Intel Pentium G3258 vs AMD FX 6300

[MiddleOfTheRoad]

Am I missing something? Looks like no difference at all.

You know some people actually read the articles, not just look at the pictures.

Hint: You might try reading the conclusion -- they saw an increase on only certain applications, not every one.

 

[MiddleOfTheRoad]

The thread you link to says nothing substantiated.
Yes it has very strong bias towards single core speeds,that's what we where talking about here wasn't it?

Yeah, right. Dolphin is insanely rigged toward Intel chips.

When the official site insults the AMD chips, you know its rigged....
https://dolphin-emu.org/docs/faq/#which-cpu-should-i-use

The bottom line, the FX performs on par with a Sandy Bridge for single threaded performance -- I don't see them trashing Sandy Bridge's IPC in their FAQ.

 

[jpiniero]

The bottom line, the FX performs on par with a Sandy Bridge for single threaded -- I don't seem them trashing Sandy Bridges in their FAQ.

You're kidding right? If Zen matches Sandy Bridge that would be a good effort given where AMD is right now.

 

[MrTeal]

Yeah, right. Dolphin is insanely rigged toward Intel chips.

When the official site insults the AMD chips, you know its rigged....
https://dolphin-emu.org/docs/faq/#which-cpu-should-i-use

The bottom line, the FX performs on par with a Sandy Bridge for single threaded performance -- I don't see them trashing Sandy Bridge's IPC in their FAQ.

How do you figure?
https://docs.google.com/spreadsheet...aEhhdUpnU01tYUtNM3ZVX2d6SkE&usp=sharing#gid=0
An FX-8350@5GHz takes 15% longer to run the benchmark than a 4.7GHz 2500k. At equal 4.2GHz clocks it takes 20% longer than the 2500k.

That's pretty strange definition of on par.

 

[Phynaz]

Yeah, right. Dolphin is insanely rigged toward Intel chips.

When the official site insults the AMD chips, you know its rigged....
https://dolphin-emu.org/docs/faq/#which-cpu-should-i-use

The bottom line, the FX performs on par with a Sandy Bridge for single threaded performance -- I don't see them trashing Sandy Bridge's IPC in their FAQ.

There is another explanation that you don't seem to want to consider.

 

[TheELF]

Negative. The single threaded score/performance is dependant on what speed you are running the memory on an FX. An FX 6300 with DDR3 1333 will perform significantly different from one running DDR3 2133.

You know some people actually read the articles, not just look at the pictures.

Hint: You might try reading the conclusion -- they saw an increase on only certain applications, not every one.

Oh I did read it! You ,I guess did not!

one application really did jump by 6%

One application was faster,probably the same passmark ram benchmark you are basing your "significantly different" speeds on.... 6% lmao.

 

[TheELF]

Yeah, right. Dolphin is insanely rigged toward Intel chips.

Find something that is insanely rigged towards AMDs chips, I told you two times already and this is the third,find something run it on one core only and lets see.

 

[TheELF]

When the official site insults the AMD chips, you know its rigged....
https://dolphin-emu.org/docs/faq/#which-cpu-should-i-use

How is telling the truth insulting?
The FX8000 is about half the speed of a high end haswell,now I am asking you to show us the speed of the FX-9000 line at 5Ghz since you are in the believe that it does a great amount of difference.

 

[MiddleOfTheRoad]

There is another explanation that you don't seem to want to consider.

Uh, yeah.... It's not like an FX-8350 performs similarly to an i7 3770k under Linux.... Oh wait, it actually does. Must be nice when the software doesn't artificially sandbag the processor
like the Windows programs that use the Intel compiler.

http://www.phoronix.com/scan.php?page=article&item=amd_fx8350_visherabdver2&num=10

 

[MiddleOfTheRoad]

How is telling the truth insulting?
The FX8000 is about half the speed of a high end haswell,now I am asking you to show us the speed of the FX-9000 line at 5Ghz since you are in the believe that it does a great amount of difference.

Only on Anandtech is the FX 9590's 1720 single threaded score half of 2300 (and that's the latest holy grail of Skylakes, the considerably newer 6700K) single threaded Passmark score.

Gotta love the mathematicians that frequent this forum.

 

[MiddleOfTheRoad]

Oh I did read it! You ,I guess did not!

One application was faster,probably the same passmark ram benchmark you are basing your "significantly different" speeds on.... 6% lmao.

6% is a larger gain than upgrading from Haswell to Skylake for the majority of tasks. Care to put your foot in your mouth some more?

 

[MiddleOfTheRoad]

Find something that is insanely rigged towards AMDs chips, I told you two times already and this is the third,find something run it on one core only and lets see.

Compared to its circa 2012 rivals (Vishera is a really old damn chip), an FX thrives in integer and multithreaded workloads. Video editing, rendering, encoding and number crunching/scientific calculations is what the server-derived FX is best suited
to do. These tasks -- the FX runs toe to toe with an Ivy Bridge. For gaming, the FX usually performs between the performance of a Nehalem to Ivy..... Depending on how the game was written/optimized.

Considering the FX-8350 debuted at $199.99 retail and the i7 3770K was priced at $315.00 -- the FX was a very compelling purchase back in 2012. I definitely would not recommend a fresh AM3+ build now that it 2015, but if you inherited a
few AM3 parts -- it remains a solid offering. My overclocked FX-8350 wasn't quite enough for me last year when I started gaming in 1440p, which is why I'm running a 4790k now. But people greatly exaggerate the performance penalty of
running FX chips. They are much better than what many journalists would have you believe / expect.

 

[TheELF]

Uh, yeah.... It's not like an FX-8350 performs similarly to an i7 3770k under Linux.... Oh wait, it actually does. Must be nice when the software doesn't artificially sandbag the processor
like the Windows programs that use the Intel compiler.

http://www.phoronix.com/scan.php?page=article&item=amd_fx8350_visherabdver2&num=10

Sure with software from 2012 ,that nobody ever heard of or knows what they do,the fx @4,6 is kinda close to the i5 of the times....if you use an OS that does not have any optimizations....

Are you at any time going to tell us what kind of single speed you get?

Only on Anandtech is the FX 9590's 1720 single threaded score half of 2300 (and that's the latest holy grail of Skylakes, the considerably newer 6700K) single threaded Passmark score.

Gotta love the mathematicians that frequent this forum.

Gotta love your mathematicians does anybody even know what Passmark is measuring,because it certainly is not single threaded speed?

I'm still waiting on you to explain your beginning statement of overclocking the FX-6300 closing the gap in single threaded performance significantly.

Compared to its circa 2012 rivals (Vishera is a really old damn chip), an FX thrives in integer and multithreaded workloads. Video editing, rendering, encoding and number crunching/scientific calculations is what the server-derived FX is best suited
to do. These tasks -- the FX runs toe to toe with an Ivy Bridge. For gaming, the FX usually performs between the performance of a Nehalem to Ivy..... Depending on how the game was written/optimized.

Considering the FX-8350 debuted at $199.99 retail and the i7 3770K was priced at $315.00 -- the FX was a very compelling purchase back in 2012. I definitely would not recommend a fresh AM3+ build now that it 2015, but if you inherited a
few AM3 parts -- it remains a solid offering. My overclocked FX-8350 wasn't quite enough for me last year when I started gaming in 1440p, which is why I'm running a 4790k now. But people greatly exaggerate the performance penalty of
running FX chips. They are much better than what many journalists would have you believe / expect.

Bla bla bla bla bla multithreading rules...
Show us your single speed at 5Ghz,go on make us laugh.

6% is a larger gain than upgrading from Haswell to Skylake for the majority of tasks. Care to put your foot in your mouth some more?

Skylake has a larger gain in more then just one program it has at least 5% everywhere, in whatever you run and not just when benchmarking ram.

 

[jhu]

Yeah, right. Dolphin is insanely rigged toward Intel chips.

When the official site insults the AMD chips, you know its rigged....
https://dolphin-emu.org/docs/faq/#which-cpu-should-i-use

The bottom line, the FX performs on par with a Sandy Bridge for single threaded performance -- I don't see them trashing Sandy Bridge's IPC in their FAQ.

uhm, no. They're using MSVC to compile it for Windows. For Linux they have instructions for using GCC. Unless there's somewhere in the source code where you can find that they explicitly cripple AMD CPUs, then your premise is false.

BTW, Sandy Bridge has about 1.5x single threaded IPC over Piledriver.

 

[Abwx]

BTW, Sandy Bridge has about 1.5x single threaded IPC over Piledriver.

The Xeon E5-2699 v3 and 2695 v3 boost their clock speed to no less than 3.6GHz when only one or two cores are active. The Xeon E5-2667 v3's maximum Turbo Boost is also the same 3.6GHz, so when only a few threads are active, the Xeon E5-2667 v3 has no clock advantage over the "mega/expensive SKUs" other than the fact that the clock speed will not drop lower than 3.2GHz if all cores are running at full bore.

http://www.anandtech.com/show/8423/intel-xeon-e5-version-3-up-to-18-haswell-ep-cores-/12

 

[jhu]

http://www.anandtech.com/show/8423/intel-xeon-e5-version-3-up-to-18-haswell-ep-cores-/12

GCC 4.8 with only -O2? What are they with -O3 and the proper -march options?

 

[Abwx]

GCC 4.8 with only -O2? What are they with -O3 and the proper -march options?

Ask AT for more precision, there s more recent slides but without Piledriver.

That said IPC is not constant with threads and app count, even with a low number, as shown by Computerase.de.

http://www.computerbase.de/2015-12/...m-multitasking-test-winrar-plus-the-witcher-3

 

[lyssword]

For desktop apps in general, fx6300 does pretty well vs i3, its the gaming where it loses.

 

[TeknoBug]

lol man one guy getting burned here, need to call the fire department.

 

[MrTeal]

Only on Anandtech is the FX 9590's 1720 single threaded score half of 2300 (and that's the latest holy grail of Skylakes, the considerably newer 6700K) single threaded Passmark score.

Gotta love the mathematicians that frequent this forum.

Got to love the substituting in another part when the poster said Haswell and PassMark lists the top Haswell part at 2530. Got also to love subbing in the FX-9590 when the fastest FX-8000 part is the FX-8370 at 1524. It's still not half as fast even in your preferred benchmark, but it's pretty much exactly the same calibre of mathematics you used when you claimed a $599 Zen would be about half the price of a $999 i7-5960X.