Intel plans to launch the first processors based on its 14nm “Broadwell” architecture before the end of 2014. The first Broadwell chips will be part of the Intel Core M processor family and they’ll be aimed at mobile devices including tablets and notebooks.
Next year we should see a wider range of chips for notebooks, desktops, and other devices. But Intel is focusing first and foremost on the mobile space — which makes sense. Broadwell chips should offer better performance while using less power than today’s Haswell processors, which could make them good choices for thin and light computers.
These are premium chips though: don’t expect a Broadwell processor to show up in a $200 tablet anytime soon. That’s a space that will continue to be dominated by ARM and Intel Atom chips.
Intel has shared some details about the upcoming Broadwell chips on its website, but for some seriously detailed analysis, check out articles from AnandTech and Ars Technica.
In terms of CPU performance, a 5th-gen Intel Core “Broadwell” processor will be just about 5 percent faster than an equivalent 4th-gen Intel Core “Haswell” processor running at the same clock speed. But there should be a more significant improvement in graphics performance, with a 20 percent boost in computer power and up to 50 percent better sampler throughput.
We should see big boosts in Intel QuickSync and Video Quality Engine performance, among other things and there will be initial support for hardware-decoding of H.265 video.
Intel’s new technology could also make it possible for chips to run at higher speeds without using more power — which means that you could eventually see more than a 5 percent boost in CPU performance in real-world applications. At launch though, Intel is hoping to get its Core M chips into small, low-power, devices such as fanless 2-in-1 tablet/notebok systems which measure as little as a third of an inch thick.
The chip maker hasn’t provided clock speeds, TDP values, or some other important details for its Core M processors yet. But with the first chips set to launch by year’s end, it won’t be long before the picture is complete.
If 14nm yields are bad, the highest margin parts will ship first, as they will still make money (even if half the wafer doesn’t yield). The shipping of the low cost parts shows that the yields are good for a particular process.
I am not a laptop user, nor a high-end tablet user. I wish Intel the best with broadwell, but don’t expect to buy one. When is the new Nexus 6/7/8 going to be announced? I really need a new tablet. Come on Google… hurry up and take my money!
Would you be opposed to a Nexus tablet running on an Intel Atom?
I just returned my Dell Venue 8 (2014 with the z3480) yesterday as the performance and battery life was worse than my 2013 Nexus 7. Android hardware fragmentation is a real issue as many phones/tablets need to have updates disabled to ensure that functionality is not suddenly broken. I will only use Nexus as it is most likely to be tested before app updates are rolled-out. I really wish Google would stick with Qualcomm… too bad they dropped the ball with 64bit SOCs being too far out.
Wouldn’t a Nexus running on an Atom be just as well tested? Plus, Intel actually provides open source drivers for their chips whereas ARM vendors like Qualcomm only provide binary drivers and can stop supporting it at anytime (ie. the case where you mention things breaking on Android updates especially when the new Android version uses a new kernel).
Faster while consuming less power sounds good to me. What kind of devices would this go into? Right now we have mostly Atom tablets and then jump to i3/i5/i7 tablets like the Surface Pro. This sounds like it would be somewhere in the middle? A tablet nearly as powerful as a Surface Pro yet thinner and fanless would be nice.
Maybe it’ll eat into the 8″-10″ screen market. Maybe we’ll see Atoms go into smaller devices or Core M 8″-10″ devices will have a big premium on them.
core m seems interesting, just imagine the umpc market had this product been around at the time.
I would have been happy with Bay Trail for the UMPC market. Right now, I want the UMPC market to just comeback (not just 8″ tablets).
I’m eyeing the DragonBox Pyra right now. I hope it has a better launch than the previous device (all issues has been fixed now though).
really dont wanna poopoo DragonBox Pyra but arm is just bad move imo, especially a dated ti part.
I like the Pyra but I think it’s ridiculous that you have to have those gaming controls on the device.
You can play emulated games just fine without the joysticks and it makes their device seem like a toy and thus less justifiable in a corporate environment.
A Pyra would be a wonderful network tool or for field engineers who are tired of dealing with crappy tablets and their slow keyboards.
the game controls are what should draw you to the pandora…else a small netbook maybe a better fit.
There is no netbook with that pocketable size. I wish there were.
true only if umpcs didnt die…
They died in part because the companies that sold them were not smart enough to tell the media that it’s impossible to type with 10 fingers so you should adjust and use fewer.
I have an HP Jornada 728 and I can touch type (i.e. without looking) about 50 wpm on it with 8 fingers. You simply cannot do that on any phone or tablet.
I agree. Having used several ARM SBCs, I know how much they’re a pain to use and keep updated (if you can update at all). I’ve been moving towards Bay Trail SBCs recently. So much easier for a non-Linux kernel dev like myself
As for the form factor (physical keyboard and mouse), I’m willing to put up with the whole ARM and Linux issues. I’d get rid of the joysticks though. It looks ugly too but oh well.
I did read somewhere that they considered going with an Intel Atom before choosing the TI. I’m not sure what were the reasons though. I’m eying the device but I’m not following it that closely (at least not until they’re close to shipping).
I think ed did contact intel but for some reason it was just unfeasible to build it based on atom.
Comments are closed.