1 / 13

PIDS summary

PIDS summary. T.Skotnicki,T.Schulz, K.De Meyer,J.Holz,H.Oda,H.Inoue. Summary on « Logic ». Present arguments in favor of « system-level » indicators Variability & power Density We have models to calculate both under Mastar and have shown in 2007 that:

jaunie
Télécharger la présentation

PIDS summary

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. PIDS summary T.Skotnicki,T.Schulz, K.De Meyer,J.Holz,H.Oda,H.Inoue

  2. Summary on « Logic » • Present arguments in favor of « system-level » indicators • Variability & power • Density • We have models to calculate both under Mastar and have shown in 2007 that: • The 8% results from the criterion of constant dynamic pawer per cm² • The relaxed Lg scaling results from variability limitations (SNM/sigma SNM) • Is 1/Ion a viable indicator ? • Ion may stagnate very soon ? • Ion/W – circumference or layout W ? • PIDS agrees on 8% improvement per year in Frequency but NOT in Ion • PIDS suggests to depart from 1/Ion and move towards variability and power and density as scaling indicators • PIDS agrees on the possible convergence between LOP and LSTP • Change of paradigm is scheduled for the 2009 edition, in 2008 – minor changes and preparation

  3. Summary on « Logic » • Methodology – MASTAR • The most recent version that served for the 2007 PIDS tables calculations is available from the ITRS web site • The « working » version with System-level calculations is available

  4. IMPROVED CV/I – - Full integration as well as Ieff already available in MASTAR

  5. Summary on « Memory » • NAND Flash – no significant change in 2008 is proposed for the 2008 edition (but still awaiting important comments from two « big » players, depend on their answer ITRS 2008 may be modified) • DRAM – no change in 2008 • NOR Flash – no change in 2008 • New Memory Devices (MRAM, PCM, FeRAM) – no change in 2008 • Discussions and surveys in preparation for 2009 • Methodology – dual: surveys for PIDS tables and formula-based for IRC summary tables

  6. PIDS - Modeling • Modeling TWG supports PIDS concern about viability of the 1/Ion as scaling indicator • CV/I is jointly considered a better indicator but still not sufficient – we should consider moving to inverter delay, variability (SRAM SNM/sigmaSNM) and power limitations (static and dynamic) • Exchange wrt models of random dopant positioning • Shall we extend PIDS tables to operation temperatures? No!

  7. PIDS - FEP • Rsd and in particular Contact (plug) resistance (Nickel contact ?) needs to be reconsidered (demand from FEP) Japan PIDS – FEP will work out • MASTAR Models: DG/FinFET – OK already present in MASTAR 5 and used for 2007 calculations • Agreement on CV/I definition • High mobility materials to be introduced in 2009 roadmap • ERD to be asked for help • ERD – clarify the question of Capacitor-less DRAM – FEP is considering it may be a potential solution to replace SRAM (still under discussion) • Japan PIDS-FEP have to discuss FeRAM tables of 2007 roadmap

  8. PIDS - ERD • ERD wishes to transfer to PIDS: • ETB (tunnel barrier) • Alternative channel materials • Floating body memory • MRAM and NW PCRAM Cross • – did not agree to transfer to PIDS, we agree to transfer to ERD transition table • ODA-san and INOUE-san will discuss memory entries and SKOTNICKI channel materials • ERD and PIDS will constitute a working group to analyse these materials and agree on models, mobility, saturation velocity, balisticity, BTBT etc • Wilfried will send us recent publications • NW –nanowires and CNT, NO they should not be transferred now, and not until Variability is taken into account (include in PIDS and ERD as criteria) and soltions are put forward to overcome the issue • MRAMs are in production (Freescale and Simens-preparation) , scaling down problems may find solutions – OK should be transfered. • Floating body – ask Rich LIU to drive discussion on its transfer and possibly generate PIDS roadmap

  9. PIDS - DESIGN • The 8% proposed by PIDS last year is accepted and proposed as the guideline till 2020 • Skotnicki and Khang will exchange models on Pdyn and delay and generate first results before the July meeting • A.Khang will provide PIDS with numbers on how many sigma (SNM/sigma SNM) can be suppressed per node by design means

  10. PIDS - ERM • Mike -Predictions on possible introduction of new materials are « optimistic » but this is what they should be to encourage labs to go on • ERM has good criteria for material evaluation • ACTION: benchmark variability perspective data from ERM with a CMOS line and recalculate SRAM SNM to compare SKOTNICKI-HERR • Send email specifying variability parameters we need for benchmarking the SRAMs • Consider providing the electrical parameter requirement (High-K material, tunnel barrier trap density and metal gate spec requirement for flash memory)

  11. PIDS - Wireless • Include I/O devices in table for analog functions • How will a I/O device look like for planar SOI and FinFETs? • How will I/O voltage evolve over time? • PIDS will provide an example I/O device in FinFET style (table 1 RF, precision analog device)

  12. PIDS - Yield • no points for further discussion

More Related