AMD CPU speculation... and expert conjecture

[mayankleoboy1]

Cinebench has been cheating benchmarks scores for many years now. I am not worried about if one piece of software runs better in a chip or in other but when the benchmarks cheat and give fakes scores. Can you confirm that Maxon is using Intel compiler for its programs? I.e. can you confirm me that Cinebench benchmarks really represent Cinema4D performance?

I cant confirm, and neither can you that Cinebench cheats against AMD. That is the problem with closed source software. And nobody can say whether a benchmark represents the perf of the actual software. So if you are denouncing Cinebench, you will have to denounce all other closed source benchmarks as well.

Yes Maxon claims in its website that Cinebench is based in professional software Cinema4D, "which is used extensively by studios and production houses worldwide for 3D content creation", but since Titanic was produced using Linux in 1997 all major studios including Dreamworks Animation, Pixar, Weta Digital, and Industrial Light & Magic have migrated to Linux. Currently more than 95% of the servers and desktops at large animation and visual effects companies use Linux, which imply that 95% of the industry is not using Cinema4D.

Thats a big claim you are making. Got any link to back it up ?

I am happy with open benchmarks or with closed benchmarks compiled with a fair compiler. I just reject the well-known unfair closed benchmarks that cheat about the scores giving fake advantage to Intel chips.

If the benchmarks are closed source, how do you know they cheat against AMD, or are unfair ? You are just making that up from your idea of what an AMD CPU should perform like.

I repeat, if Intel chips are so superior why do the biased benchmarks exist?

Because marketing. And AMD does not have the money to influence benchmarks.

As a final note. AMD already introduced Steamroller optimization in the main compilers as GCC. When SteamRoller was released compilers will use its potential since day one.

That has been the case with Bulldozer and Piledriver as well. Usually, compiler support is added months before actual product is released.

 

[jdwii]

MSI GX70 3BE: Richland A10 APU and Neptune 8970M Gaming Notebook
http://www.anandtech.com/print/6949/msi-gx70-3be-richland-a10-apu-and-neptune-8970m-gaming-notebook

nvidia's latest gpu scam exposed? 770 is the new 6950? amd calls nvidia out? charlied is rolling on the floor laughing hysterically(roflH)? as usual, take it with a grain of salt.
http://semiaccurate.com/2013/05/13/you-can-upgrade-an-nvidia-gtx680-to-a-gtx770-with-a-simple-bios-flash/

If true how is it like the 6950(which was only flashable to a 6970 and not all of them or mine) A 6950 is using a newer design compared to the 6800,5800 series as well.

For Nvidia to do this to a high-end part is sad and even makes me know more that their rushing this 700 series vs Amd were their design is actually going to be GCN 2.0

 

[truegenius]

Once I put the Sabertooth and a 1000 W 80+ Gold PSU in I was able to hit a stable 5.0Ghz overclock on my "old crappy" 965 BE!!

Your going to love the Sabertooth 990FX R2.0!! Keep in mind if you don't have adequate airflow watch your overclocks till you get a better case and cooling (to hit 5.0Ghz on my 965 BE I had to set the vcore to 1.6v (well above the recommended 1.55v max) and was hitting 61C - 62C is max for Phenom II).

the numbers mason, i know when you are lying

5GHz it is not possible to hit stable 5GHz on 965

 

[mayankleoboy1]

http://www.phoronix.com/scan.php?page=news_item&px=MTM3MTA

Anyone heard of "Hainan" GPU from AMD, a.ka. HD8800 ?
AMD has started adding code to the open source AMD drivers.

 

[JAYDEEJOHN]

On the Anand link, it has the "new" 8970m, and they go on to say:
"The bad news then is that a 10% clock speed increase from Trinity isn't going to be enough to close the gap in many titles, depending on the resolution and quality settings. Looking at Trinity vs. Ivy Bridge with 7970M, I've seen Intel outperform AMD by 50% or more, particularly in titles that pound the CPU (e.g. Skyrim and StarCraft II); on other games, however, it's basically a wash at high quality 1080p settings, so as a more budget-friendly gaming notebook the GX70 has potential."
http://www.anandtech.com/print/6949/msi-gx70-3be-richland-a10-apu-and-neptune-8970m-gaming-notebook

My question is, why fork out the money for a decent gpu if thats not why youre using it?
And isnt it for eye candy as well?
Not sure what theyre getting at here, but it looks crummy for AMD by their wording vs logic

And here:
The other bad news is that my continuing experience with Enduro is that it's not all that it's cracked up to be, but going pure AMD helps quite a bit. Getting updated drivers with an AMD APU and dGPU is easier, and AMD dGPUs simply cooperate with AMD iGPUs better it seems. I've done some testing with the latest 13.5 Beta2 mobile drivers issue on several other laptops (including the MSI GX60), and for most mainstream applications and games they have been fine. However, there are still times when everything doesn't work quite as smoothly as I'd like.

Isnt this a AMD product? Why the redundancy ?
And then we get the smoother comment.
And it infers AMD doesnt work well with Intel
I hope they have a huge multi system revirew to explain all this, as thats the picture Im getting

 

[jdwii]

Once I put the Sabertooth and a 1000 W 80+ Gold PSU in I was able to hit a stable 5.0Ghz overclock on my "old crappy" 965 BE!!

Your going to love the Sabertooth 990FX R2.0!! Keep in mind if you don't have adequate airflow watch your overclocks till you get a better case and cooling (to hit 5.0Ghz on my 965 BE I had to set the vcore to 1.6v (well above the recommended 1.55v max) and was hitting 61C - 62C is max for Phenom II).

the numbers mason, i know when you are lying

5GHz it is not possible to hit stable 5GHz on 965

I missed that part actually i can get my CPU stable at 3.9Ghz with only 1.30625V stable with 24 hours on prime95 with my 212+ heatsink with temps only at 51C.

But however 5Ghz even on water(with a Phenom II x4) doesn't seem likely now yes he might have it there but i bet its not stable(at 63C or below) using prime for at least 4 hours.

 

[jdwii]

On the Anand link, it has the "new" 8970m, and they go on to say:
"The bad news then is that a 10% clock speed increase from Trinity isn't going to be enough to close the gap in many titles, depending on the resolution and quality settings. Looking at Trinity vs. Ivy Bridge with 7970M, I've seen Intel outperform AMD by 50% or more, particularly in titles that pound the CPU (e.g. Skyrim and StarCraft II); on other games, however, it's basically a wash at high quality 1080p settings, so as a more budget-friendly gaming notebook the GX70 has potential."
http://www.anandtech.com/print/6949/msi-gx70-3be-richland-a10-apu-and-neptune-8970m-gaming-notebook

My question is, why fork out the money for a decent gpu if thats not why youre using it?
And isnt it for eye candy as well?
Not sure what theyre getting at here, but it looks crummy for AMD by their wording vs logic

Not trolling but why would you even read an article there of course they're going to say that, When Intel comes out with their haswell their probably say its such a huge difference that the product made the A10 useless compared to the I3. Then they're going to still use horrible benchmarks such as sysmark. If i had it my way toms would buy that site and its whole database of servers and use some military base software to get rid of everything from them once and for all.

 

[mayankleoboy1]

If i had it my way toms would buy that site and its whole database of servers and use some military base software to get rid of everything from them once and for all.

meaning Toms isnt biased ? Toms sometimes publish reviews that look like paid promotions of softwares. And all that front page promotion to write a "Windows 8 guide " .

Anandtech is quite Intel 'friendly' . And every one of them is a fucking Apple fanboi. Every time Apple releases anything, they are squealing like little girls.

 

[de5_Roy]

On the Anand link, it has the "new" 8970m, and they go on to say:

Spoiler

"The bad news then is that a 10% clock speed increase from Trinity isn't going to be enough to close the gap in many titles, depending on the resolution and quality settings. Looking at Trinity vs. Ivy Bridge with 7970M, I've seen Intel outperform AMD by 50% or more, particularly in titles that pound the CPU (e.g. Skyrim and StarCraft II); on other games, however, it's basically a wash at high quality 1080p settings, so as a more budget-friendly gaming notebook the GX70 has potential."
http://www.anandtech.com/print/6949/msi-gx70-3be-richland-a10-apu-and-neptune-8970m-gaming-notebook

My question is, why fork out the money for a decent gpu if thats not why youre using it?
And isnt it for eye candy as well?

imo because it brings down the price and in gpu bound situations, plays on par with intel rigs (higher perf/price in gpu bound scenarios). i think this laptop is a prime example how amd can get a bigger chunk of the bom(moar moniez) by going full-amd. and additional publicity - new richland, new 8000 series gpu - marketing. they (also toms, with gx60) have already shown that ivy with discreet gfx can outperform apu with discreet gfx, especially in the upper-mid+ prices.

Not sure what theyre getting at here, but it looks crummy for AMD by their wording vs logic

Spoiler

And here:
The other bad news is that my continuing experience with Enduro is that it's not all that it's cracked up to be, but going pure AMD helps quite a bit. Getting updated drivers with an AMD APU and dGPU is easier, and AMD dGPUs simply cooperate with AMD iGPUs better it seems. I've done some testing with the latest 13.5 Beta2 mobile drivers issue on several other laptops (including the MSI GX60), and for most mainstream applications and games they have been fine. However, there are still times when everything doesn't work quite as smoothly as I'd like.

Isnt this a AMD product? Why the redundancy ?
And then we get the smoother comment.
And it infers AMD doesnt work well with Intel
I hope they have a huge multi system revirew to explain all this, as thats the picture Im getting

enduro is still [strike]maturing[/strike] lacking while nvidia has better optimus tech.
apus seem to hit thermal ceiling fast (needs much better cooling) and throttle as seen on toms' gx60 review. but, if the apu is kept under 60c (despite amd's stated 100c max.temp.), it'll perform at its peak.... at least that's the impression i got.
where did you get the idea that amd doesn't work well with intel? that statement points to amd's own usual lack of software support.

(this following part is not addressed to you)
and no cralfs, amd having less resources than intel is irrelevant for driver development.

 

[sarinaide]

MSI GX70 3BE: Richland A10 APU and Neptune 8970M Gaming Notebook
http://www.anandtech.com/print/6949/msi-gx70-3be-richland-a10-apu-and-neptune-8970m-gaming-notebook

nvidia's latest gpu scam exposed? 770 is the new 6950? amd calls nvidia out? charlied is rolling on the floor laughing hysterically(roflH)? as usual, take it with a grain of salt.
http://semiaccurate.com/2013/05/13/you-can-upgrade-an-nvidia-gtx680-to-a-gtx770-with-a-simple-bios-flash/

Lol, not even rebadging, now just flashing and selling dubious products that are the same product people already have anyways.

 

[jdwii]

If i had it my way toms would buy that site and its whole database of servers and use some military base software to get rid of everything from them once and for all.

meaning Toms isnt biased ? Toms sometimes publish reviews that look like paid promotions of softwares. And all that front page promotion to write a "Windows 8 guide " .

Anandtech is quite Intel 'friendly' . And every one of them is a fucking Apple fanboi. Every time Apple releases anything, they are squealing like little girls.

Well for me its alot easier to test software than hardware, so i care more about hardware reviews and extremely little for software reviews.

 

[mayankleoboy1]

I am already wondering how Nvidia will do the PR over this leak. How are they going to convince the reviewers that Gtx770 is a new product. :lol:

I smell a red colored rat in this leak....

 

[noob2222]

Cinebench has been cheating benchmarks scores for many years now. I am not worried about if one piece of software runs better in a chip or in other but when the benchmarks cheat and give fakes scores. Can you confirm that Maxon is using Intel compiler for its programs? I.e. can you confirm me that Cinebench benchmarks really represent Cinema4D performance?

I cant confirm, and neither can you that Cinebench cheats against AMD. That is the problem with closed source software. And nobody can say whether a benchmark represents the perf of the actual software. So if you are denouncing Cinebench, you will have to denounce all other closed source benchmarks as well.

It is known that cinebench uses ICC

What does ICC offer AMD in terms of performance?

http://www.agner.org/optimize/blog/read.php?i=49

Tests with AMD bd 8150
MS 17.00 1 SSE2 code path: 724 clock cycles
Intel 12.1.3 1 generic code: 1950 clock cycles
Intel 12.1.3 1 Intel forced path: 720 clock cycles

obviously this test was to show near maximum benefit with code path and won't always be this obvious, but the truth is still evident.

with ICC and intel's checking system for GenuineIntel, amd gets the slowest code path available (yes at times slower than MSVC, but generally near identical). Benchmarking with this compiler is not representative to AMD's performance on an even playing field.

Cinebench may not be intentially "cheating", ICC is doing that for them.

 

[kettu]

But only Kaveri with steamroller cores is confirmed for 2013. Steamroller based FXs and Opterons are still scheduled for 2014. No? That was the reasoning behind the FX Centurion rumour at least.

http://techreport.com/news/24594/amd-roadmap-confirms-steamroller-based-opterons-this-year

It's probably Kaveri based opteron though.

 

[mayankleoboy1]

It is known that cinebench uses ICC

Spoiler

http://www.agner.org/optimize/blog/read.php?i=49
Tests with AMD bd 8150
MS 17.00 1 SSE2 code path: 724 clock cycles
Intel 12.1.3 1 generic code: 1950 clock cycles
Intel 12.1.3 1 Intel forced path: 720 clock cycles
obviously this test was to show near maximum benefit with code path and won't always be this obvious, but the truth is still evident.
with ICC and intel's checking system for GenuineIntel, amd gets the slowest code path available (yes at times slower than MSVC, but generally near identical). Benchmarking with this compiler is not representative to AMD's performance on an even playing field.
Cinebench may not be intentially "cheating", ICC is doing that for them

.

Maybe Cinebench people are using assembly code specially for AMD procs, and using ICC only because its the best compiler. Since CB is so CPU dependent, this probability is very very high.Maybe they manually removed the CPU check. Maybe they added other shims for AMD procs.
We only know that CB is compiled by ICC. We dont know what and how is the code written by CB developers.

Edit : I was just checking the link you gave. It mentions only some libraries that give separate code for Non-intel procs. Those libraries are high performance, and obviously tuned for the internal architecture of Intel procerssors.
Or do you think that the internal arch of AMD is the same as Intel, just because both support x86 ? Or does AMD hand out the details of its architectures to general public and competitors ? Or the marketing PR architecture slides we see at review time is sufficient to make a high performing, tuned library ?

 

[JAYDEEJOHN]

@de5_roy
They say AMD works better with AMD, and even then its not great, so whats that saying how AMD works with other chips?
Thats my point, and left rather ambiguous by Anand.

 

[juanrga]

Cinebench has been cheating benchmarks scores for many years now. I am not worried about if one piece of software runs better in a chip or in other but when the benchmarks cheat and give fakes scores. Can you confirm that Maxon is using Intel compiler for its programs? I.e. can you confirm me that Cinebench benchmarks really represent Cinema4D performance?

I cant confirm, and neither can you that Cinebench cheats against AMD. That is the problem with closed source software. And nobody can say whether a benchmark represents the perf of the actual software. So if you are denouncing Cinebench, you will have to denounce all other closed source benchmarks as well.

You cannot confirm your claims, but I can confirm the mines. It is well-known that Cinebench, Sysmark, and other benchmarks cheat against AMD

http://sharikou.blogspot.com.es/2009/12/ftc-accuses-intel-of-rigging-benchmarks.html

Yes Maxon claims in its website that Cinebench is based in professional software Cinema4D, "which is used extensively by studios and production houses worldwide for 3D content creation", but since Titanic was produced using Linux in 1997 all major studios including Dreamworks Animation, Pixar, Weta Digital, and Industrial Light & Magic have migrated to Linux. Currently more than 95% of the servers and desktops at large animation and visual effects companies use Linux, which imply that 95% of the industry is not using Cinema4D.

Thats a big claim you are making. Got any link to back it up ?

http://en.wikipedia.org/wiki/Linux#Market_share_and_uptake

I am happy with open benchmarks or with closed benchmarks compiled with a fair compiler. I just reject the well-known unfair closed benchmarks that cheat about the scores giving fake advantage to Intel chips.

If the benchmarks are closed source, how do you know they cheat against AMD, or are unfair ? You are just making that up from your idea of what an AMD CPU should perform like.

There are ways to detect if a closed source benchmark is cheating or not. Sysmark and Cinebench are biased

http://sharikou.blogspot.com.es/2009/12/ftc-accuses-intel-of-rigging-benchmarks.html

PCmark gives up to a 40% more score when believes that a given chip is "made by Intel".

http://www.zdnet.com/blog/hardware/pcmark-2005-memory-benchmark-favors-intel-cpus/2346

I repeat, if Intel chips are so superior why do the biased benchmarks exist?

Because marketing. And AMD does not have the money to influence benchmarks.

You are contradicting yourself. If intel chips are so good. Marketing dept. only needs to emphasize how good the chips are.

But, if the chips are not so good as you believe, then Intel needs to cheat about the benchmarks giving fake scores and fake figures.

It is not about money. There are companies so big like Intel and none of them has been accused of continuous cheating/lying about benchmarks or about illegal practices.

http://en.wikipedia.org/wiki/Intel#Competition.2C_antitrust_and_espionage

http://en.wikipedia.org/wiki/Intel#Competition

http://en.wikipedia.org/wiki/AMD_v._Intel

As a final note. AMD already introduced Steamroller optimization in the main compilers as GCC. When SteamRoller was released compilers will use its potential since day one.

That has been the case with Bulldozer and Piledriver as well. Usually, compiler support is added months before actual product is released.

One of the problems of Bulldozer was the lack of software prepared for its innovative technology. AMD is not repeating the same mistake:

it is evident that the company is more concerned about ensuring that peculiarities of the Steamroller cores are taken into consideration by software designers on the first place.

http://www.xbitlabs.com/news/cpu/display/20121011232630_AMD_Secretly_Rolls_Out_Steamroller_Support_Patch_for_Compilers.html

 

[juanrga]

On the Anand link, it has the "new" 8970m, and they go on to say:
"The bad news then is that a 10% clock speed increase from Trinity isn't going to be enough to close the gap in many titles, depending on the resolution and quality settings. Looking at Trinity vs. Ivy Bridge with 7970M, I've seen Intel outperform AMD by 50% or more, particularly in titles that pound the CPU (e.g. Skyrim and StarCraft II); on other games, however, it's basically a wash at high quality 1080p settings, so as a more budget-friendly gaming notebook the GX70 has potential."
http://www.anandtech.com/print/6949/msi-gx70-3be-richland-a10-apu-and-neptune-8970m-gaming-notebook

My question is, why fork out the money for a decent gpu if thats not why youre using it?
And isnt it for eye candy as well?
Not sure what theyre getting at here, but it looks crummy for AMD by their wording vs logic

And here:
The other bad news is that my continuing experience with Enduro is that it's not all that it's cracked up to be, but going pure AMD helps quite a bit. Getting updated drivers with an AMD APU and dGPU is easier, and AMD dGPUs simply cooperate with AMD iGPUs better it seems. I've done some testing with the latest 13.5 Beta2 mobile drivers issue on several other laptops (including the MSI GX60), and for most mainstream applications and games they have been fine. However, there are still times when everything doesn't work quite as smoothly as I'd like.

Isnt this a AMD product? Why the redundancy ?
And then we get the smoother comment.
And it infers AMD doesnt work well with Intel
I hope they have a huge multi system revirew to explain all this, as thats the picture Im getting

Anandtech is well-known for its Intel/Apple bias. They are not professional about benchmarks, quality, price, power consumption... In their forums they make all kind of attacks against AMD, and ban people who says they the truth. However posters saying that people who chose AMD instead of Intel are "mentally insane" (sic) are allowed to post and even applauded...

Google is your friend

 

[de5_Roy]

@de5_roy
They say AMD works better with AMD, and even then its not great, so whats that saying how AMD works with other chips?
Thats my point, and left rather ambiguous by Anand.

they mentioned gpus and enduro, so i am guessing that they're hinting at gpu switching. since both gpus are amd's, software support is(or should be) easier. meanwhile the usage experience, while functioning, isn't smooth enough - that's my assumption. i think relatively higher amount of stuttering or slowdown is involved, probably the cpu can't handle it well enough yet. the writer has hinted at enduro's shortcomings and lack of improvement for a while. iirc amd launched enduro with quite a bit of fanfare.
imho, the vagueness of that statement does lead to assuming amd doesn't work well with intel. i think it's a failure on both amd's and intel's lack of mutual cooperation and software support(both are well-known for it), if true.

 

[de5_Roy]

Anandtech is well-known for its Intel/Apple bias. They are not professional about benchmarks, quality, price, power consumption... In their forums they make all kind of attacks against AMD, and ban people who says they the truth. However posters saying that people who chose AMD instead of Intel are "mentally insane" (sic) are allowed to post and even applauded...

Google is your friend

https://pbs.twimg.com/media/BHNkE2yCYAEmtGF.png:large

is that screenshot supposed to mean something? anything? i just typed tomshardware intel -> google corrected it to tom's hardware intel -> the second and 4th suggestions are 'intel bias', 3rd and 4th if you type tom's hardware bia... :lol:
some friend google turned out to be...

edit:
s/a is on a roll
http://semiaccurate.com/2013/05/14/why-would-nvidia-release-an-unlocked-gtx680-to-gtx770-bios/
tpu's take
http://www.techpowerup.com/183936/geforce-gtx-680-can-be-flashed-to-gtx-770.html

First Richland desktop part available in Europe (with 'benches')
http://www.fudzilla.com/home/item/31376-first-richland-desktop-part-available-in-europe

 

[JAYDEEJOHN]

@de5_roy
They say AMD works better with AMD, and even then its not great, so whats that saying how AMD works with other chips?
Thats my point, and left rather ambiguous by Anand.

they mentioned gpus and enduro, so i am guessing that they're hinting at gpu switching. since both gpus are amd's, software support is(or should be) easier. meanwhile the usage experience, while functioning, isn't smooth enough - that's my assumption. i think relatively higher amount of stuttering or slowdown is involved, probably the cpu can't handle it well enough yet. the writer has hinted at enduro's shortcomings and lack of improvement for a while. iirc amd launched enduro with quite a bit of fanfare.
imho, the vagueness of that statement does lead to assuming amd doesn't work well with intel. i think it's a failure on both amd's and intel's lack of mutual cooperation and software support(both are well-known for it), if true.

Agreed, AMD isnt out of the woods yet here, and my POV is, RR is fixing all the little things while keeping a firm grip at the helm.
If you look at the talent theyve brought in, as well as new up and comers we havnt heard about yet, I just dont see AMD as being known as tolerable, at least, not any more.
You can say, you have only so much to work with, but the things you can, you can do very well, and thats the legacy of some of the new/old hires

 

[cowboy44mag]

Once I put the Sabertooth and a 1000 W 80+ Gold PSU in I was able to hit a stable 5.0Ghz overclock on my "old crappy" 965 BE!!

Your going to love the Sabertooth 990FX R2.0!! Keep in mind if you don't have adequate airflow watch your overclocks till you get a better case and cooling (to hit 5.0Ghz on my 965 BE I had to set the vcore to 1.6v (well above the recommended 1.55v max) and was hitting 61C - 62C is max for Phenom II).

the numbers mason, i know when you are lying

5GHz it is not possible to hit stable 5GHz on 965

I missed that part actually i can get my CPU stable at 3.9Ghz with only 1.30625V stable with 24 hours on prime95 with my 212+ heatsink with temps only at 51C.

But however 5Ghz even on water(with a Phenom II x4) doesn't seem likely now yes he might have it there but i bet its not stable(at 63C or below) using prime for at least 4 hours.

I don't keep my overclock at 5Ghz and the vcore has to be set too high and heat does become a problem. I live in Pa, and we have just started to have a little warm weather here and there, when I had it overclocked to 5Ghz I turned the heat off to my "man cave" (family room in basement) and let it run overnight and next day (almost 24 hrs) ambient room temps were probably 40F or below, and like I said not too many people put two delta fans on a Hyper 212 EVO, it makes a difference. I'm not saying it was totally stable at that overclock as past running Prime 95 and playing a little Skyrim to enjoy the overclock I didn't keep it there for long. A lot of overclockers will also say its not stable unless it can post Intel Burn Test too, it passed Prime 95, however I shut Burn Test down after a few minutes as it was jumping to nearly 70C, way too high. Most overclockers won't push the vcore past 1.55v, they will stop there and start to find highest stable overclock at 1.55. I had mine a little past 1.6v, which most "experts" say will fry the processor within seconds. Those must be the same "experts" who say 5Ghz is impossible, because my processor is fine.

It is also noteworthy that just about all people who had Phenom IIs have upgraded as they upgrade the "supporting hardware" (ie motherboards, ram, psu, ect) to bigger and better processors. I couldn't even hit a true 4Ghz overclock on my old motherboard, psu, ram configuration. It wasn't until I put the Sabertooth 990FX R2.0, a 1000W 80+ Gold psu, and Crucible Tactical RAM that I was able to hit that overclock. 5Ghz is more than possible, as these chips have been pushed to at least 8Ghz by overclock enthusiasts (they were of course using insane cooling).

With summer coming and warm weather coming there is no way I can keep a 5Ghz overclock, which is why I set it to 4.5Ghz which is more manageable, however I my have to set to 4.2 or 4.1 if summer temps get too high (one summer almost every freaking day was high 90s or over 100F). Running air conditioners just to keep a overclock gives too much money to the hated electric company.

People tend to forget some very important rules about overclocking. The first and most important is every chip is different, no two processors exactly the same will overclock the same. One may only be stable to 3.8Ghz, one may hit 4.2, and one may hit 4.5, ect.. But the other thing is most people don't keep a 4 year old processor as their "workhorse" and see what an older processor can do with new supporting hardware. Most people have never tried to over clock a Phenom II 965 or 980 using a Sabertooth 990FX R2.0. A lot of remaining Phenom IIs in everyday use are hecta core which don't tend to overclock as high as the Phenom II quad cores so right away people jump on the "impossible" bandwagon. Anytime in history that someone said "that's impossible" someone else did it. 5Ghz was never meant to be an everyday overclock, just wanted to see how high I could push it.

 

[8350rocks]

http://www.fudzilla.com/home/item/31264-amd-creates-custom-chip-division

Looks like AMD has setup a shop to do more custom solutions like PS4. Apparently, they will offer integrated x86 and ARM based APUs.

That's catering to a niche that doesn't really exist from a manufacturer yet. Creating a custom solution without having to burn copious amounts of R&D cash. Maybe that will take off, maybe not, but the idea has some merit.

 

[sarinaide]

The A4 4000 at $40 is a fantastic part, I used to get the Athlon II 740 at $50 but this A4 offers the benefit of close enough core performance and some bustly graphics to go with it. If a person is building a low cost iTX system for HTPC or media systems this is a virtual steal.

 

[truegenius]

^ and jaguar will steal this title from a4