amd fx cpus were future proof...

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

fx still has legs?

  • Hell yeah!

  • no really

  • Hell no!


Results are only viewable after voting.

master_shake_

Diamond Member
May 22, 2012
6,425
292
121
That's probably the most honest assessment I've ever heard on the topic.

The FX 8 cores are probably the Dodge Vipers of CPU's. Is the new Corvette faster than a Viper? Sure. Is the Viper a gas guzzler? Absolutely. Does a Viper honestly struggle in any performance metric? Nope.


so wrong.

an fx cpu is like a brz, it's a sports car unless you compare it to other sports cars.
 
Apr 20, 2008
10,067
990
126
After reading this sort of thing over and over and over, and countless benchmarks, I still can't figure out why I've had such a great experience with FX CPUs. I've been at this for almost twenty years now, from 386's to this, so it can't be lack of experience. And I'm just as bright and technically savvy as the next person among this crowd. I bought an i7-4790K and a nice z97 board recently and I intend to see how differently it performs than the higher end FX chips. I'm a believer in putting ones money where their mouth is.

I've also never felt compelled to use the word "turd" in discussing any item or ideal that another person happens to favor. In fact I've never even typed it in all these years. Churchill said if you must kill a man it costs nothing to be polite, thus far I've found that to be accurate.

It's a combination of sheep mentality where people parrot what others have said, financial interest (notice how many people are obsessed with quarterly reports, you NEVER see such fanaticism in any other field) and in my opinion, fanboys.

Consumers and actual enthusiasts alike have a completely different take on it. Just looking at the steam hardware survey is enough information for the gaming side. If AMD has 25% of the gaming marketshare alone despite being outsold in laptops and OEMs by a huge margin, they're still very competitive. Gaming is supposedly their weakest link and yet 1 of 4 systems has their processors.

If AMD had products that weren't competitive they'd be in the same spot as Windows phones. Slashed prices, near zero marketshare.
 

Killrose

Diamond Member
Oct 26, 1999
6,230
8
81
I recently went with AMD again after 7yrs with a Gigabyte Mboard with a 690 chipset and various CPU's that came to a head with a 3core PhenomII in it for the last few years. I remember when I bought that how we AMD hopefuls were betting on Phenom to pull thru.

Bought a MSI 970 gaming with a FX 8320e and like it so far. I very well knew that even a regular i5 in a cheap Mboard that would at least let it do its Turbo core dance would beat out my new build, but I got a rare deal on both the MSI Mboard and FX cpu and some ebay memory so I went with it all.

Next time you Intel boys see me in the Bar you owe me a beer for helping to keep some sort price controls on your cpu parts, because god only knows what Intel would be charging you if AMD was not even able bring something to the bench :)
 

crashtech

Lifer
Jan 4, 2013
10,695
2,294
146
It should be OK to be disrespectful of a product brand, they are corporate creations and not members of one's family. I think a lot of people lose sight of that and get way too emotionally invested in brands. This is especially puzzling when the product is an inanimate hunk of silicon, it's not even a football team or a politician.
 
Apr 20, 2008
10,067
990
126
It should be OK to be disrespectful of a product brand, they are corporate creations and not members of one's family. I think a lot of people lose sight of that and get way too emotionally invested in brands. This is especially puzzling when the product is an inanimate hunk of silicon, it's not even a football team or a politician.

Do you know how childish it is to see a grown adult name calling something they don't like? The fact that it isn't a sports team or politician makes it reek of fanaticism. You've got to be way too emotionally involved in brands to start name calling an inanimate hunk of silicon.

When a reasonable person doesn't like something that has no effect on their life they thoughtfully discuss it or just outright ignore it.
 

MiddleOfTheRoad

Golden Member
Aug 6, 2014
1,123
5
0
What satire? He's right.

He's never been right about the FX. It's just the rantings of a heavily biased forum member. There is nothing my i7 does than my FX 8320 can't do -- and the ancient FX is way closer in real world usage than Shintai can accept.

Again, I will remind people of the FX-8350 Phoronix review where the FX outperformed the i7 3770K for most of the benchmarks. The Vishera FX was a reasonably good performer by 2012 standards. Obviously, the world has moved on now and AMD is stuck with a power hungry old node -- but his revisionist history is great satire. I've never met a person who bought a Vishera back in 2012 that wasn't very satisfied with the chip.
 

MiddleOfTheRoad

Golden Member
Aug 6, 2014
1,123
5
0
so wrong.

an fx cpu is like a brz, it's a sports car unless you compare it to other sports cars.

The Suburu BR-Z has gotten extremely positive reviews from every magazine I've read.
The Miata and BR-Z are extremely fun, nimble sports cars to drive. If picking the BR-Z was intended as a insult, that's a pretty epic fail.
 

erunion

Senior member
Jan 20, 2013
765
0
0
The Suburu BR-Z has gotten extremely positive reviews from every magazine I've read.
The Miata and BR-Z are extremely fun, nimble sports cars to drive. If picking the BR-Z was intended as a insult, that's a pretty epic fail.

This is why car analogies should be avoided. You are now arguing overs cars and not the subject at hand.
 

TheELF

Diamond Member
Dec 22, 2012
4,027
753
126
He's never been right about the FX. It's just the rantings of a heavily biased forum member.

Again, I will remind people of the FX-8350 Phoronix review where the FX outperformed the i7 3770K for most of the benchmarks.

Come on.... you can't call someone heavily biased and then go and refer to a benchmark that focuses purely on parallel workloads(if you're talking about this one), and obscure ones at that,the kind no one (no average user anyway) will ever use.
And even then with double the cores and with benches that are purely paralleled the lead of the 8350@4,6 isn't all that impressive.
I mean its much better at some benches about the same at others and much worse at some.
A clear winner that's not.
 

ShintaiDK

Lifer
Apr 22, 2012
20,378
146
106
Next time you Intel boys see me in the Bar you owe me a beer for helping to keep some sort price controls on your cpu parts, because god only knows what Intel would be charging you if AMD was not even able bring something to the bench :)

We already know what Intel would charge. They charge it today.
 

crashtech

Lifer
Jan 4, 2013
10,695
2,294
146
Do you know how childish it is to see a grown adult name calling something they don't like? The fact that it isn't a sports team or politician makes it reek of fanaticism. You've got to be way too emotionally involved in brands to start name calling an inanimate hunk of silicon.

When a reasonable person doesn't like something that has no effect on their life they thoughtfully discuss it or just outright ignore it.
I much prefer to see someone trash a brand name than to be personally insulting, something I have witnessed often here. The part that is sad and truly childish is to see those who identify with a brand so strongly that to insult it is to insult them, a pathetic displacement of one's sense of honor.
 
Aug 11, 2008
10,451
642
126
Yea, I think the expression is "the pot calling the kettle black". Some of those who complain most about other posters should perhaps examine their own.
 

Killrose

Diamond Member
Oct 26, 1999
6,230
8
81
We already know what Intel would charge. They charge it today.

What kinda beer do you recommend from your part of the world? Maybe i'll un-cap one this weekend. I'll have a swig as I wait for frames to load on my FX system while gaming, LOL.
 

monstercameron

Diamond Member
Feb 12, 2013
3,818
1
0
you guys are so cold, why would you spend hard earned cash on something you aren't passionate about?
you have every right to be filled with passion, the problem is how to express it.
 

Killrose

Diamond Member
Oct 26, 1999
6,230
8
81
you guys are so cold, why would you spend hard earned cash on something you aren't passionate about?
you have every right to be filled with passion, the problem is how to express it.

You're just the guy I need to talk too. What should I get my girlfriend for Valentines Day?
 

DrMrLordX

Lifer
Apr 27, 2000
23,041
13,138
136

Hey now, this thread is about the mighty FX. We've done enough harassing AMD over their "oversized" iGPUs. Besides, reread the graphs before posting conclusions.

You're just the guy I need to talk too. What should I get my girlfriend for Valentines Day?

An FX-8310!!!!

. . .

no. FX-8310 with a nice Chianti?
 
Last edited:

xthetenth

Golden Member
Oct 14, 2014
1,800
529
106
you guys are so cold, why would you spend hard earned cash on something you aren't passionate about?
you have every right to be filled with passion, the problem is how to express it.

I did. I'm passionate about monitors. However I fully recognize that there are other very good options and that them being very good doesn't deduct from how good my choice is.

yea, humans aren't my forte :sneaky:

Perfect. What sort of roses do you recommend for an LG 34UM95?
 

MiddleOfTheRoad

Golden Member
Aug 6, 2014
1,123
5
0
Come on.... you can't call someone heavily biased and then go and refer to a benchmark that focuses purely on parallel workloads(if you're talking about this one), and obscure ones at that,the kind no one (no average user anyway) will ever use.

Hmmm.... Let's think about this for a second. I can't use benchmarks that focus purely on parallel workloads.... When scientific computing (which is what we've been discussing for the last 2 days over several threads) is nearly always parallel workloads.

Or the fact that I referenced Linux benchmarks -- when 83% of the world's Top 500 SuperComputers use Linux.... for drum roll please..... mostly scientific computing / simulation.

Thanks for the input from the rocket scientist of this forum.

But for those who actually have a little common sense, these are some good reads:
http://www.zdnet.com/article/fast-faster-fastest-linux-rules-supercomputing/
http://www.phoronix.com/scan.php?page=article&item=amd_fx8350_visherabdver2&num=1
 

Enigmoid

Platinum Member
Sep 27, 2012
2,907
31
91
Hmmm.... Let's think about this for a second. I can't use benchmarks that focus purely on parallel workloads.... When scientific computing (which is what we've been discussing for the last 2 days over several threads) is nearly always parallel workloads.

Or the fact that I referenced Linux benchmarks -- when 83% of the world's Top 500 SuperComputers use Linux.... for drum roll please..... mostly scientific computing / simulation.

Thanks for the input from the rocket scientist of this forum.

But for those who actually have a little common sense, these are some good reads:
http://www.zdnet.com/article/fast-faster-fastest-linux-rules-supercomputing/
http://www.phoronix.com/scan.php?page=article&item=amd_fx8350_visherabdver2&num=1

Depends. On a large scale yes. On a small scale there are a TON of individual scientists writing code for research papers and such who run singlethreaded code (myself) because we can't be bothered to parallelize. You are writing code that will take maybe a couple of days to run (time is important because you are waiting on results) but considering the time and effort to parallelize code it simply isn't worth it.

AMD chips simply are not worth the bother for me. Low FP and singlethreaded performance.

Not the mention the TON of problems that can occur when multithreading floating point code. Breaking code into threads and putting them back together will possibly give different results depending on rounding errors.

Generally FP calculations are not even commutative

(a+b)+c != a+(b+c)

due to rounding.

Thus if you need a ton or reproducible accuracy MT is pretty much impossible.

https://software.intel.com/sites/default/files/article/164389/fp-consistency-102511.pdf
 

jhu

Lifer
Oct 10, 1999
11,918
9
81
Depends. On a large scale yes. On a small scale there are a TON of individual scientists writing code for research papers and such who run singlethreaded code (myself) because we can't be bothered to parallelize. You are writing code that will take maybe a couple of days to run (time is important because you are waiting on results) but considering the time and effort to parallelize code it simply isn't worth it.

AMD chips simply are not worth the bother for me. Low FP and singlethreaded performance.

Not the mention the TON of problems that can occur when multithreading floating point code. Breaking code into threads and putting them back together will possibly give different results depending on rounding errors.

Generally FP calculations are not even commutative

(a+b)+c != a+(b+c)

due to rounding.

Thus if you need a ton or reproducible accuracy MT is pretty much impossible.

https://software.intel.com/sites/default/files/article/164389/fp-consistency-102511.pdf

Interesting. Do you use x87 then for better precision or do you just assume 64-bit FP for all your data?
 

Enigmoid

Platinum Member
Sep 27, 2012
2,907
31
91
Interesting. Do you use x87 then for better precision or do you just assume 64-bit FP for all your data?

Depends, i'm just a student. Personally I've just recently started running physics calculations and am in the evaluation process (coding and testing a bunch of stuff separately before putting it together) but multithreading is definitely difficult. Sometimes I can get away with single precision other times double is needed and even then care is needed.

With what I'm doing a certain degree of imprecision is expected, given the theoretical model makes certain assumptions, the calculated result will differ from anything experimentally verifiable. However, the issue is consistency. It may be slightly imprecise due to fp rounding errors but as long as its always imprecise in that fashion its ok.

Double Precision Fortran for me and what I am doing is best. Very accurate and extremely fast (and very good for matrices).
 

escrow4

Diamond Member
Feb 4, 2013
3,339
122
106
Given the sheer number of cores, they will probably have legs for a long time. Just dont expect great performance at any one thing. On a positive note, there is nothing that would absolutely choke and die on an 8 core FX that would fly on a 4790K. The FX will run everything, unlike my G3258 which flies most of the time but will absolutely choke and die if I try to launch multiple apps while an encode is running in the background and I'm streaming a game. :rolleyes:

You expect a poky old G3258 to run multiple apps whilst encoding and streaming? D:

The FX may run everything but it will do so slower than Intel whilst guzzling more juice. Look at the recent techspot budget battle for recent power figures. I second Shintai here, FX now more than ever is a failure. Broadwell and Skylake will embarrass it further (if that is even possible).