What's new

Broadwell "U" - class of SP4 (not Core-M)

Sorry to jump in the middle here. I posted a new topic in the general forum about the Core M and got redirected to this thread.

Are the reports that heat throttling the i5/7 SP3 to the performance level of the i3 untrue? I was set to buy either the i5 or i7 versions until I read several separate benchmark reports that the only real difference between all SP3 versions was storage space because of heat management issues. I ask because I was excited about the possibility that the Core M-5Y71 configured up to the 6W TDP in basically the same form factor as the SP3 would be able to run full throttle comfortably in a vented and active-cooled space. Wouldn't the 5y71 outperfrom the 4300U given heat management issues?

I found this forum after a few days of sporadic research on this topic, but everyone here seems to be really down on the Core M. I know all the devices on the market now that run the M are configured for 4.5W in a passive-cooled shell, but I was thinking the SP4 might be the first to juice it up to 6W so we could see performance like this: http://www.anandtech.com/show/8515/quick-look-at-core-m-5y70-and-llama-mountain

The 5y71 I am looking at: http://ark.intel.com/products/84672/Intel-Core-M-5Y71-Processor-4M-Cache-up-to-2_90-GHz
Pre-release statements and via their Reference Tablet (Which has a heavy heat sink back) Intel made it seem as though the Core-M would compare favorably with a Haswell i5 like in the SP3. to date nothing released by any vendor with a Core-M and benchmarked has shown to be anywhere near the SP3s performance.

As far as reports about the i3/i5/i7 having the same performance that is categorically untrue when looking across all benchmarks. An individual benchmark which only test one function may show one or the other better but taken as a whole the 3 fall in line performance wise. Depending on your specific workload you may or may not *notice* much difference but that's not an accurate indictor. I have seen a room of people at work watch a video demonstration of a performance problem and they failed to pick up on it until I pointed it out and the problem wasn't really that subtle. Benchmarks are good at making comparisons but you cant pin all your hopes or expectations on only one. On the other hand users perceptions & accounts can widely vary given the same or different performance levels making them an unreliable indicator.

Microsoft could have used the lower powered Y series parts instead of the U series and achieved less throttling but I think its safe to say they got better overall from the sometimes throttled i5U/i7U than they would have with the i5Y/i7Y.
 
Last edited:
Pre-release statements and via their Reference Tablet (Which has a heavy heat sink back) Intel made it seem as though the Core-M would compare favorably with a Haswell i5 like in the SP3. to date nothing released by any vendor with a Core-M and benchmarked has shown to be anywhere near the SP3s performance.

As far as reports about the i3/i5/i7 having the same performance that is categorically untrue when looking across all benchmarks. An individual benchmark which only test one function may show one or the other better but taken as a whole the 3 fall in line performance wise. Depending on your specific workload you may or may not *notice* much difference but that's not an accurate indictor. I have seen a room of people at work watch a video demonstration of a performance problem and they failed to pick up on it until I pointed it out and the problem wasn't really that subtle. Benchmarks are good at making comparisons but you cant pin all your hopes or expectations on only one. On the other hand users perceptions & accounts can widely vary given the same or different performance levels making them an unreliable indicator.

Microsoft could have used the lower powered Y series parts instead of the U series and achieved less throttling but I think its safe to say they got better overall from the sometimes throttled i5U/i7U than they would have with the i5Y/i7Y.

Good to know about the SP3 throttling. I'll still probably hold off at least until I see what the SP4 will offer, but I'll keep the 3 on the table.

I know the llama-mountain reference tablet had a specially designed aluminum back which allowed them to configure at 6w instead of 4.5w, which is what I was getting at with the Surface. From Tom's: "...OEMs have the option to configure all Core M processors to one of three TDP targets: 3 Watts, 4.5 Watts, or 6 Watts...an enclosure with active cooling might support a 6 Watt TDP...have to note that consumers will have to be aware that the processor model number alone no longer guarantees a set level of performance." So, each Core M processor is like a three-way light, only the switch is set in place at the factory. All the available Core M devices right now are configured for 4.5W, which is why I'm speculating about the SP4 which could be the first device to utilize the 6W configuration. We can't really use a device like the Yoga 3 as an example, because it's been limited by Lenovo in a way the SP4 wouldn't have to be.

http://www.tomshardware.com/news/intel-core-m-broadwell-benchmarks,27656.html
 
Good to know about the SP3 throttling. I'll still probably hold off at least until I see what the SP4 will offer, but I'll keep the 3 on the table.

I know the llama-mountain reference tablet had a specially designed aluminum back which allowed them to configure at 6w instead of 4.5w, which is what I was getting at with the Surface. From Tom's: "...OEMs have the option to configure all Core M processors to one of three TDP targets: 3 Watts, 4.5 Watts, or 6 Watts...an enclosure with active cooling might support a 6 Watt TDP...have to note that consumers will have to be aware that the processor model number alone no longer guarantees a set level of performance." So, each Core M processor is like a three-way light, only the switch is set in place at the factory. All the available Core M devices right now are configured for 4.5W, which is why I'm speculating about the SP4 which could be the first device to utilize the 6W configuration. We can't really use a device like the Yoga 3 as an example, because it's been limited by Lenovo in a way the SP4 wouldn't have to be.

http://www.tomshardware.com/news/intel-core-m-broadwell-benchmarks,27656.html
I have a feeling Intel may have throttled up that reference platform beyond 6w. :)

images2Y3G3HTZ.jpg
imagesF4PUL3DH.jpg
 
Back
Top