[SOLVED] Choosing Laptop CPU Advice

[Chittychittybangbang]

Hello, i want to ask about which one are better, an I5 1135g7 (with mx450), i7 1165g7 (also mx450) laptop or a ryzen 7 5700u (igpu) laptop.
Im going to use this laptop for college (im majoring chemistry, but i want to also learn coding), so probably medium gaming-medium workload.

The pricing:
Ryzen < i5 << i7

 

[cdrkf]

It's an interesting choice... the Ryzen 7 is the fastest processor of the bunch in multi thread, and similar in single thread so it should be the better option for work (especially things like code compilation). The Ryzen 7 is also the more power efficient processor so if the laptops have similar size batteries you will get more battery life out of it than the Intel based laptops. The graphics performance of the igpu isn't nearly as good as an mx450 though (the MX450 is ~ 50% faster) so either Intel laptop will perform better in games.

 

[Chittychittybangbang]

What about the difference between the i5 and i7? Should i even consider the i7 (the price difference is pretty far)

 

[faalin]

Two things you should look at when buying a laptop the CPU and GPU. Get the best you can afford as its very hard or cant be replaced once you bought it. Memory and Hard drive space are normally the only things that can be upgraded on a laptop.

 

[SamirD]

What about the difference between the i5 and i7? Should i even consider the i7 (the price difference is pretty far)

The single thread performance of the i7 is better, but the i5 is pretty solid too. Keep in mind that if you have workloads with more threads, the ryzen will be better, but for single thread performance, the i5 will be faster and the i7 will be much faster:
https://www.cpubenchmark.net/compar...-1165G7-vs-AMD-Ryzen-7-5700U/3830vs3814vs4156

 

[cdrkf]

The single thread performance of the i7 is better, but the i5 is pretty solid too. Keep in mind that if you have workloads with more threads, the ryzen will be better, but for single thread performance, the i5 will be faster and the i7 will be much faster:
https://www.cpubenchmark.net/compar...-1165G7-vs-AMD-Ryzen-7-5700U/3830vs3814vs4156

That isn't true.. Passmark tends to favor Intel parts, however across a broader selection of appliactions the 5700U is about on par with Tiger Lake in single thread, and destroys it in multi thread:

View: https://www.youtube.com/watch?v=edpdmbmYOj0&t=1031s


16:20 in that vid comparing the 5800u vs 1185G7 (both slightly higher clocked versions of these chips).

 

[geofelt]

For college, consider other factors.
If you will be lugging it around, look for a lightweight unit.
Probably 14"
Look for a unit with a 400-500 nits screen that can be viewed in difficult lighting conditions.
See that the battery life will be sufficient for your purposes.

Do not buy anything without a ssd; sata vs pcie matters little.
It may be good strategy to buy a unit with minimal ram and drive, planning to upgrade it yourself.
Makers charge 2x the cost of such upgrades.

Now, to the processor, any of your choices are going to be capable enough.
The 4 core/8 thread I5 1135g7 should be plenty.

Do not go overboard on discrete graphics.
Your first job is college, not gaming.

I see many complaints about gaming laptops not performing well.
Usually gaming while plugged in.
One common cause is thermal throttling.
Laptop coolers must, of necessity be small and light.
They are also relatively underpowered.

Seems to me that mx450 will do the job.

 

[SamirD]

That isn't true.. Passmark tends to favor Intel parts, however across a broader selection of appliactions the 5700U is about on par with Tiger Lake in single thread, and destroys it in multi thread:

View: https://www.youtube.com/watch?v=edpdmbmYOj0&t=1031s


16:20 in that vid comparing the 5800u vs 1185G7 (both slightly higher clocked versions of these chips).

Every bit of code is optimized here or there. I've not had passmark's results deviate from my own real world observations, so it's generally what I use to compare cpus. However, there is always manipulation and corruption so I wouldn't doubt if there were biases that have been paid for--it wouldn't be the first time.

Nevertheless, even passmark is just a 'rough judge' at best as particular applications can be optimized for one platform or another or even one cpu or another. Only way to truly test is run the same workload on 2x systems.

 

[cdrkf]

Every bit of code is optimized here or there. I've not had passmark's results deviate from my own real world observations, so it's generally what I use to compare cpus. However, there is always manipulation and corruption so I wouldn't doubt if there were biases that have been paid for--it wouldn't be the first time.

Nevertheless, even passmark is just a 'rough judge' at best as particular applications can be optimized for one platform or another or even one cpu or another. Only way to truly test is run the same workload on 2x systems.

With laptops it's very hard to make a true apples to apples comparison - the chassis / cooling solution design can have a huge impact on the overall performance profile (there was a vid the other day with LTT looking at an Intel reference laptop with the new Tigre Lake 8 core part - the 'i9' was slower than the same gen 'i7' laptop they had been sent by Gigabyte due to the latter having a markedly better cooling solution).

In the case of the OP - the best bet would be to look at reviews for the specific laptop models they are considering.