Originally Posted by QueueCumber
I have no idea what your argument is, or what you are twisting my post into. My only intent was to list additional qualities computer engineers seek to advance when designing new chip architectures, at least in computer engineering, where I have some training...
Ah sorry Queue I thought you were supporting Wills comments that every year a related product that is released is due to marketing and convincing the consumer they need it, when in reality the marketing and engineer side of a product are split.
Like you I work in manufacturing and rely on the chip manufacturers and other component manufacturers for the products we sell, the companies I have worked for are high end technology solutions-designs-products.
That aside going to carry on one last time to try to clarify for some here (not directed at you Queue) that their assumptions on marketing is plainly wrong and they really should stop saying that any CD player released over X or comes out next year is all down to marketing requirements and need, or that the jump to FPGA or different improved architecure is just a marketing exercise and nothing to do with engineering needs/demands.
For Will and a few others who keep falling back on the conclusion any new CD player-chipset-dac is pure marketing requirement and not engineering, I was trying to share with them where there is a differentiation between a product design phase and product positioning that is more marketing related in the way they suggest.
This is from my own experience as part of the engineer team designing and implementing a new architecture and product, being fortunate to work closely with real CTOs, also working with senior sales management, and also the side where we discuss and present to the big boy investors, such as Goldman Sachs top clients and analysts, amongst others.
In all of this I have never come across ANY situation where a product is designed from the ground up or has different architecture-functionality because marketing says we need a new product that does the same as before but looks newer.
Staying with the processor-chip side, as this seems an easier discussion to follow without going into technical aspects, and fits with the evolution that we are seeing with FPGA based DACs.
In its simplest form it can be broken down as follows:
1. Design process with the potential of a phased deployment of a new product, meeting engineering requirements of either this manufacturer due to improvements in architecture/functionality/performance, or other manufacturers who will be using this chip/chipset. A clear example using Intel/AMD is a new chip developed to 45nm while existing ones are 65nm. A clear engineering benefit is reduced heat for greater processing power, but this does not reduce costs for the manufacturer.
A consideration here is the cost of R&D-production lines-parts and storage management,etc that will mean this will cost more in the short term than existing products. This in a way is comparable to the FPGA designs we see now, or going back 6-8 years ago when Chord Electronics went for this new design. Also this may also include more noticable revisions to a product that tie in with engineering needs again, such as the DAC64 to consumers is seen consistently as the same product comparable price over the years but by the time it was discontinued its architecture-functionality had gone through four revisions.
2. Fabrication convergence between existing older lines and the newer product, for Intel and AMD this means moving the existing 65nm over to 45nm. Here the benefit to the manufacturer is cost reduction as the production line/process is now aligned, while also responding to other manufacturers engineering needs of reduced heat (and depending upon circumstances reduced watts). This is just one example of production convergence but is a good clear example.
3. Product positioning where minor revisions are done that assist in keeping the chipset wanted and in need, this can be price tiers with some functionality/performance disabled, or what you usually see with Intel/AMD increased clock speeds that come out every 3-6months.
This is to delay any new product design as long as possible (due to the cost), while getting back as much money as possible for the current product/s.
Now this does fit more into the marketing ideas of some here, where they feel a product is churned out again with little or minor differences/improvements. It does not cost a manufacturer too much to do this while benefits of potential sales can be sustained or improved.
Why am I going through this lengthy tale that is also my real work experience with the comparable above example?
Because I was trying to point out that we have seen a jump in terms of advancement of discrete signalling processing and supported architectures of DACs-chipsets in both CD Players and seperate DAC products.
The clear example is that FPGA used to be more niche in high end or was only a few audio manufacturers going this route, the cost for doing this in terms of R&D (think costs of top engineers in this field on top of time and other resource requirements) and then setting up production processing is so much more than staying with existing-traditional architecture and chipsets.
However some it seems like to suggest that this was purely done as an exercise in terms of sales marketing and not engineering requirements that Chord Electronics and others such as dCS felt was necessarily.
Right or wrong they went this path for engineering purposes, and now they are moving on again to a next stage of FPGA we are seeing this architecture and design appearing in cheaper alternatives (trickle down effect) using a lesser version of what they have done but still more advanced than traditional chipset or architecture-functionality-and performance.
Yes I do agree that products are churned out for the purpose of marketing (as point 3 explains), and yes there are plenty of similarly designed products using same functionality and architecture that do sound incredibly similar in DACs and CD Players, however there are also many that do not for good reasons due to engineering and architecture improvements/differences.
For those who are cynics about engineering design and where marketing that they feel drives it all fits in, try chatting to a real CTO sometime and consider those 3 simple points above when you look at a product deployment plan for a technology.
OK well some are going to be happy (again not directed or because of you Queue), because this is the very last post I am going to do in either CD or Amp section of AVSF, in fact I will not bother ever entering those sections ever again.
However if interested in more feel free to send me a PM and I will be happy to discuss any details.