7-Way Linux Laptop Comparison From Sandy Bridge To Broadwell

[Ayceman]

That Toshiba i5 performed well (unsurprising given it's a 5200), but if it's like another I own (AMD this time), and it has an Elantech touchpad, you'd better steer clear of it. The Elantech driver is buggy - on some boots it doesn't detect it.

[tigerroast]

The Sandy Bridge laptop was surprisingly competitive! Of course, it's rather dismal in the power usage/efficiency department, but still. Fantastic for being the last Intel architecture that doesn't need a key to be supported by Coreboot (Libreboot, even).

[brauliobo]

thank you very much for this benchmarks, so basically the performance doubled from sandy bridge to haswell/broadwell

[Slartifartblast]

Some odd results there for the ASUS i7-4558U Xenotic benchmarks, I'd expect Iris 5100 graphics to perform much better than that. Was it some sort of driver issue or throttling ?

[schmidtbag]

Once again poor benchmark design - items description should be about real HW, not about manufactures name of products in graphs.

For someone who whines as much as you do, I'm surprised you didn't check your facts more thoroughly.

[Michael]

Once again poor benchmark design - items description should be about real HW, not about manufactures name of products in graphs.

It's not manufacturer names in product graphs. Just the vendor and then appended the CPU model, not the laptop model number.

What would make a sense, its a architecture comparison high end processors at same clock / or same TDP as did for example Anandtech too show technological progress for every Core i architecture generation.

If I had all that hardware, of course I would do such, this was just limited based upon the laptops I have available...

[bug77]

Unfortunately pts_Graph doesn't have an easy way at the moment for being able to auto-determine generations of a chip based upon the result identifier strings. Patches welcome though by anyone interested.

Maybe attaching some metadata would help?

These tests pretty much mirror my (and probably everybody else's) experience: I went from an i5-2500k to an i5-6600k and there's no discernible difference between the two.
The advantages are in rather unexpected areas: USB 3.1 ports (which are not a given, since they require an additional chip), M.2 slots, SATA3 across the board, PCIe 3.0 (useless to me, but probably useful for others) and, most surprisingly, the network seems to work better now (pages load better when going through a proxy). Another advantage for Linux users is that newer motherboards (newer than my trusty old P67 board) can update the UEFI from within the UEFI. Previously this was either tied to a Windows-only application or required copying an image to an USB stick first, making it as inconvenient as it gets to update if you were only running Linux.

[ruthan]

Once again poor benchmark design - items description should be about real HW, not about manufactures name of products in graphs.

What is the sense of the test, that show that there is some evolution in Hardware? - We know it. Are all those notebook for same budget at the time, what about displays and other components - they have to impact final price.
What would make a sense, its a architecture comparison high end processors at same clock / or same TDP as did for example Anandtech too show technological progress for every Core i architecture generation.

[Michael]

You should have used one color per generation. It's very arduous to read the diagrams otherwise.

Unfortunately pts_Graph doesn't have an easy way at the moment for being able to auto-determine generations of a chip based upon the result identifier strings. Patches welcome though by anyone interested.

[jf33]

You should have used one color per generation. It's very arduous to read the diagrams otherwise.