F E A T U R E // T U R I N G - C E L E B R A T I N G T E C H N O L O G I C A L A D V A N C E M E N T
particular actions and behaviours on
the participants. All of which which
result in the final product pricing.
The factors that contribute to the
price of any silicon are of course
related to the R&D effort and
financial resources which brought
about that that silicon. However, the
consequence of a market economy,
and more specifically the ever
changing market conditions has a
sizable impact on prices. If you
examine INTEL, NVIDIA, AMD or any
such firm. The primary driver for the
business entity is profit. That is
literally the point of being in
business. There must be a business
case for products and other related
endeavours. In the actions and
undertakings of these firms, the
business case is sometimes easy to
see, but other times it isn’t,
especially from the outside, where
we stand. That evaluation is
deserving of it's own nuanced , but
numbers based discussion.
The price at which we end up buying
these RTX GPUs is driven by any
number of factors and not
necessarily technological ones.
That’s to say, even though these two
parts are intertwined, one has to
recognise that there are at least two
or more parts to this. What we can
examine and dissect is the end
product and how it works and
performs. A price discussion
requires
(if it’s to be answered rationally)
information and data that is unlikely
to be public or even available to
anyone within the firms whose
contribution and responsibilities
aren't concerned with that.
Consider for a moment that in
1994, gaining access to a vehicle
with Satellite navigation cost (in
today's money) $97,000 USD (BMW 7
Series). Today, a far superior GPS
system can be had for about $18,000
in the Fiat 500.
Some car somewhere had to be first.
There will be a time where there
are GPUs many times faster or
more capable than the latest Turing
GPU at ray tracing. For us to even
have a baseline, a RTRT capable
GPU has to exist first, regardless of
the price. As such, Evaluation of
these GPUs in exactly the same
manner as we did past processors
makes it easy to miss the bigger
bigger, hence stifling the
conversation and the exchange of
knowledge. From the consumer
side, far removed from the technical
details of it all, it’s perfectly rational
to wonder if the market can deal
with such pricing on graphics cards.
In appreciation, criticism and
analysis of all sorts, nuance is
important and it is incumbent on
those capable to unpack what it is
that these processors mean and
what they represent. Not only in the
software of today, but how they
(along with the API extensions)
literally shape tomorrow.
“THERE WILL BE A TIME WHERE THERE ARE
GPUS MANY TIMES Faster or more
capable than the latest turing gpus.”
20 The OverClocker Issue 45 | 2018