ARM, TSMC Announce 7nm Chip Collaboration

By Lucian Armasu
published

ARM and TSMC announced that they will collaborate on making 7nm chips a reality in the next few years. The two companies are already working together to create 16nm and 10nm chips.

IBM was the first to announce the creation of a 7nm chip, although the innovative processes it used to create it also meant that mass production wouldn’t be possible for a few more years, due to the high cost. Chances are that IBM’s 7nm chips could arrive sometime in 2018, or in 2019 at the latest.

Intel has already delayed its 10nm chip production to the second half of 2017, which means its 7nm chips won’t arrive until late 2019, or even early 2020. That gives IBM and other companies the opportunity to surpass Intel in cutting-edge process technology for the first time.

It’s not clear when TSMC will be mass-producing 7nm chips. However, knowing that its 10nm chips are likely to appear early next year, then chances are that its 7nm chips will be ready sometime in 2019, potentially surpassing Intel with quicker production of 7nm chips, too.

“TSMC continuously invests in advanced process technology to support our customer’s success,” said Dr. Cliff Hou, vice president, R&D, TSMC. "With our 7nm FinFET, we have expanded our Process and Ecosystem solutions from mobile to high performance compute. Customers designing their next generation high-performance computing SoCs will benefit from TSMC’s industry-leading 7nm FinFET, which will deliver more performance improvement at the same power or lower power at the same performance as compared to our 10nm FinFET process node. Jointly optimized ARM and TSMC solutions will enable our customers to deliver disruptive, first-to-market products," he added.

ARM also said that its chips will be optimized for TSMC's 7nm process node, to enable the "industry's lowest-power architecture across all performance points."

TSMC doesn’t seem to be using the extreme ultraviolet (EUV) lithography that IBM used for its 7nm chips, but the company is still researching it for smaller process nodes. The EUV lithography, which has a much smaller wavelength of only 13.5nm compared to the existing technology (193nm wavelength) will be necessary to design smaller and smaller chips in the future.

Lucian Armasu is a Contributing Writer for Tom's Hardware. You can follow him at @lucian_armasu. 

Follow us on Facebook, Google+, RSS, Twitter and YouTube.

Lucian Armasu
Lucian Armasu
Lucian Armasu is a Contributing Writer for Tom's Hardware US. He covers software news and the issues surrounding privacy and security.
11 Comments Comment from the forums
  • bison88
    That gives IBM and other companies the opportunity to surpass Intel in cutting-edge process technology for the first time.

    This is misleading. TSMC/Samsung's process isn't equal to Intel's, not in the slightest. Their 16nm/14nm is significantly larger design wise than Intel's 22nm. It's been known for a few years that both TSMC and Samsung are playing the node game to try and "steal" the lead from Intel. They are going to be behind for awhile to come.
    Reply
  • hannibal
    IBM may do it but when TSMC has been in shedule? Also the ability to make 7nm, and an ability to make 7nm as mass production is a different thing...
    And yep, the nm to nm comparison is not very easy because of huge differencies in overall production technology.
    But if the power usage goes down, it is a good thing to mobile devices!
    Reply
  • jasonelmore
    intel's 14nm still uses the 22nm interconnect. TSMC and Samsung's 14/16nm still uses the 28nm interconnect.

    The interconnect is the true indicator of manufacturing process complexity.
    Reply
  • jasonelmore
    Also IBM does not own a foundry.. they were bought out by global foundries.
    Reply
  • aldaia
    Things are not that simple. Once upon a time, a 0.1um process meant that you could measure something (the transistor length) and find that it was indeed 0.1um. Now we have processes called 14nm but there is nothing that is 14nm at all on them.

    There are no true indicators of manufacturing process complexity. Defining a node by a few dimensions of pitches (mainly M1P and GP) is an oversimplification. Transistor density, derived from GP X M1P, puts Intel’s 14nm at 37% denser than TSMC 20nm. The comparison of real chips between 20nm A8x and 14nm Core-M reveals A8x is 48% denser than Core-M.
    Reply
  • hdmark
    not that they will actually pass intel, but what happens if they did? I don't have much knowledge on this subject, but would this mean there would be desktop processors outside of intel/amd that compete? or is this mainly for mobile/tablet type devices?

    just trying to figure out the significance of all of this
    Reply
  • goodguy713
    Honestly the only real benefit I see coming from this is longer battery life on phones and laptops. But with only minor performance gains what's the point at some point they will have to take a different approach smaller will only work for so long. I'd be more interested in 7nm GPUs than CPUs especially considering my two year old droid maxx can do pretty much everything I need it too there's no reason to up grade other than just for the sake of getting a new phone ..
    Reply
  • Spectre694
    17669295 said:
    not that they will actually pass intel, but what happens if they did? I don't have much knowledge on this subject, but would this mean there would be desktop processors outside of intel/amd that compete? or is this mainly for mobile/tablet type devices?

    just trying to figure out the significance of all of this

    A large part of Intel's dominance and success is due to the fact that it has (until recently) been able to consistently stay at least a process node and usually some sort design change (finfet etc) ahead of it's competition. Those equal performance and power advantages. Even for Samsung GloFo TSMC to match Intel at a process would both be a first and put them on somewhat more equal footing.

    That and wafers have a fixed cost so smaller node equals more chips on the same wafer so long as yields are good.
    Reply
  • aldaia
    Maybe this article from EETimes (http://www.electronics-eetimes.com/news/tsmc-arm-aim-7nm-data-centers) shows some light. Some interesting quotes:

    ARM said it is preparing a generation of “future ARM technology designed specifically for data centers and network infrastructure and optimized for TSMC 7nm FinFET,”

    Last year, TSMC said it expects to start making 7nm chips in 2017.
    “ARM is very active in trying to get into data centers and 7 nm will be a key technology node for many of these activities…High production volumes are possible in 2019 and 2020,”
    Reply
  • alidan
    Honestly the only real benefit I see coming from this is longer battery life on phones and laptops. But with only minor performance gains what's the point at some point they will have to take a different approach smaller will only work for so long. I'd be more interested in 7nm GPUs than CPUs especially considering my two year old droid maxx can do pretty much everything I need it too there's no reason to up grade other than just for the sake of getting a new phone ..

    competitive pricing.
    do you think an i7 is really worth 300+$
    similar performance on a similar node kills all current price points because there is no "this is a premium product" advantage.
    Reply