Return of AMD FX: My OC'd AMD FX 8150 review with OC'd 6990 - Daily Results!!

[Abwx]

Don't join your bud there, for logic and reasoning will destroy your "argument". Let's have real FX numbers and testing, but lets add 8-10% to i7 920's scores to make accurate comparisons, eh? Laughable.

So if frequencies are different , then
it is skewed ??..

Look , this slide is totaly skewed according to you...
Ever compained to Anand about it ?..

 

[chimaxi83]

No, the skew comes from testing overclocked hardware against stock clocked hardware. Nice try, but don't be dense.

That chart is not skewed because the hardware being tested is all at stock retail clocks. Very easy to make comparisons in that case when the person doing the testing knows what they're doing and can keep some semblance of hardware continuity between tests, like Anand did. Here's the hardware used in the slide you decided to post:

Motherboard:
ASUS P8Z68-V Pro (Intel Z68)
ASUS Crosshair V Formula (AMD 990FX)
Intel DX79SI (Intel X79)

Hard Disk:
Intel X25-M SSD (80GB)
Crucial RealSSD C300

Memory:
4 x 4GB G.Skill Ripjaws X DDR3-1600 9-9-9-20

Video Card:
ATI Radeon HD 5870 (Windows 7)

Video Driver:
AMD Catalyst 11.10 Beta (Windows 7)

Desktop Resolution:
1920 x 1200
OS:
Windows 7 x64

The skewed and biased results in this thread come from a load of different CPU's, GPU's, and clocks, with user polzyp comparing his highly overclocked hardware to stock hardware the majority of the time. Unlike the review from which you took that slide, this thread doesn't maintain any continuity. The testing methodology is... non existent. The hardware changes with almost every test.

It's pretty easy to see, so easy even a caveman can do it. If you can't see that, then you probably don't need to be in this discussion if all you're doing is trying to defend these results with your broken argument.

 

[Abwx]

Hardware sites always test overclocks on CPUs and then show
what is gained in perfs in comparison to stock and other overclocked
CPUs and it doesnt create the fuss that i can see by there..

Whatever his methodology , he didnt change the numbers , they
are actually accurates , so it is not skewed at all , contrary to your
rethoric that try to make some twisted amalgam.

As a side note , i would add that it happens that BD at sock speed
has trouble competing with a 2600K performance wise, that s quite obvious ,
but it also happen than although being late in IPC , it can often be quite
obviously better than a 2500K as soon as the software is threads heavy ,
so the 2500K buyers did not really make the best choice , they better
would have go 2600K....

And it just happens that most if not all of the threads trolls trashers
in this thread are 2500K owners...
Talk of how miserable and ridiculously vanitous are people
under guise of technological discussion...

 

[Haserath]

Hardware sites always test overclocks on CPUs and then show
what is gained in perfs in comparison to stock and other overclocked
CPUs and it doesnt create the fuss that i can see by there..

Whatever his methodology , he didnt change the numbers , they
are actually accurates , so it is not skewed at all , contrary to your
rethoric that try to make some twisted amalgam.

As a side note , i would add that it happens that BD at sock speed
has trouble competing with a 2600K performance wise, that s quite obvious ,
but it also happen than although being late in IPC , it can often be quite
obviously better than a 2500K as soon as the software is threads heavy ,
so the 2500K buyers did not really make the best choice , they better
would have go 2600K....

And it just happens that most if not all of the threads trolls trashers
in this thread are 2500K owners...
Talk of how miserable and ridiculously vanitous are people
under guise of technological discussion...

Do you believe this review was a fair comparison?

 

[gallag]

try not to take this guy to seriously, read his posts over at amd zone.

" I guess some people just take it personal when they see their intel chips not fairing so well for once

."

" The moderator that closed the post said it was on the basis that I did not admit intel has the better chip"
IS THIS TRUE?

"the majority of people on AT appreciated my work, it was really a select few people (three) who has a problem with my findings. O well, my most popular forum post closed, but I like it here over there any day !

you guys are the sh*t ^^"

"People seem to be especially upset with my DIRT 3 / AVP results, because Dozer does so well when compared to lower clocked intel counterparts. Some argue the 990x is too old to compare to and therefor it is unfair, but then they argue that FX is worse than core2quad, how ironic. I love it when people accuse me of fabricating results when in reality FX is kicking butt

.

"

"Ill be revisiting DIRT 3 by request @ 4.0 Ghz.. give the intel fans a "fair fight"

"

***even his context here is very us vs them.***

"People seem to want me to use 5 year old benchmarks to show how FX isnt go good, so to be fair I will include some. And I will be doing some explanation why performance is sub par."

***i.e if intel wins i will explane it away lol***

 

[BallaTheFeared]

I don't have a problem with his findings, he's not walking away from my i5-2500k in any tests, and in many of them he's getting smoked. The tests he uses are selected because they perform well for Bulldozer.

The game tests are actually just gpu tests, while truecrypt, encoding, and rendering all play into bulldozers eight core design.

I don't think anyone should be surprised Bulldozer is fairing well in some hand picked benchmarks, comparing it to a lower clocked i5-2500k is even fine, it only takes me a few minutes to beat or nearly match his 8150 with a cheaper cpu that has a massive advantage at all the less threaded workloads he doesn't even attempt to compare.

You can see in the few tests where he talked the most about beating the i5-2500k I quickly came in with results that either beat his, or reduced the difference substantially. In Fritz and Cinebench my i5-2500k "easily" beat his 8150 results, while in 7zip i shrunk is "100%" down to just a few small percent difference... to the tune of about 10%.

He does use misinformation to his advantage, an example of this was the same 7zip bench where he states the 3.7GHz turbo i5, when everyone knows with four cores loaded turbo is 3.4GHz.

But he took what I posted well, and accepted it without much fuss. I respect that aspect of his posting, and I can appreciate how hard he's trying to show his favorite company in better light. It's your job as a reader to pick through what he's saying and discern for yourself what is real and what is shrouded in misleading information. We already know Bulldozer does better when GPU limited, getting a few more FPS than Intel systems do. AMD knows this too, that's why AMD marketed Bulldozer in situations where it would be gpu bottlenecked. This user is aware of that too, that's why he used AvP, Heaven, and 3DMark which are all gpu limited tests with no real cpu workload for rendering. The only exception there is 3DMarks combined, where the i5-2500k is under performing considerably while the the 8150 still lags behind drastically.

 

[grkM3]

it looks like the op used the fastest ram also.i can bench my 2600k wuth 2200ddr3 at 4.8 if you guys want and i even have the amd patch on my system that boosted my multy threaded benching also.

I get 26500 memory bandwith in aida64 and can break 10.40 in cinebench at 5.22ghz.

 

[polyzp]

Hardware sites always test overclocks on CPUs and then show
what is gained in perfs in comparison to stock and other overclocked
CPUs and it doesnt create the fuss that i can see by there..

Whatever his methodology , he didnt change the numbers , they
are actually accurates , so it is not skewed at all , contrary to your
rethoric that try to make some twisted amalgam.

As a side note , i would add that it happens that BD at sock speed
has trouble competing with a 2600K performance wise, that s quite obvious ,
but it also happen than although being late in IPC , it can often be quite
obviously better than a 2500K as soon as the software is threads heavy ,
so the 2500K buyers did not really make the best choice , they better
would have go 2600K....

And it just happens that most if not all of the threads trolls trashers
in this thread are 2500K owners...
Talk of how miserable and ridiculously vanitous are people
under guise of technological discussion...

This guy knows what Im trying to do.. alot of people really not reading what I am even typing to respond to the allegations of skewing results, so I wont even bother responding again. I feel like a broken record. Its very obvious that I can compare whatever CPU i like to my results, it doesnt make the results skewed.. Do your own google comparisons if you dont like it! Simple as that haha. And if you find any sandy bridge 4.8 ghz comparisons with a 6990, then please post them so I can use these instead. Intel fanboys need to stop whining about me comparing to a 3.6 ghz 980x, when this is a thousand dollar CPU and so what if its clocked to 3.6? Tell that the the reviewers who didnt oc it higher, not me .. haha If it were up to me I would compare everythng to an equally clocked 2500k or 2600k, but the data isnt there.

 

[polyzp]

I don't have a problem with his findings, he's not walking away from my i5-2500k in any tests, and in many of them he's getting smoked. The tests he uses are selected because they perform well for Bulldozer.

The game tests are actually just gpu tests, while truecrypt, encoding, and rendering all play into bulldozers eight core design.

I don't think anyone should be surprised Bulldozer is fairing well in some hand picked benchmarks, comparing it to a lower clocked i5-2500k is even fine, it only takes me a few minutes to beat or nearly match his 8150 with a cheaper cpu that has a massive advantage at all the less threaded workloads he doesn't even attempt to compare.

You can see in the few tests where he talked the most about beating the i5-2500k I quickly came in with results that either beat his, or reduced the difference substantially. In Fritz and Cinebench my i5-2500k "easily" beat his 8150 results, while in 7zip i shrunk is "100%" down to just a few small percent difference... to the tune of about 10%.

He does use misinformation to his advantage, an example of this was the same 7zip bench where he states the 3.7GHz turbo i5, when everyone knows with four cores loaded turbo is 3.4GHz.

But he took what I posted well, and accepted it without much fuss. I respect that aspect of his posting, and I can appreciate how hard he's trying to show his favorite company in better light. It's your job as a reader to pick through what he's saying and discern for yourself what is real and what is shrouded in misleading information. We already know Bulldozer does better when GPU limited, getting a few more FPS than Intel systems do. AMD knows this too, that's why AMD marketed Bulldozer in situations where it would be gpu bottlenecked. This user is aware of that too, that's why he used AvP, Heaven, and 3DMark which are all gpu limited tests with no real cpu workload for rendering. The only exception there is 3DMarks combined, where the i5-2500k is under performing considerably while the the 8150 still lags behind drastically.

This point on turbo is actually THE ONLY good point I have seen from anyone yet who has criticism of my findings. Ill have to change this, I will change this. Everyone else is uselessly trying to pick apart my review, without any solid points whatsoever. The Ghz value thing is gettting old btw, I never said my review is the end all and be all for performance evaluation. If I had l2n results at 6.5 ghz, I could STILL compare to stock intel if I like, because you get a % comparison between OC and a stock intel... That doesnt mean its unfair... Its nothing more than showing a 6.5 ghz FX performs X against a 3.3 ghz intel.. So what? If you read through my review, I never say anywhere than AMD FX is definitly better than any of its intel counterparts hands down for all overclockes no matter what the GHz, because this would be wrong. In reality my results just show what they show, take them as is, nothing more.

 

[polyzp]

try not to take this guy to seriously, read his posts over at amd zone.

" I guess some people just take it personal when they see their intel chips not fairing so well for once

."

" The moderator that closed the post said it was on the basis that I did not admit intel has the better chip"
IS THIS TRUE?

"the majority of people on AT appreciated my work, it was really a select few people (three) who has a problem with my findings. O well, my most popular forum post closed, but I like it here over there any day !

you guys are the sh*t ^^"

"People seem to be especially upset with my DIRT 3 / AVP results, because Dozer does so well when compared to lower clocked intel counterparts. Some argue the 990x is too old to compare to and therefor it is unfair, but then they argue that FX is worse than core2quad, how ironic. I love it when people accuse me of fabricating results when in reality FX is kicking butt

.

"

"Ill be revisiting DIRT 3 by request @ 4.0 Ghz.. give the intel fans a "fair fight"

"

***even his context here is very us vs them.***

"People seem to want me to use 5 year old benchmarks to show how FX isnt go good, so to be fair I will include some. And I will be doing some explanation why performance is sub par."

***i.e if intel wins i will explane it away lol***

I remember in my old post here on AT people said something about me working for AMD because I hadnt posted on AMDzone, when in reality I just missed it, and to date I have posted there. And its true, alot of the benchmarks showing amd in worst light are quite old.

Take me as seriously as you like, I am just a regular person doing a comparison review.

And yes i felt like my thread was closed for inappropriate reasons.

How can a topic like this not be "us vs. them". At the end of the day even if i never intended it to be that, it will become that naturally.

 

[grkM3]

here is 7zip and its the first time I ever ran it.I dont even know how long it runs but I let it do 11 passes and stopped it.

I have no idea what numbers to post up.

 

[Keysplayr]

One moment.

I'm counting 20 sites this was cut and pasted to by Panos. "Polyzp".

 

[BallaTheFeared]

Heh

 

[grkM3]

From my own comparing a 2600k at 4.9 will beat a maxed out run bd and tie it at 4.8 in zip 7

For fun i benched zip 7 at 5.2 and broke 31k but system crashed as i was saving it.

I should run a bunch of maxed out 5.3ghz 2600k benches and make a review vs a stock clocked bd and the one at 4.2 etc lol.

Cant wait for his reviews agains ivy bridge in a few months

 

[polyzp]

One moment.

I'm counting 20 sites this was cut and pasted to by Panos. "Polyzp".

for a link to all the forums i have posted on visit my blog, its on the front page. Nothing to hide here :\ . Whats so wrong about opening discussion? Its interesting to see the level of intel fanboyism :wub:, and amd fanboyism :wub: across various sites to say the least.

 

[MrTeal]

I really don't have a problem with you comparing an FX to an i7 at different clocks. The 8150 obviously ships with higher stock clocks than a 2500k or 2600k, and if it overclocks to 4.8GHz great on you, it's interesting to see what kind of scaling you get.

The biggest problem I see with your tests is that you're grabbing graphs off the internet, running your system on the same test, and then posting up results pasted in there as if they're comparable. There is a reason that a hardware site link AT has a consistent test bench. In GPU2012, each of those graphics cards is benched on the same system, with the GPU (and drivers) being the only independent variable. That's what you should strive for. With the testing your doing, you're not running even close to the same hardware in most cases. Anything that's memory bandwidth sensitive on your system would show an increase due to running CL7 - 2200MHz. While most good sites bench with an SSD these days, most probably aren't as fast as your Revodrive 3 X2. Even if you went out and bought hardware to match theirs, you can't guarantee that your settings are exactly the same for every test.

These simply aren't really useful comparisons. As an enthusiast your budget is probably limited, but I think you would hear much less grumbling is you had a 2500k and decent Z68 board, and benched against it at max OC with all the same hardware and settings and just presented those two results. Your sample would obviously be smaller, but it would be a much more valid test than running against 3 year old CPUs with undetermined other hardware just because that's what you could find a graph for.

 

[Rvenger]

for a link to all the forums i have posted on visit my blog, its on the front page. Nothing to hide here :\ . Whats so wrong about opening discussion? Its interesting to see the level of intel fanboyism :wub:, and amd fanboyism :wub: across various sites to say the least.

The only fanboyism you are going to get is AMDzone. Trust me if AMD came out with a CPU that put Intel's in the dust, at least 90% of the enthusiasts on this forum would switch to it. AMD could release an AMD K6 and badge it as a piledriver and AMDzone will say its faster than a Haswell.


Bulldozer, clock for clock is significantly slower than SB no matter how you put it. My Celeron G530 'felt' faster than the FX6100 in my sig right now, at almost half the clock speed.

 

[exar333]

The only fanboyism you are going to get is AMDzone. Trust me if AMD came out with a CPU that put Intel's in the dust, at least 90% of the enthusiasts on this forum would switch to it. AMD could release an AMD K6 and badge it as a piledriver and AMDzone will say its faster than a Haswell.


Bulldozer, clock for clock is significantly slower than SB no matter how you put it. My Celeron G530 'felt' faster than the FX6100 in my sig right now, at almost half the clock speed.

To be honest, if AMD was within 5% across the board of Intel's performance, I would suggest a very good % of enthusiasts would switch to AMD. I probably would myself. I like to support AMD, if they have a competitive product.
Look back to the P4/AXP days; MANY people here all had AXP rigs, even if the P4 was better in some applications.

 

[Idontcare]

The only fanboyism you are going to get is AMDzone. Trust me if AMD came out with a CPU that put Intel's in the dust, at least 90% of the enthusiasts on this forum would switch to it. AMD could release an AMD K6 and badge it as a piledriver and AMDzone will say its faster than a Haswell.


Bulldozer, clock for clock is significantly slower than SB no matter how you put it. My Celeron G530 'felt' faster than the FX6100 in my sig right now, at almost half the clock speed.

There are two kinds of fanboys, those who are "pro" their favorite team and then those who are "anti" a specific team.

This is where the phrase "haters gonna hate" comes.

There is nothing wrong with being Pro-AMD or Pro-Intel. Those fans look to the positive in their favorite gear without needing to resort to knocking on the competition to make them feel better about their own personal preference.

The challenge in being an Anti-AMD or an Anti-Intel fanboy is that it requires a fair amount of negativity and this tends to come in the form of thread-crapping in threads of people who are Pro-AMD or Pro-Intel.

And that is when the confluence of Pro-fanboy and Anti-fanboy results in the kind of excessive thread crapping and baiting that happened in the last (the original) "Return of AMD FX" thread.

Mixing Pro and Anti fanboys is an explosive affair, but it usually just generates a lot of waste heat in the process, very little of the energy is harnessed for productive purposes.

It is threads like this where the "Anti" fanboys are best off just staying out of the thread so they avoid thread-crapping on the "Pro" fanboy's parade.

Personally I have thoroughly enjoyed my Intel 2600K. I am very much Pro-Intel with this 2600K. But I am not Anti-AMD. If I had a bulldozer here in my house I'd probably enjoy torturing it like I did my 2600K.

I have nothing against bulldozer, and nothing against people who like bulldozer. I just happen to enjoy this 2600K for now because I have one and it meets all my needs. For all I know bulldozer would leave me feeling the same way if I had one. So why knock it if I haven't tried it?

 

[Rvenger]

To be honest, if AMD was within 5% across the board of Intel's performance, I would suggest a very good % of enthusiasts would switch to AMD. I probably would myself. I like to support AMD, if they have a competitive product.
Look back to the P4/AXP days; MANY people here all had AXP rigs, even if the P4 was better in some applications.

Yep same here. I owned about 5 different AXP CPUs. I still have an old barton based Sempron at home somewhere. I bought my first intel P4 back in 2002, and I couldn't get it to stop bluescreening when I overclocked it. The CPU was dead slow when it was at stock speed so I just sold it off and bought another AXP lol.

 

[AtenRa]

My Celeron G530 'felt' faster than the FX6100 in my sig right now, at almost half the clock speed.

Yeap, my quad threaded ATOM D2700 felt faster than Celeron G530 [/sarcasm]

 

[polyzp]

There are two kinds of fanboys, those who are "pro" their favorite team and then those who are "anti" a specific team.

This is where the phrase "haters gonna hate" comes.

There is nothing wrong with being Pro-AMD or Pro-Intel. Those fans look to the positive in their favorite gear without needing to resort to knocking on the competition to make them feel better about their own personal preference.

The challenge in being an Anti-AMD or an Anti-Intel fanboy is that it requires a fair amount of negativity and this tends to come in the form of thread-crapping in threads of people who are Pro-AMD or Pro-Intel.

And that is when the confluence of Pro-fanboy and Anti-fanboy results in the kind of excessive thread crapping and baiting that happened in the last (the original) "Return of AMD FX" thread.

Mixing Pro and Anti fanboys is an explosive affair, but it usually just generates a lot of waste heat in the process, very little of the energy is harnessed for productive purposes.

It is threads like this where the "Anti" fanboys are best off just staying out of the thread so they avoid thread-crapping on the "Pro" fanboy's parade.

Personally I have thoroughly enjoyed my Intel 2600K. I am very much Pro-Intel with this 2600K. But I am not Anti-AMD. If I had a bulldozer here in my house I'd probably enjoy torturing it like I did my 2600K.

I have nothing against bulldozer, and nothing against people who like bulldozer. I just happen to enjoy this 2600K for now because I have one and it meets all my needs. For all I know bulldozer would leave me feeling the same way if I had one. So why knock it if I haven't tried it?

nice post! Yeah i really dont want this thread to turn into what the last one did.

I have said about ten times now throughout the internet that people need to take my results with a grain of salt, as they should with ANY results or review they see online.

By the time my review is finished it will have more data points on Bulldozer's performance than any other review.

Alot of the benchmarks I have done simply have not been done before (look for yourself) and posted online. (Given ram speeds, gpu clocks, cpu clocks, etc)

 

[polyzp]

So far my favourite piece of data i have found is,

Here we see that the FX Patch brings a decent boost in performance, averaging +5.17% across all 12 tests. The largest performance increase comes in the .NET Arithmetic - Dhrystone test, where we see a +24.6% difference.

 

[polyzp]

This is actually my first AMD rig, up until now I was all intel. I have never been happier about an upgrade.

 

[dguy6789]

I'm gonna finish writing up my review comparing a 3.8Ghz P4 to a 1.8Ghz Athlon 64 and post it on the forums in a few days. You guys won't believe the results; the P4 is faster. It's crazy isn't it? It shows that everyone who thought the Athlon 64 was way better than the Pentium 4 thought wrong.