IBM and ARM have extended their partnership to develop chips using 14-nm processing technology.
Monday IBM announced a partnership with UK chip developer ARM to develop 14-nm chip processing technology. The news confirms the continuation of an alliance between both parties that launched back in 2008 with an overall goal to refine SoC density, routability, manufacturability, power consumption and performance.
For the uninitiated, the idea is to build smaller, faster chip designs that provide better power management, resulting in longer battery life and better multimedia support than the current crop of ARM-based chips. With the help of ARM's design team, IBM will take ARM's intellectual property (IP) and cram it into IBM's miniscule manufacturing processes.
"ARM’s Cortex processors have become the leadership platform for the majority of smart phones and many other emerging mobile devices," said Michael Cadigan, general manager, IBM Microelectronics. "We plan to continue working closely with ARM and our foundry customers to speed the momentum of ARM technology by delivering highly advanced, low-power semiconductor technology for a variety of new communications and computing devices."
Through the previous ARM/IBM collaboration on the 32-nm and 28-nm, ARM has already delivered eleven test chips that provide concrete research structures and early silicon validation. However, using a 14-nm manufacturing process is quite a drop when compared to ARM's current crop of Cortex processors used in Nvidia's Tegra 2 and Samsung's Hummingbird chips, both of which are using 45-nm technology.
"IBM has a proven track record of delivering the core research and development that is relied upon by major semiconductor vendors worldwide for their advanced semiconductor devices. Their leadership of the ISDA alliance, which features a diverse set of top-tier companies as members, is growing in importance as consolidation trends in the semiconductor manufacturing industry continue," said Simon Segars, EVP and general manager, ARM physical IP division. "This agreement will ensure we are able to deliver highly tuned ARM Artisan Physical IP solutions on advanced ISDA process technologies to meet the early time-to-market our customers demand."
With the 32-nm and 28-nm samples currently out in the field for testing, it's uncertain when we'll see the first samples of the 14-nm process in action. Both ARM and IBM did not offer a projected "availability" date, so stay tuned.
I think that's for high-power devices, or IBM didn't hear about that lol.
It's so strange that we're getting so close to the absolute minimum our manufacturing technology can reach, I think within the next 5-10 years we'll be seeing chips with circuitry created at the atomic level, if it isn't already done that way right now.
From what I understand, the limit is nearer to 3 nm, at which point quantum tunneling allows electrons to jump the barriers between circuits. At that point, I imagine our current technology path for computer design will reach a dead end.
Of course, quantum tunneling isn't necessarily a barrier to our progress. Already, since decades ago, scientists and engineers have been putting forth ideas to use quantum tunneling to our advantage to build 3-dimensional processing units that would have even more performance potential than our current designs!
I thought this was ARM and IBM.
I remember the same thing. Back when we were still measuring processing technology in micrometers, it seemed everyone was saying the limit was at 0.1 µm. I remember being very confused when we moved beyond that without hearing anything more about it.
The limitations we'll be facing in the near future seem a bit more fundamental than whatever difficulties were being discussed 15 years ago, though. I don't imagine we'll be going much smaller for a long time once we reach that point. There are always other routes for progress, however.
I'm pretty sure the Tegra 2 is manufactured at 40 nm.
Simple designs scale down easier, but more complex ones are tricky. You can't just shrink a die and use the very same mask.
Though it would be awesome cool for that. Might actually be workable in a laptop or phone.
I think that IBM/ARM is pushing hard this way to keep Intel out of the mobile CPU biz, or at least in AMD's old shoes. Might be interesting to see how this develops.
Even if they won't be able to go further then 1,3,10 nm for a while, they will find ways of improving the manufacturing process and lowering the price so we'll be able to cope for a few more decades with multiple cores, and gpus, until quantum computers kick in.
The neighborhood of 10 nm is the foreseeable limit at this point in time.
Quantum effects must be mastered before going smaller than 10nm. All bets are off regarding rate of progress for CPUs once 10nm is reached.
Few more decades??? Hardly! Intel expects to achieve 11 nm in the year 2015. That is only 4 years away. After that there will be no more die shrink progress without amazing new science. I think we are about to see computer evolution take the form of fiber optics and chip consolidations(merging cpu/gpu, for instance) more common than die shrinks.
So if Sandy Bridge can do 5GHz on air just by switching to 32nm, then 5nm (assumed limit) should be able to do around 10GHz (8GHz to 15GHz is my guess)? Then Carbon will allow us to go near 40 to 75GHz on a 5nm process. So if we can get the carbon process down and just switch the silicon methods over to it (which is completely concievable in the next 15 years), then we'll be running octocore (I'm totally guessing overhead makes more than 8 threads impractical) processors at 50GHz. That should give us about a 40x boost over OC'd Sandy Bridge quad cores even if they were running multi-threaded apps (50GHz/5GHz*200%[Improvement per cycle efficiency]*2[8 cores vs 4]).
The point being--we've got a lot of improvement in the conceivable future without switching to an as-yet-unknown manufacturing process.
No at 12nm you get electrons flowing through PNP gates aka electrons going bye bye into well nowhere
I'm so happy they're wrong.
We keep pumping billions of dollars with countless engineers. Maybe we will reach a limit soon, maybe we won't. For the time being, I'm more than satisfied just to know that they're trying.