Am I wrong or right?

[Bit8xPr0]

Okay, so I'm in an argument with these people on a Facebook page. They're saying that in no situation can a FX 6300 beat a FX 8350. I'm trying to tell them that my FX 6300 is better then a FX 8350 when clocked at 4.3 GHZ and using applications that do not utilize more then 6 cores. Is this not how it works? If the application is not made to use the cores how will having more solve a bottleneck. 

[Godlygamer23]

If the 6300 is clocked higher and the program utilizes no more than 6 cores, the 6300 will perform better. They use the same architecture.

[Unhelpful]

Okay, so I'm in an argument with these people on a Facebook page. They're saying that in no situation can a FX 6300 beat a FX 8350. I'm trying to tell them that my FX 6300 is better then a FX 8350 when clocked at 4.3 GHZ and using applications that do not utilize more then 6 cores. Is this not how it works? If the application is not made to use the cores how will having more solve a bottleneck. 

If the 6300 is clocked higher and the application uses no more than 6 cores the 6300 will do better... Just remember you can OC an 8350 as well. So the 8350 is generally better. So I would say you are wrong.

 

http://www.anandtech.com/show/6396/the-vishera-review-amd-fx8350-fx8320-fx6300-and-fx4300-tested/4

 

Also don't argue on facebook. That is the easiest way to go full retard.

[Juganog]

no situation can a FX 6300 beat a FX 8350.  what if the 8350 doesnt work? or if i smash it

[Sauron]

If they're saying in "NO SITUATION" an FX 6300 can beat an 8350 and don't believe an overclocked 6300 can beat a stock 8350 in applications with less than 6 threads, then yeah, you're right. Of course the 8350 can be overclocked too though...

[Bit8xPr0]

no situation can a FX 6300 beat a FX 8350.  what if the 8350 doesnt work? or if i smash it

true

[TheRetiredSlave]

If an application uses 6 or less cores, the winner will be whichever CPU is overclocked higher (which could be either the 8350 or the 6300). If the application can use 7 or 8 cores, the 8350 will be faster (unless the 8350 is severely underclocked).

 

 

Edit: I suppose that it could be argued that the 8350 would win in a less-than 6 thread task because it would have more L1 and L2 cache available. This could be attributed to the fact that two cores share the same L1 and L2 cache in a single module, so the 8350 could be able to spread the task across its extra module. (frankly the difference would probably be minimal at best, so the winner would still primarily be decided by the clockspeed)

[Godlygamer23]

Edit: I suppose that it could be argued that the 8350 would win in a less-than 6 thread task because it would have more L1 and L2 cache available. This could be attributed to the fact that two cores share the same L1 and L2 cache in a single module, so the 8350 could be able to spread the task across its extra module. (frankly the difference would probably be minimal at best, so the winner would still primarily be decided by the clockspeed)

I don't think that's how it works. If the program doesn't utilize those extra cores, they aren't used and neither are the resources. Per core, the resources are the same.

[Mush Brain]

The six core will win when clocked higher than the eight core in apps that use no more than 6 threads.

[Bit8xPr0]

Yeah, so him telling me to get an FX 8350 to fix a bottleneck over my FX 6300 @4.3 GHZ didn't make sense. I mean yeah you could overclock the 8350 as well. 

[rtpb5642]

okay OC ur 6300 to 4.5... im just gunna OC my 8350 to 5.3...

[TheRetiredSlave]

I don't think that's how it works. If the program doesn't utilize those extra cores, they aren't used and neither are the resources. Per core, the resources are the same.

 

Not with the architecture used in AMD's recent CPUs.. The 8350 has 4 modules, and each module has 2 cores that share the same L1 and L2 cache (and of course one large block of L3 in the middle of all of the modules)..

 

I could be wrong about this (I'm definitely not an expert), but I believe that if one core of a module is inactive, the active core in that module can take over the unused L1 and L2 cache.. (And, like I said previously, the difference would likely be so minimal that the winner in a real-world test would mainly be decided by the overclock)

 

 

What you're saying would be true about Intel's CPUs, where the L1 and L2 caches are bound to a single core.

[Godlygamer23]

Not with the architecture used in AMD's recent CPUs.. The 8350 has 4 modules, and each module has 2 cores that share the same L1 and L2 cache (and of course one large block of L3 in the middle of all of the modules)..

I could be wrong about this (I'm definitely not an expert), but I believe that if one core of a module is inactive, the active core in that module can take over the unused L1 and L2 cache.. (And, like I said previously, the difference would likely be so minimal that the winner in a real-world test would mainly be decided by the overclock)

What you're saying would be true about Intel's CPUs, where the L1 and L2 caches are bound to a single core.

I was thinking more along the lines of the extra module left over assuming a 6 threaded task. The modules are discrete and therefore the modules can't access each other's resources. Do you understand what I'm saying?

[Sintezza]

FX8350 is better then the FX6300 hands down.

[i_build_nanosuits]

Also to add to this discussion: people are talking 6 cores and 8 cores CPU...but those are 3 modules and 4 modules CPU with only ONE FPU (floating point unit) PER MODULE...

so even with a quad-threaded task the FX8 will be noticeably faster than the 3 module if floating point instructions are to be processed.

 

If we where to compare this to intel cpu's for example...then this would be a 3 core and a 4 core CPU with ''hyper-threading'' in an hardware format (well it's not exactly that but kind of) so yeah, the FX 8 core is noticeably better.

[Sintezza]

Also to add to this discussion: people are talking 6 cores and 8 cores CPU...but those are 3 modules and 4 modules CPU with only ONE FPU (floating point unit) PER MODULE...

so even with a quad-threaded task the FX will be noticeably faster than the 3 module if floating point instructions are to be processed.

 

If we where to compare this to intel cpu's for example...then this would be a 3 core and a 4 core CPU with ''hyper-threading'' in an hardware format (well it's not exactly that but kind of) so yeah, the FX 8 core is noticeably better.

 

Well im not even going to bother to talk about the CMT architecture and how it exaly work.

Because i think that people should learn and do their own research.

I see so manny people make the same mistake over and over again wenn they talk about the FX processors.

 

But yeah thanks for adding this lol

 

Back to the main question, "OP" you are completely wrong

[David1521]

 

Also don't argue on facebook. That is the easiest way to go full retard.

 LMAO so true

[Linux User]

In applications that use 4-6 cores an fpu may help though

[Godlygamer23]

FX8350 is better then the FX6300 hands down.

When utilizing 6 threads with a higher clockspeed? No. The available cache(per module) is the same so that would not be a factor.

[Sintezza]

When utilizing 6 threads with a higher clockspeed? No. The available cache(per module) is the same so that would not be a factor.

 

You understand how the CMT architecture works?

[Godlygamer23]

You understand how the CMT architecture works?

If you have something to explain, you can elaborate.

[Sintezza]

If you have something to explain, you can elaborate.

 

6 threads will never be faster on a FX6300 then a FX8350, simply because the FX8350 has 1 extra module so also 1 extra fpu.

[Godlygamer23]

6 threads will never be faster on a FX6300 then a FX8350, simply because the FX8350 has 1 extra module so also 1 extra fpu.

So how is that extra FPU used when it's assigned to the unused cores? It resides in the unused module. The modules do not share resources except for the L3 cache.

[i_build_nanosuits]

So how is that extra FPU used when it's assigned to the unused cores? It resides in the unused module. The modules do not share resources except for the L3 cache.

6threads WILL take advantage of the extra module by using the 4th fpu to speed things up...thats what shes trying to explain to you i thought i had been clear enough but she's right..

[Godlygamer23]

6threads WILL take advantage of the extra module by using the 4th fpu to speed things up...thats what shes trying to explain to you i thought i had been clear enough but she's right..

Is there some sort of reference to that? Another question. Are you implying that during single threaded loads, all four FPUs are being used in the 8350?