3DMarks Scores of Surface Pro 3 Models (i3/i5/i7)

[bkydcmpr]

I posted the scores on another forum, and copying it here.
i7 scores found on 3dmarks website, i5 and i3 scores are from Anandtech:
(why this forum doesn't support table?)
i3 i5 i7 Link
Fire Strike 506 642 722 http://www.3dmark.com/fs/2476013
Cloud Gate 3402 4130 4773 http://www.3dmark.com/cg/1953109
Ice Storm 32327 42129 43525 http://www.3dmark.com/is/2159115
Performance 847 851 1105 http://www.3dmark.com/3dm11/8530312

 

[be77solo]

Big question is going to be what the i7 scores are after a couple of repeated runs. I fear that will show the same throttling issues the i5 exhibits now, ie the follow up runs are quite a bit slower.

The SP2 by comparison remained around the same score.

 

[dleuen]

Very interesting. I'd be curious to see what the temperature readings are on the various devices during the tests. Wonder if the i7 throttles at the same level.

 

[megatronium]

Achievement unlocked!

 

[bkydcmpr]

Big question is going to be what the i7 scores are after a couple of repeated runs. I fear that will show the same throttling issues the i5 exhibits now, ie the follow up runs are quite a bit slower.

The SP2 by comparison remained around the same score.

no doubt the score will drop in repeated test. but i7 seems to be doing better than i5, anandtech said the i5 model couldn't finish the performance test without throttling and ended up a score very close to i3, while i7 apparently finished it with a decent score.

however I don't know the condition of the i7 tests, e.g. if it's undervolted.

 

[infer66]

Here are the SP2 i5 4200U benchmarks just for comparison.

3DMark 11 998
Ice Storm 42145
Cloud Gate 4537
Fire Strike 654

 

[Moonsurface]

Interesting and encouraging for me. What sort of real world application/processing other than gaming would you have to run to get to the sort of processing requirements that these tests have?

 

[efinimazda]

My SP3 i7 256GB scored P1035 using 3DMark 11. I have connected standby disabled and undervolted the dynamic CPU and graphics. Strange thing is that the driver firmware is not supported so the test in considered invalid... not sure if this is correct but I have the latest driver for the HD5000 installed.

 

[ctitanic]

Interesting and encouraging for me. What sort of real world application/processing other than gaming would you have to run to get to the sort of processing requirements that these tests have?

Only if you run some Virtual Machines and even in that case it depends on what VM. Android will make your SP3 to run hot. Hangouts on Chrome will make the SP3 run hot. Summarizing any bad coded software using the GPU can make your Surface to run hot.

 

[grumpy]

Summarizing any bad coded software using the GPU can make your Surface to run hot.

C'mon now. You really want to blame "bad" software for the SP3's throttling issues... The heart of the SP3's problems are an inadequate cooling solution brought about by MS's ill-fated desire to chase the iPad form factor. The software need not be "bad coded" to cause issues with the SP3. Any processor intensive applications will cause it to run hot.

And spare me the wrong tool for the job manta. Better engineering compromises would have resulted in better performance and no one would have been bothered by an extra 3mm other than those who obsess over edge on photo comparisons with the iPad. If your workflow is such that it does not include tasks that couldn't be handled by an IOS or Android device, then it is likely you will be unaffected by the crippled performance. Of course, if that is the case, a non-pro Surface would have been a better choice. The Surface Pro line is meant to be the tablet that can do more for more people and the SP3 was a step backwards.

 

[GreyFox7]

C'mon now. You really want to blame "bad" software for the SP3's throttling issues... The heart of the SP3's problems are an inadequate cooling solution brought about by MS's ill-fated desire to chase the iPad form factor. The software need not be "bad coded" to cause issues with the SP3. Any processor intensive applications will cause it to run hot.

And spare me the wrong tool for the job manta. Better engineering compromises would have resulted in better performance and no one would have been bothered by an extra 3mm other than those who obsess over edge on photo comparisons with the iPad. If your workflow is such that it does not include tasks that couldn't be handled by an IOS or Android device, then it is likely you will be unaffected by the crippled performance. Of course, if that is the case, a non-pro Surface would have been a better choice. The Surface Pro line is meant to be the tablet that can do more for more people and the SP3 was a step backwards.

Interesting opinion but if you're trying to increase the acceptance of the platform you want to target a wider audience not the few users that already bought in especially on a short product cycle. The people that don't care about a Lighter and Thinner tablet can continue to consume the SP2 while the larger majority of people that wont consider the SP2 are more likely to be attracted to the SP3. It's not all about you.

 

[ctitanic]

C'mon now. You really want to blame "bad" software for the SP3's throttling issues... .

No. I just pointing to a fact. Try Microsoft Sudoku. That's my example of bad coding making the Surface to run hot.