Surface Pro 1 for a Surface 3?

[Geodude074]

I wouldn't say that the SP1 "blows away" the Surface 3 in terms of performance. The Core i5 in the SP1 might be more powerful but the Atom x7 in the Surface 3 has the potential to offer more powerful graphics. Early benchmarks (not Geekbench) are putting the Atom x7 CPU right up there (actually a bit ahead) with the Core M in the unreleased MacBook. So the differences between the SP1 and SP3, in terms of performance, could be marginal if they are even noticeable at all. Don't forget that the SP1 runs a Core i5 that's now two generations behind instead of one like the SP2/SP3.

Where are these benchmarks? I'd like to see them. The benchmarks I've seen show that the Core M outperforms both the Core i3 AND the Core i5 in the Surface Pros.

For the Atom x7 to be close to the Core M would mean that it's more powerful than the Core i5.

System Performance - The ASUS Zenbook UX305 Review

 

[onlysublime]

The Core M processors (based on Broadwell) are definitely NOT faster than the Haswell i5 or i7 processors. The Core M processors are designed to slot in-between the Atom series and the Core series. Remember that the Haswell i5 and i7 processors used in Surface Pro machines are the U series which is Intel's performance series. The Y series is designed for low power consumption. There are Core i5 and Core i7 processors that are Y series but none were used in Surface Pro machines.

The Haswell Core i3 is a Y series processor. The Core M processors fit under the Y series of processors which stress low power consumption. The Core M processors are designed to be fanless. So if you're expecting Core performance from the Core M processors, you will be disappointed.

The Core M processors (again, Y series processors) are designed to replace and be faster than the Y series Haswell processors. The only Y series Haswell processor used by Surface Pro 3 is the Core i3. The Core M is not designed to replace the U series Haswell processors which is what is in the Surface Pro 1, 2, and 3. The Surface Pro 2 uses Haswell U series Core i5. The Surface Pro 1 uses Ivy Bridge U series Core i5.

Ultimately, if you're looking for ultimate performance, you're going to wait for the U series of Broadwell or Skylake. Which will most likely still be branded the Core processors (i5, i7). Since Broadwell was so late, it ran into the debut of Skylake. And Skylake is the tock of Intel's tick-tock release cycle. Skylake is where the big improvements in performance will be.

As for the benchmarks listed by Geodude... they are misleading as well. The only processor that the Core M Broadwell beats in terms of Surface Pro machines is the Core i3 which is the Haswell Y series processor. Of course, the Core M (Y series) is designed to be faster than the processor it replaces (Haswell Y series).

Core i5 or i7 Haswell U series is still faster than Core M Broadwell (and Core i5 Ivy Bridge is faster as well). The PCMark benchmarks have to be carefully used because they are not pure CPU/GPU tests and include other factors such as SSD or eMMC speeds.

 

[colinkiama]

No, it's not the same man. Surface pros really can replace your laptop. They can run (almost) everything but the surface 3 is actually an amazing tablet that COULD replace your laptop. It's lacking in power. But windows store apps, office and old games should run fine. It really depends what you want though. If you want a iPad killer then the surface 3 is perfect. If you want an ultrabook in the form factor of a tablet, the surface pro 3 is perfect.

 

[P_Devil]

Where are these benchmarks? I'd like to see them. The benchmarks I've seen show that the Core M outperforms both the Core i3 AND the Core i5 in the Surface Pros.

For the Atom x7 to be close to the Core M would mean that it's more powerful than the Core i5.

System Performance - The ASUS Zenbook UX305 Review

They're everywhere. I think you are also forgetting that the Core M in the new MacBook isn't the same as the Core M in the Yoga Pro 3 or the Core M in the Asus. The main issue with the Core M is that CPU performance is heavily dependent on the power supplied to it. So a 1.5GHz Core M that is under powered, to save battery life, can be outperformed by a 1GHz Core M that is being fed more power. Just look at the CPU Mark scores for the MacBook and Surface 3. The MacBook's are somewhere around 1300 for single-threaded and 3000 for multi-threaded whereas the Surface 3 is 1000 for single-threaded and 3400 for multi-threaded.

 

[dirtyvu]

P devil. Just stop. You make a lot of nonsensical statements.

 

[Geodude074]

As for the benchmarks listed by Geodude... they are misleading as well. The only processor that the Core M Broadwell beats in terms of Surface Pro machines is the Core i3 which is the Haswell Y series processor. Of course, the Core M (Y series) is designed to be faster than the processor it replaces (Haswell Y series).

Core i5 or i7 Haswell U series is still faster than Core M Broadwell (and Core i5 Ivy Bridge is faster as well). The PCMark benchmarks have to be carefully used because they are not pure CPU/GPU tests and include other factors such as SSD or eMMC speeds.

Except for these benchmarks, where Core M clearly pulls ahead of Core i5.





Overall, yes Core i5 is faster than Core M. But the performance difference isn't that drastic. You're looking at about roughly 10-15% better performance from a Core i5. That trade-off is definitely worth it though for longer battery life, fanless design, and a considerably cheaper pricepoint aka Asus Zenbook UX305 for $700.

 

[dirtyvu]

Can you read? Pcmark is not a pure cpu/gpu benchmark. It is a performance benchmark. What it illustrates is how important the motherboard chipset is, how important drivers are, how important the system design is, and how important the ssd, hdd, or emmc is. But if you just want to compare CPU, it's not a good test. Just like you can stick a Ferrari engine in a Mustang. You aren't going to get Ferrari performance.

 

[onlysublime]

Yes, just changing drivers will cause changes in your PCMark results. Sometimes a new driver causes a DECREASE in performance. This doesn't happen with a pure CPU test. Of course, with GPU, the drivers are very important as the GPU has to interface to the rest of the system.