the chip, and even then the same problems arise; you may have gotten a bad chip, etc.
m
0
l
a c 535 à CPUs
July 6, 2013 3:38:10 PM

and even then, it's about your chip - hence luck of the draw - Asus did testing with couple hundred Haswell and found 70% could OC 4.5, 30% to 4.6, 20% to 4.7 and 10% could hit 4.8........so what if you draw from that lower 30% and can only do 4.3 or even less......and Joe Blow gets one in the upper 5% and hit's 4.9 or better - test on a single CPU don't mean squat.....I know when I speak on say the 3570K, I'm generally thinking an overall average of all the ones I've seen (at times I think too many)
______________
Edit: in case anybody ask about the Asus testing, had to dig this up:

http://www.hardocp.com/article/2013/06/01/intel_haswell...
m
0
l
a b à CPUs
July 6, 2013 3:40:12 PM

Tradesman1 said:
and even then, it's about your chip - hence luck of the draw - Asus did testing with couple hundred Haswell and found 70% could OC 4.5, 30% to 4.6, 20% to 4.7 and 10% could hit 4.8........so what if you draw from that lower 30% and can only do 4.3 or even less......and Joe Blow gets one in the upper 5% and hit's 4.9 or better - test on a single CPU don't mean squat.....I know when I speak on say the 3570K, I'm generally thinking an overall average of all the ones I've seen (at times I think too many)
______________
Edit: in case anybody ask about the Asus testing, had to dig this up:

http://www.hardocp.com/article/2013/06/01/intel_haswell...


First of all, I agree. Second of all, that's an awesome find. You seem to be quite good at this ;) 
m
0
l
a c 535 à CPUs
July 6, 2013 3:45:26 PM

It's my job, that's why I like the forums, helps me keep up with all the hardware, most of my clients are upper tier users that have a good understanding of systems, and want high end systems (which makes me happy, I get paid to play), but if they have questions they want answers, constantly researching, networking with other builders/consultants/It folks and folks that work for the manufacturers
m
0
l
a b à CPUs
July 6, 2013 3:47:40 PM

Tradesman1 said:
It's my job, that's why I like the forums, helps me keep up with all the hardware, most of my clients are upper tier users that have a good understanding of systems, and want high end systems (which makes me happy, I get paid to play), but if they have questions they want answers, constantly researching, networking with other builders/consultants/It folks and folks that work for the manufacturers


That's pretty awesome... It's also one of the reasons why I like the forums; although this isn't my job, I love keeping up with all the current tech and whatnot, and solving peoples' problems with it is probably the best way to do so.
m
0
l
a c 535 à CPUs
July 6, 2013 3:49:00 PM

Yeppper ;) 
m
0
l
!