http://chipdesignmag.com/sld/blog/2010/07/22/good%e2%80%99-vs-%e2%80%98good-enough%e2%80%99/ Deep Insights for Chip Architects and Engineers Wed, 15 Jun 2011 14:29:30 +0000 hourly 1 http://wordpress.org/?v= http://chipdesignmag.com/sld/blog/2010/07/22/good%e2%80%99-vs-%e2%80%98good-enough%e2%80%99/comment-page-1/#comment-8707 Gary Stringham Fri, 17 Sep 2010 17:32:34 +0000 http://chipdesignmag.com/sld/?p=3330#comment-8707 Agile software programming teaches to only program what is needed, not what MIGHT be needed. That works fine for software because updates and patches can be distributed. That behavior too time and cost prohibitive for hardware. So I worked with hardware engineers and persuaded them to include features that I MIGHT need. Months later when integrating my software with real hardware, I ended up using some of those features, which avoided delays and respins. Because the hardware had the extra stuff in it, it was “good enough” to ship. Agile software programming teaches to only program what is needed, not what MIGHT be needed. That works fine for software because updates and patches can be distributed. That behavior too time and cost prohibitive for hardware. So I worked with hardware engineers and persuaded them to include features that I MIGHT need. Months later when integrating my software with real hardware, I ended up using some of those features, which avoided delays and respins. Because the hardware had the extra stuff in it, it was “good enough” to ship.

]]> http://chipdesignmag.com/sld/blog/2010/07/22/good%e2%80%99-vs-%e2%80%98good-enough%e2%80%99/comment-page-1/#comment-8566 Andrew Haines Wed, 01 Sep 2010 13:30:15 +0000 http://chipdesignmag.com/sld/?p=3330#comment-8566 Increasing software intensity is driving good-enough in another way. Product functionality is now a combination of hardware and software not hardware alone. This drives testing up to the system level. If system level testing succeeds then the hardware is good enough even if it has latent errors that might have been a issue earlier. If errors are never exposed, they are not errors. It's good enough. Increasing software intensity is driving good-enough in another way. Product functionality is now a combination of hardware and software not hardware alone. This drives testing up to the system level. If system level testing succeeds then the hardware is good enough even if it has latent errors that might have been a issue earlier. If errors are never exposed, they are not errors. It’s good enough.

]]> http://chipdesignmag.com/sld/blog/2010/07/22/good%e2%80%99-vs-%e2%80%98good-enough%e2%80%99/comment-page-1/#comment-8249 System-Level Design » Blog Archive » Quantum Computers Move A QuBit Closer To Reality Thu, 22 Jul 2010 16:18:13 +0000 http://chipdesignmag.com/sld/?p=3330#comment-8249 [...] In the early days of computers, hardware was king. Software was written well after the hardware platform was built. Today, system designers have to balance the simultaneous co-design and co-verification of both hardware and software systems. In the near future, as hardware becomes more and more of a commodity, the software may eclipse hardware in the product life cycle to become the primary system driver (see ‘Good’ Vs. ‘Good Enough’). [...] [...] In the early days of computers, hardware was king. Software was written well after the hardware platform was built. Today, system designers have to balance the simultaneous co-design and co-verification of both hardware and software systems. In the near future, as hardware becomes more and more of a commodity, the software may eclipse hardware in the product life cycle to become the primary system driver (see ‘Good’ Vs. ‘Good Enough’). [...]

]]>