Phoronix : FX 8350 competitive with Core i7 3770K under Linux.

[grimpr]

http://www.phoronix.com/scan.php?page=article&item=amd_fx8350_vishera&num=1

From the initial testing of the brand new AMD FX-8350 "Vishera", the performance was admirable, especially compared to last year's bit of a troubled start with the AMD FX Bulldozer processors.

For many of the Linux computational benchmarks carried out in this article, the AMD FX-8350 proved to be competitive with the Intel Core i7 3770K "Ivy Bridge" processor. Seeing the FX-8350 compete with the i7-3770K at stock speeds in so many benchmarks was rather a surprise since the Windows-focused AMD marketing crew was more expecting this new high-end processor to be head-to-head competition for the Intel Core i7 3570K on Microsoft's operating system.

 

[T_Yamamoto]

"Stock" speeds

And if this is true (unlikely) I'll be surprised

 

[Yuriman]

Stock vs stock the i7 is still a much better chip in general, but given the pricepoint of the FX, I think it's no longer pure lunacy to buy one.

Waiting for some power consumption numbers.

 

[grimpr]

Looks like the optimizations of the GCC compilers are starting to pay off for AMD and its new chips.

 

[Olikan]

not sure if true...but good
http://www.cowcotland.com/articles/1199/test-processeur-amd-fx-8350.html

 

[inf64]

Thanks for that post Olikan. Not sure why it wouldn't be true,unless that website "doctored" the results on purpose (what would they gain by doing that ?). Their test selection is kind of odd. The most useful part of the review (IMO) is the power draw: 8350 draws the same amount of power as 8120 and draws 20+W less than 8150. Also its idle power is considerably lower.

 

[Gigantopithecus]

Their test selection is kind of odd.

For a generally lower-level Windows user forum, sure, but not for people who do more with their computers than slay trolls and edit photos.

I have to build three DNA sequence analysis Linux workstations before the end of the semester and it's clear from the results of Phoronix's benchmarks that a 3770K might not be worth the $100+ premium over the 8350. That's $300+ for the lab that might be better spent elsewhere.

 

[inf64]

I was referring to the review that Olikan posted , not the Phoronix one.

 

[Olikan]

Thanks for that post Olikan. Not sure why it wouldn't be true,unless that website "doctored" the results on purpose (what would they gain by doing that ?). Their test selection is kind of odd. The most useful part of the review (IMO) is the power draw: 8350 draws the same amount of power as 8120 and draws 20+W less than 8150. Also its idle power is considerably lower.

who knows man, this is not a professional review, the guy might do something wrong just by the lulz, or just because he sucks at benching

 

[blckgrffn]

I'm looking forward to seeing how hard it is to get this guy thermally stable @ ~4.6 and how much heat (power usage) you have to dissipate at that clock speed.

BD seemed to OC well, but the eight module parts seemed hard to get really stable above 4Ghz due to the amount of heat you had to deal with.

As a plus for science, you should get to ride on AVX optimizations for Intel with Vishera since it is now implemented the same way on Vishera (at least, that's my understanding...)

 

[frozentundra123456]

http://www.phoronix.com/scan.php?page=article&item=amd_fx8350_vishera&num=1

Nice link, but if you read the entire article, performance was all over the place. Some tests it won, some it was competitive overclocked, and some it did really poorly. So agsin, a selective quote, but at least you linked the entire article so one could read it if they desired.

 

[blckgrffn]

Nice link, but if you read the entire article, performance was all over the place. Some tests it won, some it was competitive overclocked, and some it did really poorly. So agsin, a selective quote, but at least you linked the entire article so one could read it if they desired.

To be fair, it was the conclusion that the author of the article put forth. Quibbling with it seems reasonable based on the data, but it had its source in the article.

I mean, Anandtech.com never has a conclusion we disagree with, right?

 

[Vic Vega]

who knows man, this is not a professional review, the guy might do something wrong just by the lulz, or just because he sucks at benching

I wouldn't consider what you see from AT or other HW web sites any more or less professional. Just because you've been doing something a long time doesn't mean your methods are good (or bad).

 

[hokies83]

This the bench with the stock 3770k with ht off vs the 4.6ghz amd chip? that is kinda joke ya know...

What happens when you OC the 3770k to 4.6ghz lmao.

It looks more even to an i7 950 clock for clock tho...

 

[Arachnotronic]

A golf clap for AMD. It manages to, in special cases, match/beat a 3770K which consumes *half* the power including a BUILT IN graphics chip and with half the cores. Oh, and the die size is monstrous.

Goes to show if you're willing to take garbage margins and brute force your way to some "victories" you can do it. However, AMD is significantly under capitalized compared to Intel, so to expect anything other than what we're seeing is silly.

 

[Idontcare]

I mean, Anandtech.com never has a conclusion we disagree with, right?

Perish the thought, young heretic, lest ye be cleansed with fire and brimstone post haste :twisted:

 

[Arkaign]

This the bench with the stock 3770k with ht off vs the 4.6ghz amd chip? that is kinda joke ya know...

What happens when you OC the 3770k to 4.6ghz lmao.

True, comparing stock v stock, great. Comparing oc vs. oc, great. Comparing stock and OC vs stock and OC, great.

Comparing stock vs. OC = bullcrap.

 

[Vic Vega]

This the bench with the stock 3770k with ht off vs the 4.6ghz amd chip? that is kinda joke ya know...

What happens when you OC the 3770k to 4.6ghz lmao.

It looks more even to an i7 950 clock for clock tho...

I'm looking at the article now. It has numbers for the FX-8350 stock and overclocked. The stock numbers seem to be slightly under i7 3770K with the overclocked numbers being slightly higher (for the tests I actually bothered looking at).

 

[Arkaign]

I looking at the article now. It has numbers for the FX-8350 stock and overclocked. The stock numbers seem to be slightly under i7 3770K with the overclocked numbers being slightly higher.

Where are the 3770K OC numbers?

I'm sure some people run the Intel K-series chips at stock clocks, but not smart ones (else why pay more for a K chip vs. a non-K chip if you don't OC?).

3.5Ghz w/3.9Ghz Turbo stock.

4ghz+ is cake, with a decent chip and cooling setup 4.5Ghz or beyond.

 

[sm625]

Anandtech is too busy reviewing fans to bother reviewing important chips like the FX-8350 or the Pentium 2117U.

 

[hokies83]

It looks 25% slower clock for clock...

That is the same diff between a i7 950 and a 3770k....

Soon as Windows based benchmarks come out ppl will be crushing them like there is no tomrrow.

 

[LoneNinja]

This the bench with the stock 3770k with ht off vs the 4.6ghz amd chip? that is kinda joke ya know...

What happens when you OC the 3770k to 4.6ghz lmao.

It looks more even to an i7 950 clock for clock tho...

Did you even take a look at the article? The FX 8350 is clocked stock and overclocked at 4.6Ghz, and the I7 3770k has hyperthreading enabled.

 

[Vic Vega]

Anandtech is too busy reviewing fans to bother reviewing important chips like the FX-8350 or the Pentium 2117U.

Or giving us one more article about Apple something.

 

[hokies83]

Did you even take a look at the article? The FX 8350 is clocked stock and overclocked at 4.6Ghz, and the I7 3770k has hyperthreading enabled.

ah NM.. still 4.2ghz vs 3.9ghz still looks 25% slower.

 

[Eeqmcsq]

According to the phoronix article date, the article was published tomorrow! We can see into the future!