GlobalFoundries’ 12LP+ 12nm Node Promises 10nm-Class Power and Performance

(Image credit: Ken Wolter/Shutterstock)

GlobalFoundries (GF) on Tuesday announced the availability of a new addition to its 12 Leading Performance (12LP) platform, called 12LP+. The company claims it will feature a noticeable increase in performance and decrease in power and area. It also contains a low-voltage SRAM bit cell.

GlobalFoundries (GF) is promising much of the power and performance benefits of 7nm but at the lower cost of the 12LP platform. It also delivers a 15% increase in logic transistor density.

More specifically, GF said the 12LP+ FinFET process provides a 20% increase in performance or a 40% decrease in power over the base 12LP platform, (which itself provided 10% improvement over 16/14nm and a 15% improvement in logic area scaling). That's the same amount of improvement that TSMC claimed when comparing its 7nm process to 16nm, (although, TSMC also had 10nm before reaching 7nm).

In a statement, GF compared its new process to 7nm and also mentioned its lower cost: "As an advanced 12nm technology, our 12LP+ solution already offers clients a majority of the performance and power advantages they would expect to gain from a 7nm process, but their NRE (non-recurring engineering) costs will average only about half as much, a significant savings."

Another new feature is a 0.5V SRAM cell, which GF claimed is high-speed and low-power, and useful for data shuffling between memory and processor, such as in artificial intelligence (AI) applications. For AI applications, GF provides a design reference package and design-technology co-development (DTCO) services. There is also a new interposer for 2.5D packaging to facilitate high-bandwidth memory (HBM). Lastly, the company also said Arm has developed both Artisan physical IP and POP IP for AI applications, which will also be made available for 12LP.

GlobalFoundries is aiming for AI and cloud chips and said it already has several customers. Tape-outs are expected in the second half of 2020, and volume production is schedule for 2021.

  • NightHawkRMX
    Seems pretty good in theory.
    Reply
  • MasterMadBones
    Finally some good news out of GF again. This actually seems so close to 7nm that it may have been wiser to name it as such (or 10nm) from a marketing standpoint. Node naming has little to do with actual feature size these days anyway.
    Reply
  • hannibal
    So we can Expect to see new 12+ version of Ryzen IO chip next year instead of 7nm IO chip? Allso maybe new versions of some older cpu and gpu?
    Reply
  • bit_user
    MasterMadBones said:
    Finally some good news out of GF again. This actually seems so close to 7nm that it may have been wiser to name it as such (or 10nm) from a marketing standpoint. Node naming has little to do with actual feature size these days anyway.
    IIRC, TSMC delivered a 2x density increase between 14 nm and 7 nm. Really, it should've been more like 4x, to justify that jump in naming, but at least it was big.

    Anyway, the point is that these guys really can't call it 7 nm, with only a 15% density improvement. You have to keep in mind that one way chips (particularly things like GPUs, FPGAs, AI accelerators, etc.) get more performance by moving to a smaller node is by using more transistors. If they only get the stated efficiency improvements with such a modest bump in transistor count, that's only roughly half of the picture.

    BTW, I wish this were the process that the RX 590 used. Then, it might've actually delivered a worthwhile benefit.
    Reply