GPU vs CPU Processing Speed

[iRun26.2]

If I upgrade to a faster core M processor (1.2GHz or 1.3GHz vs 1.1GHz) will the internal GPU be also be that much faster?

 

[mikeo007]

Barely noticeable. The GPU isn't much of a performer in these Core M chips, so don't expect much out of it.

 

[Serban]

The HD5300 it's almost equal to the 2012 HD4000

 

[iRun26.2]

Barely noticeable. The GPU isn't much of a performer in these Core M chips, so don't expect much out of it.

But does the GPU clock speed go up proportionally to the CPU clock speed?

 

[KrisLord]

But does the GPU clock speed go up proportionally to the CPU clock speed?

These Intel processors have more configuration options than before. They don't use a fixed spec, so we could find all 3 macs use the same processor model but with minor changes to CPU clock. As they're so customisable, it's not possible to say whether the GPU will be faster.

The below link shows the Core M range.

http://ark.intel.com/products/series/83614/Intel-Core-M-5Y-Mobile-Processor-Series

 

[T5BRICK]

These Intel processors have more configuration options than before. They don't use a fixed spec, so we could find all 3 macs use the same processor model but with minor changes to CPU clock. As they're so customisable, it's not possible to say whether the GPU will be faster.

The below link shows the Core M range.

http://ark.intel.com/products/series/83614/Intel-Core-M-5Y-Mobile-Processor-Series

According to the information you linked, the CPUs that I suspect Apple will be using are the 1.1GHz/2.6GHz M-5Y51 and 1.2GHz/2.9GHz M-5Y71. They both list their GPU base clock at 300MHz and the max clock as 900MHz.

http://ark.intel.com/products/84669/Intel-Core-M-5Y51-Processor-4M-Cache-up-to-2_60-GHz

http://ark.intel.com/products/84672/Intel-Core-M-5Y71-Processor-4M-Cache-up-to-2_90-GHz

 

[crsh1976]

The HD5300 it's almost equal to the 2012 HD4000

Synthetic benchmark averages put the HD 4000 well above the HD 5300 - well, that's relative, they're both low-end GPUs (the type of low end where slow and slower are hard to tell apart).

 

[iRun26.2]

Synthetic benchmark averages put the HD 4000 well above the HD 5300 - well, that's relative, they're both low-end GPUs (the type of low end where slow and slower are hard to tell apart).

Can you provide a source for that, please?

 

[crsh1976]

Can you provide a source for that, please?

Here's a rough comparison: http://www.videocardbenchmark.net/compare.php?cmp[]=2921&cmp[]=2

As I said, it's relative, it's not like the HD 4000 is tons better, more like they're both adequate for basic use (2D stuff, video, etc) and both suck bad at anything intensive like 3D gaming.

 

[dusk007]

http://www.notebookcheck.com/Intel-HD-Graphics-4000.69166.0.html
http://www.notebookcheck.com/Intel-HD-Graphics-5300.125574.0.html

They are actually pretty much equal.
Compare the snthetics like 3dmark cloud gate but also take a good look at gaming benches. The thing is with these integrated GPUs they need to take away power from the CPU to fuel themselves. So benchmarks that disproportionately put load on just the GPU in an unrealistic way will not show realistic performance estimates. Therefore look at games they are more realistic and also show when CPU limitations make the weak gpu meaningless anyway.

Regarding the initial question. Actually a faster cpu clock might hurt the gpu speed. There is 4.5 W total TDP. If the CPU is allowed to clock higher the gpu clock speed has to give. They might be higher binned chips which perform better in any case but if they aren't the gpu will just be as fast or even slower. What is certain is that it will never be faster on the faster clocked chip.