I wonder how the benchmarks comparing the Snapdragon processor that doesn't have the Meltdown and/or Spectre CPU flaw will compare with the junk CPUs that do have it and require a software workaround that will have a stated 30% performance impact. I'm looking at buying a new laptop this year and I'm looking ARM.

Quote from: Blackbeard on December 11, 2017, 09:16:59
FYI, Ultrabookreview made a mistake, and updated one of the results. :p

CineBench R11.5 32-bit: CPU 1.50 cb, CPU Single Core 0.50 cb.

Looks like around Intel's Cherry Trail Atom performance in this benchmark.

Thanks, Blackbeard. The article has been updated. :)

FYI, Ultrabookreview made a mistake, and updated one of the results. :p

CineBench R11.5 32-bit: CPU 1.50 cb, CPU Single Core 0.50 cb.

Looks like around Intel's Cherry Trail Atom performance in this benchmark.

Quote from: Eduardo Lucas on December 10, 2017, 17:22:12
You don't even need core processors to trash this. Yoga Book with a two year old Atom already makes it almost useless. Lower power, higher battery life? Really? In a emulated environmnent? Who said the WHOLE key IA32-AMD64 library will be otimized to ARM? Who said they will be able to keep the CPU from not draining under constantly heavt workload?

Not to mention "Things will improve on the software side". They have been testing it for one year already, and are unable to score more than a atom X5 in Geekbench 4?

It clearly seems that Microsoft is promising something which is not really that visible today. This has a good chance of failing, in the same way that Intel failed with Android. 

We will reserve how the ARM fares when compared to the Atoms for some time. The core OS with all libraries HAS been ported to ARM. That is not surprising since Windows 10 always had ARM compatible libraries, courtesy, Windows 10 Mobile. The other statements seem to be presumptions. I mean, lets get this laptop tested fully, shall we?

You don't even need core processors to trash this. Yoga Book with a two year old Atom already makes it almost useless. Lower power, higher battery life? Really? In a emulated environmnent? Who said the WHOLE key IA32-AMD64 library will be otimized to ARM? Who said they will be able to keep the CPU from not draining under constantly heavt workload?

Not to mention "Things will improve on the software side". They have been testing it for one year already, and are unable to score more than a atom X5 in Geekbench 4?

It clearly seems that Microsoft is promising something which is not really that visible today. This has a good chance of failing, in the same way that Intel failed with Android. 

Quote from: digitalguy on December 09, 2017, 22:32:21
The confusion was created by ultrabookreview.... I am 99% sure he used Cinebench R15 which is expressed with those points, for instance my quad core i7 laptop (3rd gen) has 6.25 in R11.5 and 558 in R15 (both multicore scores). Clearly those are R15 scores, not R11.5... (unless as I said he forgot a dot after the first figure of each score, which I highly doubt)
It doesn't matter much which i7 you take, if you use another metrics the difference will be in the order of hundred times...
My point was just to say that this chip should compare to Core M... rather than a U series (let alone a quad core u series), since it's a fanless chip without special cooling (some fanless U chips exist, but have very special cooling solutions, like those from Microsoft or Acer)

PCWorld reports that they could not install Cinebench on the NovaGo. This could mean that they tried installing R15 but it did not install as the emulation layer cannot emulate x64 apps just yet. Therefore, the figures posted by UBR should have been from R11.5. Like anon said, this is not a 'direct' performance comparison to the 8550U. It was only to point out the anomaly in the reported scores. We are coming up with a more detailed analysis of the scores in comparison to existing CPUs, which should give a better perspective.

The confusion was created by ultrabookreview.... I am 99% sure he used Cinebench R15 which is expressed with those points, for instance my quad core i7 laptop (3rd gen) has 6.25 in R11.5 and 558 in R15 (both multicore scores). Clearly those are R15 scores, not R11.5... (unless as I said he forgot a dot after the first figure of each score, which I highly doubt)
It doesn't matter much which i7 you take, if you use another metrics the difference will be in the order of hundred times...
My point was just to say that this chip should compare to Core M... rather than a U series (let alone a quad core u series), since it's a fanless chip without special cooling (some fanless U chips exist, but have very special cooling solutions, like those from Microsoft or Acer)

Quote from: digitalguy on December 09, 2017, 18:23:00
Man, comparing a Snapdragon 835 to a quad core Core i7-8550U? Really? Come on...

You're confused. The comparison is because the reported numbers indicate ~200x better performance than an 8550u (so there must be a mistake in the data).

That said, I would like to see a direct comparison to a 8550u when it becomes available.

Man, comparing a Snapdragon 835 to a quad core Core i7-8550U? Really? Come on... This is a fanless device that works without much cooling. I would compare it to a Core M, and while the geekbench is lower than a Core M3, the Cinebench is higher than a any Core M and close to a core i5 of 4th-5th gen...
Now it's not clear if it's Cinebench R11.5 and they forgot the dot or R15 and they got wrong the version number, but either way this beats the core M on sustained load...