FX6300 + GTX770

[kefan77]

quick question:

will stock running FX6300 bottleneck single GTX770?

my assumption is yes, but need 2nd opinion.

how much would that change if the CPU is overclocked (say 4.5)

 

[Marine-RX179]

It would be game dependent. But general answer would be yes.

Particularly if the game doesn't use all 6 cores, even with 4 cores used, it could bottleneck the 770 by an average of 40-50% (depending on graphic settings which you use), more if they games use less than 4 cores. The FX CPUs are fairly underpowered if you not overclock them...clock for clock and per core performance, they are only on par with Core2/Phenom II level at best.

Also, you'd better off getting the 7970 which would be faster and cheaper, and you could put the money saved toward CPU budget (if you haven't got the FX6300 and 770 yet). IMO you would get much better overall performance with an i5 and a 7950, than a FX6/8 with a 770/7970. If you already have the FX6300, then just focus on overclocking it.

 

[swiftmetal]

As above these are poor chips if you can get an 8320 and have a decent motherboard I would do that first and then clock it to at least 4.5 GHz and then you should not have as bigger bottle neck and please save your self some money and get something like 7950/7970 as at the price they are now they are a steel and out perform the NVIDIA cards at a higher price point.

 

[CAT-THE-FIFTH]

As above these are poor chips

No they are not,as I know loads of people who use them in builds. The FX6300 at around £80 to £85 has been one of the best budget CPUs under £100 for a while. In games like WoW and SC2 which are lightly threaded,it will destroy the older AMD CPUs,even at the same clockspeed. Both GameGPU and Toms Hardware tested this.

 

[ubersonic]

No they are not,as I know loads of people who use them in builds. The FX6300 at around £80 to £85 has been one of the best budget CPUs under £100 for a while.

Just because a lot of people use them and they are good value for money when compared to other new chips doesn't stop the architecture being poor, a chip launched at the end of 2012 should not have lower clock for clock performance per core than ones from 2007, it's that simple.

In games like WoW and SC2 which are lightly threaded,it will destroy the older AMD CPUs,even at the same clockspeed. Both GameGPU and Toms Hardware tested this.

In WoW the FX6300 gets less than 1 FPS more than the equally clocked AMD Phenom II X4 970 BE, less than 2 FPS more than the lower clocked AMD Phenom II X6 1100T BE, and lower FPS than the AMD Phenom II X4 980 BE, In addition it get beat by the Intel Pentium G620.

You can ignore the THG testing on WoW it's worthless, Anand do a much more competent job.

 

[Cool Breeze]

As above these are poor chips if you can get an 8320 and have a decent motherboard I would do that first and then clock it to at least 4.5 GHz and then you should not have as bigger bottle neck and please save your self some money and get something like 7950/7970 as at the price they are now they are a steel and out perform the NVIDIA cards at a higher price point.

Er thats complete BS about the 7950/7970 a GTX 770 beats these all day long, and I come from the stand point of having owned all of em.

 

[CAT-THE-FIFTH]

Just because a lot of people use them and they are good value for money when compared to other new chips

Which are the only important metrics.

doesn't stop the architecture being poor, a chip launched at the end of 2012 should not have lower clock for clock performance per core than ones from 2007, it's that simple.

Not relevant at all,since you need to compare same priced CPUs. That would be the Core i3 for the £80 to £85 FX6300 and the Core i3 for the £110 to £115 FX8320.

Also,what is hilarious is that many of you would take a lower clocked Core i5 instead of a higher clocked Core i3,even if the latter had better single core performance. None of you would disable half the cores on a Core i5 to overclock it 10% higher would you?? Nope??

But then none of you would ever use anything under a Core i5 anyway,so its a moot point too when you are talking about cheaper CPUs.

So,in the end its just pathetic E-PEEN and a case of people just trying to feel better. Anything good to be said about any AMD CPU,needs to be burned with fire!!

It also comes from people who have no clue of ever using a single of these systems in real life at all. I have used recent Core i3,Core i5,Core i7 and FX63**,FX83** and A6,A8 and A10 based systems in the realworld too.

BTW,I have a Xeon E3 and a GTX660,so you cannot try that other angle also.

In WoW the FX6300 gets less than 1 FPS more than the equally clocked AMD Phenom II X4 970 BE, less than 2 FPS more than the lower clocked AMD Phenom II X6 1100T BE, and lower FPS than the AMD Phenom II X4 980 BE, In addition it get beat by the Intel Pentium G620.

You can ignore the THG testing on WoW it's worthless, Anand do a much more competent job.

Of course you ignore the THG results,because it does not fit your biased opinion. Or the fact the G620 gets wasted in many newish games too. Even the Core i3 has its own issues and I had a Core i3 myself.

You don't seem to understand Anandtech bench has a load of problems. One of them being they mismatch test results with different patches and different test areas. For example,all reviewers now test BD and PD CPUs under Win8 or Win7 with the patches. Anandtech tested the FX8150 with the patch which increased performance in a number of games. Then they forgot to update the bench tables.

Its the same with their GPU bench tables too. They keep mismatching stuff all the time using different drivers. They keep forgetting to update the tables.

Its is a low priority for them and it shows.

As the site has expanded Anand does less and less of the work himself and farms it out to part time writers from all over the world,and it is showing now.

At least TH retests all the games in one go.

None of these sites are perfect,but thats the way the cookie crumbles.

Also,I actually know quite a few WoW players who use AMD system. So I actually noted performance on their systems when they were raiding.

The PD based upgrades they did,lead to better performance than either the Athlon II or Phenom II based systems they used previously. However,YMMV.

 

[gpuerrilla]

In WoW the FX6300 gets less than 1 FPS more than the equally clocked AMD Phenom II X4 970 BE, less than 2 FPS more than the lower clocked AMD Phenom II X6 1100T BE, and lower FPS than the AMD Phenom II X4 980 BE, In addition it get beat by the Intel Pentium G620.

You can ignore the THG testing on WoW it's worthless, Anand do a much more competent job.

WoW is dead. It's coming up to 10 years old. These types of games benefit far more from a decent internet connection and less people in the game area.

 

[gpuerrilla]

So,in the end its just pathetic E-PEEN and a case of people just trying to feel better.
...
It also comes from people who have no clue of ever using a single of these systems in real life at all.
...
Of course you ignore the THG results,because it does not fit your biased opinion.
...
all reviewers now test BD and PD CPUs under Win8 or Win7 with the patches. Anandtech tested the FX8150 with the patch which increased performance in a number of games. Then they forgot to update the bench tables.

Points I agree with - it's getting tiring not so much from the usual suspects but these random jib ins from people with no weight behind their points who are just like sheep following the biased factoids they absorb.

Yesterday a stereotypical intel user posted on the Cinebench thread - comparing a £500 CPU to the £150 FX8 without applying the brain. Yeah the BD/PD era didnt quite perform as expected but they are still a good performing component for the price.

 

[Marine-RX179]

As above these are poor chips if you can get an 8320 and have a decent motherboard I would do that first and then clock it to at least 4.5 GHz and then you should not have as bigger bottle neck and please save your self some money and get something like 7950/7970 as at the price they are now they are a steel and out perform the NVIDIA cards at a higher price point.

Just to clarify I DID NOT say the FX6300 is a poor chip. The FX6300 is great chip for its asking price £80 ish, but if people were running them at stock clock, then they are doing it wrong...as these chip really should all be overclocked to reach its full potential.

My point was just that rather than paying £330+ for GTX770, OP would be better off spending £230-£250 getting a faster 7970 instead, or just £180 on a 7950, and put the £100+ saved toward a FX8, or a i5 platform. The price premium that the GTX770 is asking for doesn't reflect in performance if comparing to the 7950/7970...and least not large enough for it to be justified.

 

[cheesyboy]

FX-6300 CPU benchmarks (attached to a GTX680 - A slightly downclocked GTX 770);
http://www.xbitlabs.com/articles/cpu/display/fx-8350-8320-6300-4300_6.html#sect0

At 1080p, there is no bottleneck. In fact, only the Pentium shows any genuine bottleneck at 1080p.

Fairly old games, to be fair.

 

[CAT-THE-FIFTH]

Maybe the OP needs to clarify whether they are considering getting an FX6300 and a GTX770,or already have an FX6300.

If it is the first case,then as others have said,perhaps spend more on the CPU and less on the card.

You can get an HD7970 or GTX680 for under £250 these days. Even an HD7950 or GTX670 for under £200.

If they have an FX6300 already the HD7950 or GTX670 would be still OK. The price difference between an HD7850 or GTX660 and them is less than £50 IIRC.

Points I agree with - it's getting tiring not so much from the usual suspects but these random jib ins from people with no weight behind their points who are just like sheep following the biased factoids they absorb.

Yesterday a stereotypical intel user posted on the Cinebench thread - comparing a £500 CPU to the £150 FX8 without applying the brain. Yeah the BD/PD era didnt quite perform as expected but they are still a good performing component for the price.

Yep. Don't get me wrong if you have a very high budget,then go Core i5 or Core i7,but for the price the AMD CPUs are nowhere as bad as some are indicating.

Anyway,I have said what I wanted to say in this thread,so I am out. I can see where it will go otherwise!

 

[teppic]

I agree that these are good chips. Overclocked to 4.5GHz they're exceptional for the money. As tomshardware's review says:

The shining star in today’s comparison is AMD’s FX-6350, which delivers solid performance in games, while besting Intel's Core i5 in a number of our other benchmark workloads. The cheaper FX-6300 is an even more attractive bargain, so long as you're willing to overclock it.

And their results exclude games like Tomb Raider, which perform exactly the same on a 6300 and a 4770K.

 

[ubersonic]

Which are the only important metrics.

Well if you think that excuses the poor performance of the architecture I guess they may be...

Not relevant at all,since you need to compare same priced CPUs.

I was.

That would be the Core i3 for the £80 to £85 FX6300 and the Core i3 for the £110 to £115 FX8320.

Both of which also have more performance per core clock for clock, but are also handicapped CPU's (on overclocking anymore) as a result of Intel resting on its laurels.

Spoiler

Also,what is hilarious is that many of you would take a lower clocked Core i5 instead of a higher clocked Core i3,even if the latter had better single core performance. None of you would disable half the cores on a Core i5 to overclock it 10% higher would you?? Nope??

But then none of you would ever use anything under a Core i5 anyway,so its a moot point too when you are talking about cheaper CPUs.

So,in the end its just pathetic E-PEEN and a case of people just trying to feel better. Anything good to be said about any AMD CPU,needs to be burned with fire!!

It also comes from people who have no clue of ever using a single of these systems in real life at all. I have used recent Core i3,Core i5,Core i7 and FX63**,FX83** and A6,A8 and A10 based systems in the realworld too.

BTW,I have a Xeon E3 and a GTX660,so you cannot try that other angle also.

I thing you've missed the point, I wasn't saying the FX chips are bad, or that the FX-4/6 are not the best chip currently on the market at their price points, I was taking exception to your statement that the FX's are not poor chips. Just because something is cheap and able to compete well with opponents that don't make an effort doesn't stop it being poor, the fact that piledriver cores are slower per clock than Core2 cores and AMD is forced to mount bunches of them onto a CPU and raise the clocks just to compete shows how poor the architecture is. If current i3's were overclockable they would walk over the FX-4's like the i3 5xx's did and in many things (like WoW) the FX-6 too.

Of course you ignore the THG results,because it does not fit your biased opinion.

Actually I ignore them because I know how their obtained and that its an inaccurate method of measuring performance, they basically hop on a flight path and watch the FPS, it's not a valid representation of playing the game and the test itself will change each time. It's the equivalent of testing an FPS by auto-running across a map during a random game.

WoW is dead. It's coming up to 10 years old. These types of games benefit far more from a decent internet connection and less people in the game area.

WoW was the example he used so I merely corrected him on it.

 

[cheesyboy]

I thing you've missed the point, I wasn't saying the FX chips are bad, or that the FX-4/6 are not the best chip currently on the market at their price points, I was taking exception to your statement that the FX's are not poor chips. Just because something is cheap and able to compete well with opponents that don't make an effort doesn't stop it being poor, the fact that piledriver cores are slower per clock than Core2 cores and AMD is forced to mount bunches of them onto a CPU and raise the clocks just to compete shows how poor the architecture is. If current i3's were overclockable they would walk over the FX-4's like the i3 5xx's did and in many things (like WoW) the FX-6 too.

That's rather an outdated view of CPU architecture. In a world where applications are all moving towards multi-threaded workloads, loading your design path in the direction of a fairly radical multi-core design is surely not "poor".

AMD's modular architecture, with certain resources shared across pairs or groups of cores has massive potential as we move towards "hetrogenous computing". That's all about spreading the workload through the most appropriate resources. Ultimately, it's the pathway to greater efficiency too - lots of slow spinning cores can very easily modify their power profile on the move to match workloads.

The problem has been a mixture of teething problems (Bulldozer), R&D challenges surrounding such a radical design, and an inhospitable/immature software climate.

Nonetheless, if AMD are only targeting entry/mid-range (which they have confirmed) and are giving a good argument for being the best choice at those price points....well, how is that poor? Is the ARM architecture poor because it can't compete with high end desktop CPUs?

 

[Marine-RX179]

FX-6300 CPU benchmarks (attached to a GTX680 - A slightly downclocked GTX 770);
http://www.xbitlabs.com/articles/cpu/display/fx-8350-8320-6300-4300_6.html#sect0

At 1080p, there is no bottleneck. In fact, only the Pentium shows any genuine bottleneck at 1080p.

Fairly old games, to be fair.

Actually those benches are clearly showing the GTX680 in tests being bottlenecked by the stock clock FX6300. It's only not bottlenecking when faster CPUs such as i5/i7 (or FX8350 even) are not giving extra frame rate over the FX-6300.

As I said, if people were running the FX6300 on stock clock and not overclocking it, they are doing it WRONG.

 

[cheesyboy]

Actually those benches are clearly showing the GTX680 in tests being bottlenecked by the stock clock FX6300. It's only not bottlenecking when faster CPUs such as i5/i7 (or FX8350 even) are not giving extra frame rate over the FX-6300.

As I said, if people were running the FX6300 on stock clock and not overclocking it, they are doing it WRONG.

You know the bit where I said "at 1080p"? Well, that should have given you a clue that I was referring to the 1080p figures.

And the very small advantage the Intel CPUs hold in some of the 1080p benchmarks is not a CPU bottleneck - since you can see by the low-res scores that the AMD CPU has more to give (as do the others). It's a platform limitation - probably memory - but not a bottleneck.

Far Cry 2 is possibly the exception. But I'm not sure 100+ fps at ultra settings is much to worry about.

 

[Marine-RX179]

You know the bit where I said "at 1080p"? Well, that should have given you a clue that I was referring to the 1080p figures.

And the very small advantage the Intel CPUs hold in some of the 1080p benchmarks is not a CPU bottleneck - since you can see by the low-res scores that the AMD CPU has more to give (as do the others). It's a platform limitation - probably memory - but not a bottleneck.

I WAS referring to 1080p. Big or small, a bottleneck is a bottleneck when the graphic card's GPU is not being ultilized at 99/100%. That's why I said people should all overclock their FX6300 by default, rather than running them on stock clock.

 

[teppic]

And if a GPU is running at 100% then you have a GPU bottleneck as much as you would with a CPU at 100%. This near obsession with having the GPU running at 100% regardless of whether it makes any real difference to games is strange to say the least.

 

[gpuerrilla]

And if a GPU is running at 100% then you have a GPU bottleneck as much as you would with a CPU at 100%. This near obsession with having the GPU running at 100% regardless of whether it makes any difference to games is strange to say the least.

On top of that, one game may max out GPU or CPU - whereas another will be the opposite. As long as your running quality settings on the high side and it's smooth (not going into the 30+ fps argument) then the system can be accounted for as pretty balanced.

If everyone had the money to burn it would be a boring same old hardware in each system. Reality is that some of us have to try the next best rung down or go for the biggest £/performance. Surely you should be glad people are attaining good experiences from cheaper hardware??