Core M 5Y70 Specifications
The first Broadwell-Y processor has started shipping with Ultrabooks like the Yoga 3 Pro. How does it perform?
Back in August of this year, Intel invited me out to Portland, Oregon to talk about the future of processors and process technology. Broadwell is the first microarchitecture to ship on Intel's newest 14nm process technology and the performance and power implications of it are as impressive as they are complex. We finally have the first retail product based on Broadwell-Y in our hands and I am eager to see how this combination of technology is going to be implemented.
If you have not read through my article that dives into the intricacies of the 14nm process and the architectural changes coming with Broadwell, then I would highly recommend that you do so before diving any further into this review. Our Intel Core M Processor: Broadwell Architecture and 14nm Process Reveal story clearly explains the "how" and "why" for many of the decisions that determined the direction the Core M 5Y70 heads in.
As I stated at the time:
"The information provided by Intel about Broadwell-Y today shows me the company is clearly innovating and iterating on its plans set in place years ago with the focus on power efficiency. Broadwell and the 14nm process technology will likely be another substantial leap between Intel and AMD in the x86 tablet space and should make an impact on other tablet markets (like Android) as long as pricing can remain competitive. That 14nm process gives Intel an advantage that no one else in the industry can claim and unless Intel begins fabricating processors for the competition (not completely out of the question), that will remain a house advantage."
With a background on Intel's goals with Broadwell-Y, let's look at the first true implementation.
Intel Core M 5Y70 Specifications
Newly branded as the Core M line, Broadwell-Y is meant to target the market looking for similar performance of today's ultra-low voltage Haswell processors but with longer battery life and more efficient designs. The Core M 5Y70 is the first, and flagship processor shipping this winter.
|Core M 5Y70||Core i5-4200U|
|ISA Extensions||AVX, SSE||AVX 2.0, SSE 4.2|
|Base Frequency||1.10 GHz||1.6 GHz|
|Max Turbo Frequency||2.60 GHz||2.60 GHz|
|TDP||4.5 watt||15 watt|
|Max Memory Size||16GB||16GB|
|Memory Types||LPDDR3, DDR3L||LPDDR3, DDR3L|
|Memory Bandwidth||25.6 GB/s||25.6 GB/s|
|Graphics||HD 5300||HD 4400|
|GPU Execution Units||24||20|
|GPU Base Frequency||100 MHz||200 MHz|
|GPU Max Dynamic Frequency||850 MHz||1.0 GHz|
|PCIe Support||2.0; x1 (6), x2 (4), x4 (3)||2.0; x1 (4), x4 (2)|
|Package Size||30mm x 16.5mm||40mm x 24mm|
|Die Size||82 mm^2||142 mm^2|
The comparison above pits the Core M 5Y70 against the Intel Core i5-4200U, the most popular Haswell processor used in Ultrabooks like the Yoga 2 Pro. Considering that Lenovo used the 5Y70 in the Yoga 3 Pro, it would seem that this is the replacement window that Intel has been trying to setup.
From a specification stand point, the Core M 5Y70 and the Core i5-4200U show some interesting dynamics. For example, while both are dual-core designs with HyperThreading (able to address 2 threads per core), the base clock speed on the 5Y70 is only 1.10 GHz, a full 500 MHz (45%) slower than the 1.6 GHz base clock on the 4200U. Notice however that the maximum Turbo clocks are the same - 2.60 GHz. Intel told us months ago that Broadwell-Y was designed to run at extremely low power and clock speeds when idle but still reach high performance and high clock speeds on demand, when needed, in short bursts.
Though I am still working on the ability to accurately log the clock speed of the Core M 5Y70 through Windows Performance Monitor, my experience has shown that a clock speed of 2.5-2.6 GHz for the CPU will only last 10-20 seconds in a best case condition. After that, the processor will throttle down to a lower clock speed (I've seen 2.2 GHz and then 1.8 GHz) to limit power consumption and maintain stability in a system that is incredibly thermally constrained.
The Intel HD 5300 graphics system has more potential compute power with 24 Execution Units (compared to the 20 EUs in the Core i5-4200U) but again the clock speed variance is a major factor in graphics and gaming performance.
Most importantly, note the significant TDP difference between these two processors. The Core M 5Y70 only requires a thermal solution designed for 4.5 watts. The Core i5-4200U required a 15 watt system design - more than 3x the potential heat to dissipate. Suddenly, Intel's claims of seeing similar performance levels on Broadwell-Y as on Haswell-ULT seem much more impressive (if they live up to that).
As an example, take a look at this photo comparing the internals of the Yoga 2 Pro and Yoga 3 Pro, the former of which uses the Core i5-4200U and the latter the Core M 5Y70.
Deconstructed Yoga 3 Pro (top), Yoga 2 Pro (bottom)
On the Yoga 3 Pro up top, look for the black heatpipe to the left of the silver shielding square - that is the location of the Core M 5Y70; that tiny little package. Now let that gaze move to the right and see the one small fan used for cooling the system. Compare that to the Yoga 2 Pro, with a much larger CPU package (see the centered silver frame outlining it) and the pair of larger fans required to keep the internals at operating temperatures. The amount of space saved, and power consumption lost, with the Core M 5Y70 and Yoga 3 Pro is astonishing.
But this isn't a review of the Yoga 3 Pro (that is coming soon), so let's dive into the specifics of our performance testing and see how the Core M 5Y70 stands up to Intel's claims.
Looks like both intel and
Looks like both intel and nvidia are pushing for more efficient chips for next year. When do we expect to see the Broadwell based ultrabooks for sale? They certainly look promising compared to older ivy bridge based chips.
This machine is for sale
This machine is for sale already.
If you are asking about the higher wattage Broadwell parts – sometime in early 2015.
For my use scenario I need a
For my use scenario I need a ultrabook with decent performance but longer battery life. I was considering the Dell XPS 13 but if the Core M has the similar performance levels as that of Haswell i5 and improved battery life(even additional 40-50mins matter for me while travelling) then I might wait for Dell to launch a SKU with this cpu instead.
The bit about wanting to see
The bit about wanting to see this processor in a laptop with a much higher capacity battery were my thoughts exactly.
How do the unit costs compare
How do the unit costs compare to the old Haswell parts? It sucks if consumers are paying the same for less performance, but more battery life.
I feel that today’s CPUs are
I feel that today’s CPUs are more than powerful enough for the average user. Especially at ~5 watts, performing close to a ~15 watt part is very impressive. This means that while watching youtube, netflix, browsing the web, or word processing will have great battery life, and good performance.
The tray price is $281 select
The tray price is $281 select models
Pricing information isn’t
Pricing information isn't available on these parts. Intel's Ark website doesn't list it either.
I, personally, care more about battery life at a usable performance level for my notebooks. I wouldn't pay MORE for it, but I would pay as much.
ryan is it possible for you
ryan is it possible for you to test against and Beema a6-6310 or a4-6410 apus. Also it seems a little meh for Intels next beast.
This is not even close to
This is not even close to what Intel’s Llama Mountain reference tablet was able to serve up performance wise. Why in the heck did Lenovo put more weight into that crazy hinge gizmo when they could have instead used their weight quota more towards a bigger battery and a better heat dissipating chassis? This has all the signs of bad engineering. There is even a fan internally for the 3.5W chip which shouldn’t be needed at all they knew what they were doing. What they should have done was configure the CPU to the optimal TDP of 6W like Intel recommends and designed accordingly instead of this giving us 3.5W sluggishness and then they would have gotten performance on par or faster than the i5 Haswell chips.
Real letdown, Lenovo! From what I heard, HP did the same thing with their new Envy X2 as well. Intel gave so many free development resources out to their partners and everyone instead opted to go this cheap and dirty route. Hopefully, Lenovo at least had the sense to follow Intel’s design guidelines (and not cut corners for once) with the ThinkPad Helix 2nd Gen due in a couple days from now.
For your information, you
For your information, you should note in the article that Lenovo is using the non-default 3.5 W TDP-down config for the Yoga 3 Pro. This is the reason for the somewhat lunderwhelming performance compared to what Intel had teased a couple months ago. Had Lenovo gone with the default 4.5W or the optimal 6W TDP configurations, the Pro 3 would be a much more powerful system. This is, of course, other flaws such as with thermal transfer in their design.
For your information, you
For your information, you should note in the article that Lenovo is using the non-default 3.5 W TDP-down config for the Yoga 3 Pro. This is the reason for the somewhat underwhelming performance compared to what Intel had teased a couple months ago. Had Lenovo gone with the default 4.5W or the optimal 6W TDP configurations, the Pro 3 would be a much more powerful system. This is, of course, ignoring other flaws such as with thermal transfer in their design.
I’m planning on trying out
I’m planning on trying out the new Lenovo Yoga 3 Pro and I hope it’s performance does not disappoint me. I’m not looking for a powerhouse, just something that can do moderate internet browsing and Steam in the background without stuttering- I have a desktop for everything else. I’m really sick of my old, hot, and heavy laptop from 2008.
$1400 for a laptop to do just
$1400 for a laptop to do just ‘some browsing’ seems rather expensive
On the one hand, comparing
On the one hand, comparing the 5Y70 to the i5 seems fair given the frequency range, IGP, and price, but then on the other hand it’s also fair to compare it to Atom given the power consumption. I’m not sure whether to be impressed or not. If anything, the graphics performance per watt seems more impressive than the CPU performance per watt.
I’m thinking that the Broadwell Atom (Cherry Trail) will be a more impressive upgrade over Bay Trail than Core-M is over Core iX.
There is no broadwell atom
There is no broadwell atom
Performance seems lower in
Performance seems lower in 3DMark than my Haswell Y powered Dell Venue Pro 11. This seems a bit off – was this tested on battery or plugged in? The Intel tablet chips have a fairly wide performance difference on battery. While I think both tests are valid, that’s a difference to be noted.
What I’m interested is
What I’m interested is actually about the future of the top processors for laptops that will come next year. 2 years ago we were here in our office between the option of a laptop with a i7-3820QM or a tower with a Xeon E5-2689. At this time we chose finally both, for understanding the differences. Because of the lower performance of the laptop, we still need the Xeon Workstation sometimes, but the laptop is almost ok for what we need. Do you think there is any chance that the new i7 Broadwell for laptops can have a similar performance of the 2 years old Xeon or Intel will focus on the lower TDP?
Yeah! On battery using the
Yeah! On battery using the most recent Intel drivers, my Haswell tablet gets 27k on Ice Storm! And it’s much higher plugged in. http://www.3dmark.com/3dm/4712367? I just don’t understand these results. i5 4210Y is somehow faster than its successor? Well then OK. I’ll take being able to do more, and charging every 5-6 hours instead of 7-8. That’s insane.
When you did your battery
When you did your battery test and you found that it was drawing 8 watts.. I have to admit that sounds VERY high. My G46VW idles at lower than 8 watts on battery and it’s a full fledged gaming laptop with a full voltage 3840QM CPU. I think some more software optimization needs to happen in Windows to get that number lower on the newer platforms.
Benchmarks returned for the
Benchmarks returned for the Broadwell 5Y70 from different sources seem to be completely inconsistent. FutureMark graphics results that appearing around September 2014 showed a performance of about 2X those listed here. I am not suggesting any error on you part. Indeed, Intel themselves seem to be responsible for this disarray. AnandTech, in their review of the 5Y70, appear to confirm that the reference unit tested used a 6W TDP SoC. The 4.5W chip will be the more likely choice (for reasons of useful battery life and simplified thermal management) in a production context for tablets and other low power computing devices.
One thing is certain – the 5Y70 is looking to be a less impressive part today, both in terms of outright performance and in terms of perf/watt, than it was thought to be just a few months ago. Things may improve with Skylake but ordinary is the best way to describe the Broadwell 5Y70.
Thank you for this review.
Thank you for this review. I’m currently deciding w/n to buy the new 5Y70 MacBook, and I was turned off by the poor Geekbench results that were leaked.
You more or less convinced now me with my usage patterns (no heavy continuous load), the 5Y70 should work well for my needs.