haswell or piledriver

[vishalaestro]

i need a recommendation for the cpu since today's games are using more no of cores is it wise to stay with amd or with intel

 

[8350rocks]

The extra 100MHz is negligible compared to 4 more shaders. From a gaming perspective, the Intel HD 4600 is still to weak anyway for anyone who wants to build a gaming rig.

Improved integrated graphics performance is more important in laptop than desktops since it basically increases the minimum performance. While the average Haswell or Richland laptop relying on just the iGPU is not going to run graphic intensive games very well, at least they can provide some level of enjoyment for games with modest graphic requirements.

That is one arena where AMD really excels, the Richland on board graphics are supposed to be right on the heels of a HD 7750 card

I think that is a bit of a stretch. Benchmarks have shown that the Radeon HD 7660D is a little faster than the desktop Radeon HD 5570 graphics card. I would called the HD 7660D a "desktop Radeon HD 5580" if such a card ever existed.

A performance increase from a "Radeon HD 5580" to a Radeon HD 5770 would be astronomical for an iGPU. The Radeon HD 7750 is basically equal to the Radeon HD 6770 and the HD 6770 is really just a re-badged Radeon HD 5770.

Not having research anything about Richland's iGPU yet, I think that it's performance will be closer much closer to the Radeon HD 6670 than the Radeon HD 6770 (HD 7750) by a wide margin. I'll refine that a bit more and state that my estimate of Richland's highest performing iGPU will be somewhere in between the Radeon HD 5670 and the Radeon HD 6670. The Radeon HD 6670 is has about 66% the performance of the Radeon HD 5770/6770/7750. A Radeon HD 5570 in turn has about 66% the performance of the Radeon HD 5670.

Additionally, Trinity has 1.5 billion transistors, granted I do not know how many transistors is devoted towards the Radeon HD 7660D. Richland's transistor count will likely be a little more than that. However, I do know that the Radeon HD 7750 has 1.3 billion transistors. Combining a Radeon HD 7750 like iGPU into the Richland APU will likely mean the transistor count will go beyond 2 billion which means a larger APU and higher production costs. Plus it will also threaten AMD's low end and mainstream graphic card sales which represents the bulk graphic card sales. Such a move by AMD will mean that Richland will need to sell at a higher price to cover increased production cost; threatening sales. And Richland will also undermine AMD's own graphic card sales. Both are bad for AMD's long term operations.

Actually, it's not bad for graphics card sales or long term operations...here's why...

AMD's top of the line A10-5800k is a $122.00 solution for the guy who is looking at a budget PC that encompasses low end graphics and a good CPU(3.8 GHz stock clock). The newest Richland (top of the line A10-6800k) is going to be right at 4.0 GHz with TC to 4.4 GHz, the onboard graphics are using GCN core technology, which is supposed to give a 20% increase in performance. This now allows someone to spend ~$140 for a CPU with 4.0 GHz rating and ~ HD 7750 graphics, making AMD the ONLY solution to build a decent $500 gaming rig. That won't run most games at high settings, but would do fine at 720p or less, especially at smaller screen resolutions.

That would put AMD APUs on the radar for anyone on an extremely tight budget, people who try the APUs and later can afford to build a solid gaming system, will likely have a fairly high % of brand loyalty after the good experience they have with the budget solutions offered by AMD.

Further, this technology will make it's way into the Laptop arena at some point on the low power chip solutions which will make AMD increase market share in the laptop/mobile market.

This just makes AMD that much stronger...

 

[hapkido]

I think it's hilarious you think a chip make (AMD in this case) can double the processing power of their current chips without increasing the number of transistors, and then cut power by about 70-80W and move them to laptops and not lose any processing power. Why wouldn't they put these 4-core + 7750 25-35W chips in their desktop lineup to begin with?

It's like you read a few reviews and wikipedia articles, but don't understand what's actually going on. 7750 is 1.5 billion transistors with 512 cores using 55W. a10-5800k is 1.3 billion transistors -- 200 million fewer transistors for both CPU and GPU than 7750 has. The new Richland a10s will have 384 GPU cores and the entire package will use about 35W in laptops. How is AMD going to pack something as fast as a 7750 into an APU that has fewer transistors and uses less power, without shrinking their processing node -- a lot?

You make up this BS and then spout it off to others and claim everyone who says you're wrong is a biased fanboy. YOU are the biased fanboy. Just stop already.

 

[8350rocks]

I think it's hilarious you think a chip make (AMD in this case) can double the processing power of their current chips without increasing the number of transistors, and then cut power by about 70-80W and move them to laptops and not lose any processing power. Why wouldn't they put these 4-core + 7750 25-35W chips in their desktop lineup to begin with?

It's like you read a few reviews and wikipedia articles, but don't understand what's actually going on. 7750 is 1.5 billion transistors with 512 cores using 55W. a10-5800k is 1.3 billion transistors -- 200 million fewer transistors for both CPU and GPU than 7750 has. The new Richland a10s will have 384 GPU cores and the entire package will use about 35W in laptops. How is AMD going to pack something as fast as a 7750 into an APU that has fewer transistors and uses less power, without shrinking their processing node -- a lot?

You make up this BS and then spout it off to others and claim everyone who says you're wrong is a biased fanboy. YOU are the biased fanboy. Just stop already.

You're not reading what I am saying, I didn't say the architecture would go into laptops/mobiles...I said the TECHNOLOGY...(i.e. Jaguar which is powering the next gen consoles for mobiles)...GCN is ALREADY a component in Jaguar/Richland series processors. Richland will be for laptops and Jaguar for mobiles.

They didn't have to change the transistor count...THEY CHANGED THE ENTIRE TECHNOLOGY SYSTEM.

Can you read? I am curious because it seems like you just type stuff to type it. Practicing your keyboard skills a bit are you?

Read this...if you ever do anything on this site but type BS...I don't know how you could've missed it:

http://www.tomshardware.com/news/Richland-APU-AMD,21318.html

Look at those HUGE performance gains...

 

[hapkido]

I think it's hilarious you think a chip make (AMD in this case) can double the processing power of their current chips without increasing the number of transistors, and then cut power by about 70-80W and move them to laptops and not lose any processing power. Why wouldn't they put these 4-core + 7750 25-35W chips in their desktop lineup to begin with?

It's like you read a few reviews and wikipedia articles, but don't understand what's actually going on. 7750 is 1.5 billion transistors with 512 cores using 55W. a10-5800k is 1.3 billion transistors -- 200 million fewer transistors for both CPU and GPU than 7750 has. The new Richland a10s will have 384 GPU cores and the entire package will use about 35W in laptops. How is AMD going to pack something as fast as a 7750 into an APU that has fewer transistors and uses less power, without shrinking their processing node -- a lot?

You make up this BS and then spout it off to others and claim everyone who says you're wrong is a biased fanboy. YOU are the biased fanboy. Just stop already.

You're not reading what I am saying, I didn't say the architecture would go into laptops/mobiles...I said the TECHNOLOGY...(i.e. Jaguar which is powering the next gen consoles for mobiles)...GCN is ALREADY a component in Jaguar/Richland series processors. Richland will be for laptops and Jaguar for mobiles.

They didn't have to change the transistor count...THEY CHANGED THE ENTIRE TECHNOLOGY SYSTEM.

Can you read? I am curious because it seems like you just type stuff to type it. Practicing your keyboard skills a bit are you?

Read this...if you ever do anything on this site but type BS...I don't know how you could've missed it:

http://www.tomshardware.com/news/Richland-APU-AMD,21318.html

Look at those HUGE performance gains...

WHAT HUGE PERFORMANCE GAINS?

I see a picture of an APU with 2 CPU cores and 384 GPU shader cores with no mention of a die shrink or that the graphics will be anywhere near a 7750.

Start providing concrete evidence or shut up. Your opinion means jack.

 

[8350rocks]

I think it's hilarious you think a chip make (AMD in this case) can double the processing power of their current chips without increasing the number of transistors, and then cut power by about 70-80W and move them to laptops and not lose any processing power. Why wouldn't they put these 4-core + 7750 25-35W chips in their desktop lineup to begin with?

It's like you read a few reviews and wikipedia articles, but don't understand what's actually going on. 7750 is 1.5 billion transistors with 512 cores using 55W. a10-5800k is 1.3 billion transistors -- 200 million fewer transistors for both CPU and GPU than 7750 has. The new Richland a10s will have 384 GPU cores and the entire package will use about 35W in laptops. How is AMD going to pack something as fast as a 7750 into an APU that has fewer transistors and uses less power, without shrinking their processing node -- a lot?

You make up this BS and then spout it off to others and claim everyone who says you're wrong is a biased fanboy. YOU are the biased fanboy. Just stop already.

You're not reading what I am saying, I didn't say the architecture would go into laptops/mobiles...I said the TECHNOLOGY...(i.e. Jaguar which is powering the next gen consoles for mobiles)...GCN is ALREADY a component in Jaguar/Richland series processors. Richland will be for laptops and Jaguar for mobiles.

They didn't have to change the transistor count...THEY CHANGED THE ENTIRE TECHNOLOGY SYSTEM.

Can you read? I am curious because it seems like you just type stuff to type it. Practicing your keyboard skills a bit are you?

Read this...if you ever do anything on this site but type BS...I don't know how you could've missed it:

http://www.tomshardware.com/news/Richland-APU-AMD,21318.html

Look at those HUGE performance gains...

WHAT HUGE PERFORMANCE GAINS?

I see a picture of an APU with 2 CPU cores and 384 GPU shader cores with no mention of a die shrink or that the graphics will be anywhere near a 7750.

Start providing concrete evidence or shut up. Your opinion means jack.

They did shrink the die to 28nm...maybe it doesn't cover it in that article, but it's EASILY findable with google.

GO READ UP ON IT.

 

[hapkido]

28nm GPU is NOT a die-shrink for AMD.

Stop making things up. Every time I prove you wrong, you go on to spout something else that's also wrong.

 

[8350rocks]

28nm GPU is NOT a die-shrink for AMD.

Stop making things up. Every time I prove you wrong, you go on to spout something else that's also wrong.

So, what you're telling me is going from 32nm trinity to 28nm richland isn't a die shrink? Well, that's the first I've heard of changing a die size not being a die shrink...

For a GPU, it's not a die shrink...for an APU, it is...

Get your facts straight, come back and we'll talk.

Edit: In the Tom's article it mentions 28nm too, they just don't specify there that it's a die shrink. If you look at wikipedia on AMD Richland, it still says 32nm because in 2012 when it was announced it was going to be 32nm, instead they held it back 6 months and went to a new die and new architecture. Trinity was 32nm process, NOT 28nm process.

 

[jaguarskx]

Some AMD representative announced that Richland A10 will be between 20-40% faster than Trinity A10. Using that, we would wait the new APU to be somewhere between a Radeon HD 6670 and a HD 7750.

If that is true, then it should validate my estimate that Richland's best iGPU performance will be between the Radeon HD 5670 and HD 6670. The below chart has the summarized performance of the Radeon HD 5570, HD 5670, and HD 6670 for various resolutions. Taking the HD 6670's performance as the baseline (i.e. 100%), the Radeon HD 5670 has 82% the performance of the Radeon HD 6670 and the HD 5570 has 64% of the Radeon HD 6670's performance.

Another way of looking at it is that the Radeon HD 5670 is 28.1% more powerful than the Radeon HD 5570 and the Radeon HD 6670 is 56.25% more powerful than the Radeon HD 5570.

As I stated before, the Trinity's Radeon HD 7660D is a little faster than the Radeon HD 5570, (by about 10% base on all the benchmarks I've seen) so if Richland's best iGPU is going to have a 20% - 40% performance gain, then it's performance will still be between the Radeon HD 5670 and HD 6670.

Source:
http://www.techpowerup.com/reviews/Axle/Radeon_HD_5670/29.html

As stated before, the Radeon HD 5770, HD 6770 and HD 7750 all basically have similar performance. Going from the Radeon HD 5570 to the Radeon HD 5770 is more than a 150% (1.5x) increase in average performance. To me that is far beyond any reasonable performance increase for an integrated graphic core performance and such claims should be viewed with a mountain of skepticism.

We'll know soon enough though...

 

[hapkido]

28nm GPU is NOT a die-shrink for AMD.

Stop making things up. Every time I prove you wrong, you go on to spout something else that's also wrong.

So, what you're telling me is going from 32nm trinity to 28nm richland isn't a die shrink? Well, that's the first I've heard of changing a die size not being a die shrink...

For a GPU, it's not a die shrink...for an APU, it is...

Get your facts straight, come back and we'll talk.

"The AMD 10 6800K is said to be based on a 32nm fabrication process..."

http://wccftech.com/amd-a10-6800k-flagship-28nm-richland-apu/

Are we done here yet?

 

[8350rocks]

Look at you sitting there "hunting for obscure sites that support your opinion"...

TOM'S HARDWARE, where we're posting, says it's a 28nm process...

Oh, and by the way, AMD...that manufacturer that makes them...they are telling you it's a 28nm process...

How dense are you?

http://wccftech.com/amds-kaveri-based-28nm-richland-apu-features-steamroller-cores-compatibility-fm2-socket/

The same guy you just quoted, in a newer post talking about 28nm...

Here's a few more:

http://www.xbitlabs.com/news/cpu/display/20130107232327_AMD_Reveals_Client_APUs_for_2013_Richland_Kabini_Temash_and_Kaveri.html

http://news.softpedia.com/news/A10-6800K-AMD-s-Top-28nm-Richland-APU-317793.shtml

http://forums.pureoverclock.com/general-discussion/19560-amd-s-kaveri-based-28nm-richland-apu-features-steamroller-cores.html

http://forums.anandtech.com/showthread.php?t=2292118

"Are we done here yet?"

 

[hapkido]

It's 32nm.

http://en.wikipedia.org/wiki/AMD_Accelerated_Processing_Unit#Desktop_and_Mobile

How about you stop trying to make me prove your point? You're wrong. You've been wrong about EVERYTHING.

 

[hapkido]

Richland and Kaveri are not the same APU. I don't know what your point is.

Even when AMD does release their 28nm APU, it's still not going to have graphics comparable to a 7750, so your whole point is moot. Stop moving the goalpost.

edit:

From AMD --
http://products.amd.com/en-us/NotebookAPUDetail.aspx?id=62&f1=AMD+Quad-Core+A10-Series+APU+for+Notebooks&f2=A10-5750M+with+Radeon%E2%84%A2+HD+8650G&f3=2500&f4=4&f5=FS1r2&f6=35+W&

"CMOS 32nm"

From Toms --
http://www.tomshardware.com/news/AMD-APU-CPU-richland,20640.html

"Just like the current Trinity APUs, the Richland APU's will drop right into the existing FM2 socket. A total of six models will be released, amongst which four are quadcore and two are dualcore. The x86 cores of the APUs will be built from either one or two 'Enhanced Piledriver' modules, which will be fabricated using a 32nm production process."

You are either uneducated or are lying. Either way, you lose credibility -- especially since you won't admit you made a mistake -- and your opinion should be disregarded.

 

[8350rocks]

They can change the die size and still fit the same socket...hello?! It wasn't a drastic change in die size...Socket AM3 processors will fit an AM3+ board even though the die size changed from the Phenom II series to the current FX series...

You're clearly not well versed in the realm of hardware...

Intel changes sockets like you change socks, but AMD supports their hardware manufacturers for longer terms so that Joe Q. Public doesn't have to buy a new board necessarily every time he wants a performance upgrade.

From Tom's dated 9 Jan 2013:

http://www.tomshardware.com/news/amd-richland-28nm-a10-6800k,20272.html

Note where it says 28nm architecture.

 

[hapkido]

They can change the die size and still fit the same socket...hello?! It wasn't a drastic change in die size...Socket AM3 processors will fit an AM3+ board even though the die size changed from the Phenom II series to the current FX series...

You're clearly not well versed in the realm of hardware...

Intel changes sockets like you change socks, but AMD supports their hardware manufacturers for longer terms so that Joe Q. Public doesn't have to buy a new board necessarily every time he wants a performance upgrade.

I just want to slap you. AMD lists the CMOS at 32nm on their website. What more proof do you need?

Moving on. The other poster provided evidence the GPU won't be anywhere near a 7750. Even when they do move to 28nm APUs, AMD is already at 28nm on their GPUs. A 28nm GPU operating at 720MHz with 384 shader units is not going to be as fast as a 28nm GPU operating at 800MHz with 512 shaders. Fewer shader units and less clock speed means it will be quite a bit slower.

The texture fillrates speak for themselves, 25.6 Gtex/s for the 7750 versus 12.8 Gtex/s for the 8650D in the Richland APU. If you can't do that math, that's half. For comparison, AMD's 5670 is 15.5 Gtex/s, which would make sense because it has 16 more shader units operating 55MHz faster.

 

[hapkido]

From Tom's dated 9 Jan 2013:

http://www.tomshardware.com/news/amd-richland-28nm-a10-6800k,20272.html

Note where it says 28nm architecture.

Stop ninja editing your posts.

The article I linked from Tom's is 15 days newer and says 32nm. AMD lists 32nm on their website.

You're also using a tactic called moving the goalpost. When you're shown to be wrong about something (they are as fast as 7750s), you try to change the argument back to something else (the processing node). You're lying about that too, but it's inconsequential. The Richland APUs are out for mobile devices. Show me benchmarks where the GPU performs on par with a 7750 and come back.

I'm done arguing with you about this. You are a liar.

 

[8350rocks]

From Tom's dated 9 Jan 2013:

http://www.tomshardware.com/news/amd-richland-28nm-a10-6800k,20272.html

Note where it says 28nm architecture.

Stop ninja editing your posts.

The article I linked from Tom's is 15 days newer and says 32nm. AMD lists 32nm on their website.

You're also using a tactic called moving the goalpost. When you're shown to be wrong about something (they are as fast as 7750s), you try to change the argument back to something else (the processing node). You're lying about that too, but it's inconsequential. The Richland APUs are out for mobile devices. Show me benchmarks where the GPU performs on par with a 7750 and come back.

I'm done arguing with you about this. You are a liar.

They're not out yet, they won't be until July...there are no benchmarks yet...

I am not lying to you...I stated the onboard graphics should be close to HD 7750 performance.

http://www.overclock.net/t/1347709/amd-richland-a10-6800k-apu-thread/20

http://www.techpowerup.com/179248/AMD-quot-Richland-quot-Desktop-APU-Lineup-Detailed.html

http://wccftech.com/amd-launching-28nm-kaveri-apu-steamroller-cores-2013/

Richland will be such a drastic improvement over Trinity, they expect anywhere from 20-40% improvement. Based on shaders alone, that's 75% of the performance of the HD 7750, with the gains in performance from GCN and piledriver architecture, they expect the iGPU for the Richland to meet about 85-90% of the performance of the HD 7750.

Which is really close.

Further I am not moving the goalpost, I stated a fact, and you came in telling me I was full of BS, though you can easily find the information yourself if you wanted to read up on it. Though clearly you don't.

So stop trolling...Richland will be close to HD 7750 onboard. You cannot dispute that.

 

[juanrga]

Another uneducated user, passing along bad information.

DESKTOP i5s are ALL 4 cores / 4 threads. i3s are 2 cores / 4 threads and i7s are 4 cores / 8 threads. Pentiums are Intel's 2 cores / 2 threads CPU on desktop.

To be fair the distinction between desktop and mobile chips is a bit arbitrary those days. He mentioned his i5-3210M (a 2 core chip), which is usually considered a mobile chip, but however used in several desktops computers.

Thus "desktop i5 are ALL 4 core" does not mean "desktop pcs with a i5 are ALL 4 core".

 

[ihog]

All the i5's that I have come across are dual cores with hyper threading, I say this on my i5 3210m. Modules? These little chips are first 8 core design. Once again AMD has done it again back in the day their original Phenom chips were the first quad core design.

http://ark.intel.com/products/family/65504
Just click on a processor, and you will find many of them have 4 physical cores. Just because you haven't come across one that has 4, doesn't mean they don't exist.

And you didn't understand what I said about AMD's 8 cores.

 

[hapkido]

From Tom's dated 9 Jan 2013:

http://www.tomshardware.com/news/amd-richland-28nm-a10-6800k,20272.html

Note where it says 28nm architecture.

Stop ninja editing your posts.

The article I linked from Tom's is 15 days newer and says 32nm. AMD lists 32nm on their website.

You're also using a tactic called moving the goalpost. When you're shown to be wrong about something (they are as fast as 7750s), you try to change the argument back to something else (the processing node). You're lying about that too, but it's inconsequential. The Richland APUs are out for mobile devices. Show me benchmarks where the GPU performs on par with a 7750 and come back.

I'm done arguing with you about this. You are a liar.

They're not out yet, they won't be until July...there are no benchmarks yet...

I am not lying to you...I stated the onboard graphics should be close to HD 7750 performance.

http://www.overclock.net/t/1347709/amd-richland-a10-6800k-apu-thread/20

http://www.techpowerup.com/179248/AMD-quot-Richland-quot-Desktop-APU-Lineup-Detailed.html

http://wccftech.com/amd-launching-28nm-kaveri-apu-steamroller-cores-2013/

Richland will be such a drastic improvement over Trinity, they expect anywhere from 20-40% improvement. Based on shaders alone, that's 75% of the performance of the HD 7750, with the gains in performance from GCN and piledriver architecture, they expect the iGPU for the Richland to meet about 85-90% of the performance of the HD 7750.

Which is really close.

Further I am not moving the goalpost, I stated a fact, and you came in telling me I was full of BS, though you can easily find the information yourself if you wanted to read up on it. Though clearly you don't.

So stop trolling...Richland will be close to HD 7750 onboard. You cannot dispute that.

ON AMD'S WEBSITE:

http://products.amd.com/en-us/NotebookAPUDetail.aspx?id=62&f1=AMD+Quad-Core+A10-Series+APU+for+Notebooks&f2=A10-5750M+with+Radeon%E2%84%A2+HD+8650G&f3=2500&f4=4&f5=FS1r2&f6=35+W&

"CMOS 32nm"

You said AMD said they're 28nm, and you keep insisting on that. That right there proves you are a LIAR. You are completely discredited.

The mobile Richland APUs are out, they were released March 19. If you're going to continue to lie its GPU is as fast as a 7750, provide your benchmarks or shut up.

 

[juanrga]

The mobile Richland APUs are out, they were released March 19. If you're going to continue to lie its GPU is as fast as a 7750, provide your benchmarks or shut up.

I don't want to enter in your personal discussion, but the performance of the just released mobile Richland APUs chips is not the same than of the future desktop Richland.

 

[hapkido]

The mobile Richland APUs are out, they were released March 19. If you're going to continue to lie its GPU is as fast as a 7750, provide your benchmarks or shut up.

I don't want to enter in your personal discussion, but the performance of the just released mobile Richland APUs chips is not the same than of the future desktop Richland.

Yes of course. I mentioned the mobile APUs because he said, "this technology will make it's way into the Laptop arena at some point on the low power chip solutions". I was disputing his thinking that a 35W laptop part would rival a 100W desktop part someone. Even then, if the new Richland APU doesn't make huge performance increases (like 2-3 times) over its Trinity predecessor, there's no reason to think the desktop Richland APU would see the 2-3 times performance increase needed to match a 7750 -- especially without a die shrink.

 

[jaguarskx]

I am not lying to you...I stated the onboard graphics should be close to HD 7750 performance.

http://www.overclock.net/t/1347709/amd-richland-a10-6800k-apu-thread/20

That is merely a thread from another forum of people speculating that the Richland iGPU is going to be as nearly as powerful as a Radeon HD 7750. They offer no substance to backup their speculation.

http://www.techpowerup.com/179248/AMD-quot-Richland-quot-Desktop-APU-Lineup-Detailed.html

The article from Techpowerup.com does not state anything about the expected of Richland's iGPU. The only the Radeon HD 7750 is only mentioned in the comment section and that only pertains to dual graphics with a Radeon HD 7750 and a link referencing a review of a low profile Radeon HD 7750.

http://wccftech.com/amd-launching-28nm-kaveri-apu-steamroller-cores-2013/

Again, there is nothing in the thread which specifically states the Richland's iGPU is anywhere near the performance level of the Radeon HD 7750. It only states that it is possible that Richland's iGPU will be 20% - 40% more powerful than Trinity's Radeon HD 7660D iGPU.

If you look at my analysis at the top of the 2nd page of this thread, a 20% - 40% increase in performance will mean that Richland's iGPU performance will be somewhere in between the Radeon HD 5670 and HD 6670.

Richland will be such a drastic improvement over Trinity, they expect anywhere from 20-40% improvement. Based on shaders alone, that's 75% of the performance of the HD 7750, with the gains in performance from GCN and piledriver architecture, they expect the iGPU for the Richland to meet about 85-90% of the performance of the HD 7750.

I have no problems with the first sentence as it is backup by the link to the 3rd article. Whether the 20% - 40% increase in performance is actually acurrate is a different issue, but it it seems reasonable enough for me not to raise a red flag.

As for the remainder of the paragraph in bold red, I have no idea where you are getting your information from or if you are simply making it up. None of the three links mentions anything about shaders. Actually, the first link to overclock.net is simply to a forum full of member speculation so that link is pure garbage.

Further I am not moving the goalpost, I stated a fact, and you came in telling me I was full of BS, though you can easily find the information yourself if you wanted to read up on it. Though clearly you don't.

So stop trolling...Richland will be close to HD 7750 onboard. You cannot dispute that.

You did not provide any facts. The closest you provided was an opinion that Richland's will probably be 20% - 40% more powerful than the Radeon HD 7660D in the Trinity APU. But your claim that Richland's iGPU is going to be close to the Radeon HD 7750 is utterly baseless. Nearly everything you have posted regarding this topic can be and has been disputed.

I don't care if you want to delude yourself into thinking that Richland's iGPU will be nearly as powerful as a Radeon HD 7750.

However, I do care if you attempt to purposely mislead other people into believing in something that is utterly incorrect by make baseless claims and trying to back them up with links that supposed backs up what you are trying to claim. The fact that none of the links provides any backup to your claims clearly indicates that you are going out of your way to mislead people.

 

[jaguarskx]

In review of this entire thread I have come to realize that I did not provide any reference regarding my statement about the Trinity's Radeon HD 7660D performance being a little better than the Radeon HD 5570. Below is a link to Anandtech's performance review and a few random charts from that review.

http://www.anandtech.com/show/6332/amd-trinity-a10-5800k-a8-5600k-review-part-1/2

 

[mohit9206]

wow you guys are discussing so much about my 7750. thanks. anyways i am here to say that richland's integrered graphics will be NOT even be close to a 7750. the highest end desktop richland apu will have graphics matching the 6670 ddr3. there you go..now stoo discussing this any furthur

 

[mohit9206]

the bad thing about amd processors is they consume much more power than intel cpu without offering any performance advantage in games. overclocking amd cpu consumes even more power than stock. so amd needs to work hard on making their architecture more efficient. amd graphics cards on the other hand are perfect.
so for OP if you are concerned about power consumption go with Intel.
as for 8 cores of amd will outperform 4 cores of intel, its wrong. by the time most games start using all 8 threads both cpu will be ancient. so intel wins this one ..