2017 13" nTB 640 iGPU vs TB 650 iGPU

[sakabaro]

I wonder how the 640 iGPU found in the 2017 nTB 13" MacBook Pro compare with the 650 iGPU found in the 13" Touch Bar models.

Some people seems to claim, it should be negligible 5-10% differences, some other seems to say the difference is bigger showing a 25% difference in a few Windows benchmarks.

Anyone has more information or was able to run an actual mac benchmark on those?

 

[NBAasDOGG]

I wonder how the 640 iGPU found in the 2017 nTB 13" MacBook Pro compare with the 650 iGPU found in the 13" Touch Bar models.

Some people seems to claim, it should be negligible 5-10% differences, some other seems to say the difference is bigger showing a 25% difference in a few Windows benchmarks.

Anyone has more information or was able to run an actual mac benchmark on those?

You'll be shocked if I tell you that the nTB 13" scores higher in Geekbench 4 when compared to 13" TB!

 

[sakabaro]

You'll be shocked if I tell you that the nTB 13" scores higher in Geekbench 4 when compared to 13" TB!

Yeah, I saw that. Crazy!

If the results are similar for graphics performance, that will be so great! (I dislike the touchbar so bad)

 

[NBAasDOGG]

Yeah, I saw that. Crazy!

If the results are similar for graphics performance, that will be so great! (I dislike the touchbar so bad)

Unfortunately, Geekbench is not always representing real performance. It's nice to wait and see CineBench and video editing performance. In reality, the 13" TB caries a 28W chip with higher transistor count and more GPU cores (Iris 650). This is supposed to be around 10-15% faster than the 15W chip inside 13" nTB like we saw in MacBook Pro 2016.

If my assumption is wrong, I ques the MacBook Pro nTB is now an affordable and fast machine

 

[Tanax]

Yeah, I saw that. Crazy!

If the results are similar for graphics performance, that will be so great! (I dislike the touchbar so bad)

You don't seriously believe the nTB is better than the TB version just because a benchmark says so?
People are reading way too much into this whole benchmark thing. Compare it with real-world tests instead and you'll see that the TB is faster. By how much though, I can't answer.

 

[NBAasDOGG]

You don't seriously believe the nTB is better than the TB version just because a benchmark says so?
People are reading way too much into this whole benchmark thing. Compare it with real-world tests instead and you'll see that the TB is faster. By how much though, I can't answer.

He doesn't believe, he said "if"...

 

[Samuelsan2001]

The 650 is about 20% faster in synthetic tests.

see here for comparison.
https://hardrealm.com/en/video/Iris-Plus-Graphics-650-vs-Iris-Plus-Graphics-640

 

[sakabaro]

The 650 is about 20% faster in synthetic tests.

see here for comparison.
https://hardrealm.com/en/video/Iris-Plus-Graphics-650-vs-Iris-Plus-Graphics-640

Cool. Thanks! I also wonder OpenCL performance, that will closer to real life performance on MacOS I suppose.

 

[jgbr]

It's cleary a trade off between models. Worth noting a faster CPU option in the nTB with lower Wattage. Can not overlook the cost of the touchbar which you could fit in more ram or the better CPU--which would change performance again!

 

[TechZeke]

Unfortunately, Geekbench is not always representing real performance. It's nice to wait and see CineBench and video editing performance. In reality, the 13" TB caries a 28W chip with higher transistor count and more GPU cores (Iris 650). This is supposed to be around 10-15% faster than the 15W chip inside 13" nTB like we saw in MacBook Pro 2016.

If my assumption is wrong, I ques the MacBook Pro nTB is now an affordable and fast machine

I agree with you about Geekbench, but the higher watt CPU is meaningless unless you are constantly taxing the CPU to its limits. MacBook Airs have been keeping up with the 13" MacBook Pros for a half a decade now. Even still, if you are constantly maxing the CPU you probably already looking at the 15" model. I guarantee for doing the vast majority of tasks, the 15W CPU is fine.

 

[sakabaro]

I agree with you about Geekbench, but the higher watt CPU is meaningless unless you are constantly taxing the CPU to its limits. MacBook Airs have been keeping up with the 13" MacBook Pros for a half a decade now. Even still, if you are constantly maxing the CPU you probably already looking at the 15" model. I guarantee for doing the vast majority of tasks, the 15W CPU is fine.

Even compiling in XCode doesn't take that long, I would say it's never hitting full CPU for more than a few minutes.