FX83## owners whats your CPU revition?

[Martini1991]

Yet,how many people have a 4.4GHZ Phenom II X6 and a 3.3GHZ NB one?? There might be people with 5.3GHZ FX8350 CPUs out there for example.

Yet a Phenom II X4 at 3.8GHZ with a 2.6GHZ NB,couldn't beat a FX6300 in SC2.

The same goes with SC2 HoTS.

Well, I did say it was far from the norm....................
I could run it faster as a 4 Quad core too.

I've ran a Callisto at 4.5GHZ (And that was on air)

I honestly don't understand why you're always gagging for an argument, making up points (Valid as they may be) no one's arguing against or trying to contradict.
Would the GameGPU end results change with a Phenom II having an overclocked CPU NB? Yes, there's no if's or buts about it. Even with the CPU NB at 2.8/3GHZ, the PD may well have parity or even be ahead (No one's said to the contrary).

 

[CAT-THE-FIFTH]

Well, I did say it was far from the norm....................
I could run it faster as a 4 Quad core too.

I've ran a Callisto at 4.5GHZ (And that was on air)

I honestly don't understand why you're always gagging for an argument, making up points (Valid as they may be) no one's arguing against or trying to contradict.
Would the GameGPU end results change with a Phenom II having an overclocked CPU NB? Yes, there's no if's or buts about it. Even with the CPU NB at 2.8/3GHZ, the PD may well have parity or even be ahead (No one's said to the contrary).

Well I replied to you original post and not the edit LOL.

Its more the point people seem to want to cling to the arguments that PD is always slower clock for clock than the Phenom II CPUs.

We have seen with better optimisation this is not always the case,even in games which should not be better on the FX CPUs(which do have narrower cores than the Phenom II).

Would the GameGPU end results change with a Phenom II having an overclocked CPU NB? Yes, there's no if's or buts about it. Even with the CPU NB at 2.8/3GHZ, the PD may well have parity or even be ahead (No one's said to the contrary).

You mean like the Phenom II X4 with the 2.6GHZ NB in the link I posted earlier:

http://forums.anandtech.com/showpost.php?p=34718546&postcount=4

No if and buts.

That is the first version.

BTW,most of my mates play SC2 and WoW. The FX CPUs even at stock showed better performance(minimums especially) against Phenom II X4 and overclocked Athlon II X4 CPUs.

I harranged them to benchmark,but none of them are enthusiasts which was annoying.

 

[Martini1991]

You've quoted my edit, which was basically a snip saying my AMD CPU was far from the norm (Which is probably why I find AMD's PD so unimpressive to be perfectly honest, I just can't get excited over performance I had years prior).

I'm not aware of anyone who claims PD's always slower, anyone who knows anything should know better (Which is why I find Humbugs post somewhat confusing, as self professed AMD fan, he *should* know better. Again, I take this approach because I'm always called a fanboy, but ironically seem to be more clued up on both the Intel and AMD platforms than my accusers, so I take "be a ***** stance".)

 

[CAT-THE-FIFTH]

You've quoted my edit, which was basically a snip saying my AMD CPU was far from the norm (Which is probably why I find AMD's PD so unimpressive to be perfectly honest, I just can't get excited over performance I had years prior).

I'm not aware of anyone who claims PD's always slower, anyone who knows anything should know better (Which is why I find Humbugs post somewhat confusing, as self professed AMD fan, he *should* know better)

Plenty of people say its always slower though. The problem is the FX8150 launch,where AFAIK it tended to be. The thing is though that years later,we had PD which did improve performance(in some titles like DiRT the BD CPUs were bugged so there were big improvements),and basically there is better optimisations now.

Now,one could make the argument that maybe no one is really optimising for the Phenom II CPUs now which is another factor,which does help the FX CPUs.

I only went by SC2 and WoW since these are titles which don't give the FX CPUs an unfair advantage in threads.

Not saying the FX always has better IPC than the Phenom II CPUs,either but it kind of all needs to be taken on a case by case basis now.

 

[Martini1991]

You mean like the Phenom II X4 with the 2.6GHZ NB in the link I posted earlier:

http://forums.anandtech.com/showpost.php?p=34718546&postcount=4

No if and buts.

That is the first version.

BTW,most of my mates play SC2 and WoW. The FX CPUs even at stock showed better performance(minimums especially) against Phenom II X4 and overclocked Athlon II X4 CPUs.

I harranged them to benchmark,but none of them are enthusiasts which was annoying.

That isn't technically a heavily clocked Phenom II, even the first Deneb I touched I ran higher than that , and doesn't that say what I said? The results change?

But just the stock default settings of the FX-6300 were able to pretty much match the heavily overclocked Phenom II

Whereas your GameGPU results showed the FX6300 being able to make a decent break stock versus stock.

I can never tell if you're agreeing or trying to argue a case to be perfectly honest.

 

[CAT-THE-FIFTH]

That isn't technically a heavily clocked Phenom II, even the first Deneb I touched I ran higher than that , and doesn't that say what I said? The results change?



Whereas your GameGPU results showed the FX6300 being able to make a decent break stock versus stock.

I can never tell if you're agreeing or trying to argue a case to be perfectly honest.

If a Phenom II X4 with a higher than average CPU-NB overclock cannot beat a FX6300 running at similar(or lower clockspeed),then its highly unlikely the Phenom II X6 with slightly lower IPC(even with a bigger CPU-NB overclock) would massively changed it.

The first is the original which was launch in 2010.

GameGPU was the latest expansion last year:

http://gamegpu.ru/rts-/-strategii/starcraft-ii-heart-of-the-swarm-test-gpu.html

 

[Martini1991]

The first is the orginal which was launch in 2010.

GameGPU was the latest expansion last year:

http://gamegpu.ru/rts-/-strategii/starcraft-ii-heart-of-the-swarm-test-gpu.html

I'm aware?

 

[Martini1991]

If a Phenom II X4 with a higher than average CPU-NB overclock cannot beat a FX6300 running at similar(or lower clockspeed),then its highly unlikely the Phenom II X6 with slightly lower IPC(even with a bigger CPU-NB overclock) would massively changed it.

The first is the original which was launch in 2010.

GameGPU was the latest expansion last year:

http://gamegpu.ru/rts-/-strategii/starcraft-ii-heart-of-the-swarm-test-gpu.html

1.) It's not a higher than average CPU-NB, it's a lower than average.
2.) I didn't mention beating, I said change the results (You know, the FPS figures, this isn't a turn around either, as I've clearly stated, this doesn't contradict that a PD can either be parity or beat the 2.8GHZ-3GHZ CPU NB Phenom II) I'm not even mentioning Thuban, merely Phenom II.

You're trying to argue with me (Or it seems that way) by saying what I'm saying (But not actually reading what I'm saying).

 

[humbug]

The CPU-NB adds a lot of heat.

I could get my P-II x6 to run at about 4.1 to 4.2Ghz with a 2800Mhz CPU-NB (for benching)

In Games that use the CPU to its full extent it was not completely stable anything past 4Ghz + 2400Mhz CPU-NB, Crysis 3, BF4 and Thief in Mantle after an hour and a half the computer would freeze, i could cure that with lashings of volts, 1.47v (70mv over 4Ghz) but then it would get too hot and eventually shut down. and mine is a pretty decently binned Thuban.

For a completely stable trouble free computer you couldn't run it over 4Ghz without some water on it.

 

[CAT-THE-FIFTH]

I'm aware?

1.) It's not a higher than average CPU-NB, it's a lower than average.
2.) I didn't mention beating, I said change the results (You know, the FPS figures, this isn't a turn around either, as I've clearly stated, this doesn't contradict that a PD can either be parity or beat the 2.8GHZ-3GHZ CPU NB Phenom II) I'm not even mentioning Thuban, merely Phenom II.

You're trying to argue with me (Or it seems that way) by saying what I'm saying (But not actually reading what I'm saying).

That is an overclocked Phenom II X4 with an overclocked CPU-NB(a 30% overclock for the CPU-NB IIRC),which makes it faster than a Phenom II X4 980BE. So even if you increased the clockspeed to 2.8GHZ(which would be a 40% overclock),I don't see it being massively better. Yet a stock FX6300 is not loosing to it and they are both at similar clockspeeds. This was in an old 890GX motherboard too,in a game that uses three threads.

The first set of results are for SC2 which was launched in 2010,before BD or PD.

The second set of results are for SC2:HoTS launched in 2013,after BD and PD.

 

[Martini1991]

You answered your own question. The first is the original game showing that clock for clock a Phenom II cannot really beat something like an FX6300.

The second is the expansion which came out three years later,showing the FX CPUs pulling away.

You're not actually reading what I'm saying, either that, or you're not understanding it.

That is an overclocked Phenom II X4 with an overclocked CPU-NB(a 30% overclock for the CPU-NB IIRC),which makes it faster than a Phenom II X4 980BE. So even if you increased the clockspeed to 2.8GHZ(which would be a 40% overclock),I don't see it being massively better. Yet a stock FX6300 is not loosing to it and they are both at similar clockspeeds. This was in an old 890GX motherboard too,in a game that uses three threads.

The first set of results are for SC2 which was launched in 2010,before BD or PD.

The second set of results are for SC2:HoTS launched in 2013,after BD and PD.

Haven't said anything to the contrary of any of that.
Again, Phenom II had dated instruction sets even when it launched, I would have absolutely no problem believing that a PD can maintain parity/capture a lead in either SC2, or its expansion. But 1.) The first results show an overclocked Phenom II against a stock FX6300, the FX6300 has parity/small win. This is in the original SC2.

However, your second example shows the FX6300 walking away. Ignore that it's SC2's expansion right now, the comparison is against a stock Phenom II. The results for the Phenom II would get higher as it has its CPU -NB overclocked, against, that's where my "No ifs or buts come in", regardless of the final result.

Only once you've got equal testing parameters can you talk about the expansion and further optimizations that benefit PD, not beforehand.

 

[Martini1991]

The CPU-NB adds a lot of heat.

I could get my P-II x6 to run at about 4.1 to 4.2Ghz with a 2800Mhz CPU-NB (for benching)

In Games that use the CPU to its full extent it was not completely stable anything past 4Ghz + 2400Mhz CPU-NB, Crysis 3, BF4 and Thief in Mantle after an hour and a half the computer would freeze, i could cure that with lashings of volts, 1.47v (70mv over 4Ghz) but then it would get to hot and eventually shut down. and mine is a pretty decently binned Thuban.

For a completely stable trouble free computer you couldn't run it over 4Ghz without some water on it.

On these forums, I was one of the first people to run a Phenom II X4 965 at 4GHZ completely stable on air, I also did it with a launch day 1055T (The 125W one!) on air at 4025MHZ. Your chip's not as good as you think it is.

They were both on a T.R.U.E Black Rev C.

They were in the Phenom II overclocking thread.

 

[CAT-THE-FIFTH]

You're not actually reading what I'm saying, either that, or you're not understanding it.

However, your second example shows the FX6300 walking away. Ignore that it's SC2's expansion right now, the comparison is against a stock Phenom II. The results for the Phenom II would get higher as it has its CPU -NB overclocked, against, that's where my "No ifs or buts come in", regardless of the final result.

Only once you've got equal testing parameters can you talk about the expansion and further optimizations that benefit PD, not beforehand.

The Phenom II X4 was running at the same clockspeed as the FX6300 and it had a CPU-NB overclock. I understand what you are saying about CPU-NB overclocking improving performance,but even with a Phenom II X4 with such an overclock,its still no better in the original SC2.

That overclocked Phenom II X4 is better than the highest bin Phenom II X4 SKU,the Phenom II X4 980BE.

Now lets look at HoTS.

The Phenom II X6 1100T is running at between 3.3GHZ to 3.7GHZ at 34FPS minimums and 48FPS average.

An FX8350 runs at between 4.0GHZ to 4.2GHZ produces 50FPS minimums and 71FPS average.

If we assume the Phenom II X6 Turbo is not working,its running at 3.3GHZ and the FX8350 at 4.2GHZ,which gives the FX8350 a 27% clockspeed advantage.

Even if you were to clock the Phenom II X6 to 4.2GHZ,I still see the FX8350 having a 16% improvement overall.

Could massively overclocking the Phenom II X6 CPU-NB help,tilt the balance?? That is still a big gap.

It also does not change the fact that even at similar clockspeed,when at stock the FX is still clock for clock faster in the game,at least going by GameGPU figures.

It even gets worse for sections of WoW(latest expansion section):

http://media.bestofmicro.com/Z/2/357662/original/world-of-warcraft-1680.png

Anandtech places the FX and Phenom II X4 CPUs broadly similar in the section they tested(not the latest expansion it seems and probably lacks optimisations).

Thats two instances of a new expansion in a Blizzard game improving performance on the FX CPUs.

You can ignore it all you like,but that is TWO new Blizzad expansions,which show this. No ifs or buts come in" at all.

Especially if running at stock,they will be a good upgrade for someone with an earlier Athlon II and Phenom II,especially if they won't overclock and have a compatible board running those games.

Intel is better OFC in both of them,although you should still be able to get decently playable performance,IMHO OFC.

 

[Martini1991]

Believe me cat, I'm understanding what you're saying fine.

Read my edit, then either nod and agree, or shake your head and disagree, and we'll leave it there, I have no interest in a semantic based argument, which is likely what this will turn into.

 

[humbug]

On these forums, I was one of the first people to run a Phenom II X4 965 at 4GHZ completely stable on air, I also did it with a launch day 1055T (The 125W one!) on air at 4025MHZ. Your chip's not as good as you think it is.

They were both on a T.R.U.E Black Rev C.

They were in the Phenom II overclocking thread.

here are 9 hours @ 4Ghz Prime95 on a CM Hyper 412S, that was 2 years ago on the now dead 1090T, and good Temps.



This one i made sure is from the same year (2010) they were the best ones, its just as well binned as my old one was.

I could run it @ 4.2Ghz all day long, even pass 2 hours of Prime95 at that

That was until Crysis 3 came along, then BF4, then Thief... not all things are equal, those, especially BF4 and Thief in Mantle find something wrong with the CPU making it need more volts, at which point you need better cooling.

To be frank i doubt any of your Phenom II's would have survived those games any better than mine does.

 

[Martini1991]

Those aren't good temperatures based on your voltage though, bear in mind that 49C isn't actually accurate.
My CPU temp never got past 62C (Because I didn't allow it) when I overclocked a 125W chip.]

I can't find any actual screenshots for my 125W Thuban, which is a shame. I do have a 4.25GHZ with 3GHZ CPU-NB 8 hour Prime screenshot like, voltage is like 1.525v (Because LLC)

I'm pretty confident my CPU would have coped.

 

[CAT-THE-FIFTH]

Believe me cat, I'm understanding what you're saying fine.

Read my edit, then either nod and agree, or shake your head and disagree, and we'll leave it there, I have no interest in a semantic based argument, which is likely what this will turn into.

I understand your saying a clocked Phenom II X6 with a big CPU-NB overclock might change what we see. Its that I am not sure if it will make a huge difference in the games I mentioned after seeing that guy trying it with his Phenom II X4 with such an overclock applied(yes,I know the Phenom II X6 could hit a higher overclock due to the memory controller improvements). Taking the WoW results into consideration,that is now two new Blizzard expansions which show this.

Maybe,we need to agree to disagree.

 

[humbug]

Those aren't good temperatures based on your voltage though, bear in mind that 49C isn't actually accurate.
My CPU temp never got past 62C (Because I didn't allow it) when I overclocked a 125W chip.]

I can't find any actual screenshots for my 125W Thuban, which is a shame. I do have a 4.25GHZ with 3GHZ CPU-NB 8 hour Prime screenshot like, voltage is like 1.525v (Because LLC)

I'm pretty confident my CPU would have coped.

what cooler?

 

[Martini1991]

what cooler?

I think that was under water.

But that's when I allowed temps to get hotter on the CPU socket, that was my 95W chip.

My 125W chip would do 4025MHZ at lower temps than your screenshot, it's a shame I have none, like I say, I did not allow it to get to 62C.

It was a T.R.U.E black rev C I used back then.

With AMD, a lot of air coolers had naff mounts back then, so the end performance isn't as good as it should be. The Rev C never had that problem.

Either way, those settings for my Thuban at 4250MHZ would have had no problem gaming away.

And I think we've all seen where I made my cooling better ;

It is a shame I never took a screenshot of my final settings for the CPU, but hey ho.

 

[CAT-THE-FIFTH]

According to HWBOT the Phenom II 1090T and 1100T on average hit around 4.1GHZ:

http://hwbot.org/hardware/processor/phenom_ii_x6_1090t_be/
http://hwbot.org/hardware/processor/phenom_ii_x6_1100t_be/

The Phenom II X16 1055T around 3.9GHZ:

http://hwbot.org/hardware/processor/phenom_ii_x6_1055t/

OTH,being HWBOT it does not mean 24/7 clockspeeds,due to the nature of the site.