( ESNUG 515 Item 4 ) -------------------------------------------- [11/29/12]

From: [ Jim Hogan of Vista Ventures LLC ]
Subject: Custom 2.0 means that SPICE runs must BOTH be fast AND accurate

Hi, John,

With Custom 1.0, SPICE simulation speed and capacity came at the expense
of precision.

For the Custom 2.0 retooling, such a tradeoff will no longer be acceptable
to analog designers.  In fig 7, we look at simulation and analysis across
a scale showing the increasing effort and precision required. 

       
          Figure 7: Custom 2.0 verification technology landscape

The applications range from custom digital, memory, clock and I/O, and mixed
signal through analog to radio frequencies (RF), microwave, and micro-
electromechanical systems (MEMS).  The analysis gets more complex at the
circuit level as well as the whole system level, ranging from one-physics to
multi-physics.  Transient analysis is required across the broadest range,
from custom digital to microwave. From high-speed I/O's through RF, we find
various flicker, shot and thermal noise analysis.  From analog through
microwave, we have harmonic analysis. 

As chip designs scale to RF, microwave and MEMS, and as the need for 
chip-package-board co-design increases to provide for speed and density 
while reducing power and form factor, newer challenges surface.  With these
trends, it is often necessary to calculate the complex electromagnetic
fields to be able to model the designs accurately.  

Electromagnetic (EM) Integrity includes signal integrity, power integrity,
simultaneous switching noise analysis, and electromagnetic interference 
analysis.  A variety of approaches exist to calculate the electromagnetic 
fields.  These include 2.5D traverse electromagnetic solvers, 2.5D boundary
element/method of moment solvers, 3D quasi-static solvers and 3D full-wave
solvers.  

The Custom 2.0 SPICE simulation challenge is basically, with high volume,
28-nm-and-below designs, the cost of failure is so high that SPICE coverage
becomes increasingly critical.

Here are some problem examples from traditional SPICE simulation: 

    - Verifying a post-layout LC-tank voltage controlled oscillator 
      (VCO) with 1.8 million elements, including transistors, resistors,
      capacitors, and inductor took 18 days.

    - Verifying post-layout SerDes I/O Macro with 12.8 million elements
      could not be run. 

    - Verifying a post-layout 12-bit ADC with 7.4 million runtime took
      36 days.

Even accounting for increasing the number of sim farms, SPICE simulation
speeds need to improve 5-10x and SPICE simulation capacity needs to improve
100x when compared with traditional SPICE simulators.  Furthermore, both
improvements must occur without giving up any accuracy.

    - Jim Hogan
      Vista Ventures, LLC                        Los Gatos, CA

         ----    ----    ----    ----    ----    ----   ----

Related Articles

  Hogan on the early days of Custom 1.0 and Cadence Analog Artist
  Key aspects, market drivers for the present Custom 2.0 retooling
  Atomic scaling problems, Variation, and the Custom 2.0 retooling
  3D FinFETs mean lots and lots of SPICE runs even with Custom 2.0
  Custom 2.0 is design data dependencies, NOT design data managment
  Hogan outlines the current players for the Custom 2.0 retooling

Join    Index    Next->Item






   
 Sign up for the DeepChip newsletter.
Email
 Read what EDA tool users really think.


Feedback About Wiretaps ESNUGs SIGN UP! Downloads Trip Reports Advertise

"Relax. This is a discussion. Anything said here is just one engineer's opinion. Email in your dissenting letter and it'll be published, too."
This Web Site Is Modified Every 2-3 Days
Copyright 1991-2024 John Cooley.  All Rights Reserved.
| Contact John Cooley | Webmaster | Legal | Feedback Form |

   !!!     "It's not a BUG,
  /o o\  /  it's a FEATURE!"
 (  >  )
  \ - / 
  _] [_     (jcooley 1991)