FEBRUARY 2019 PCWorld 37Nvidia’s Turing GPU
architecture (go.
pcworld.com/ti80)
make the RTX 2060’s
cores much more
capable than the GTX
1060’s, as we’ll see in
the benchmarks
section.
Opposite, you
can see the
GeForce RTX 2060’s
specs, and how they compare against the
GTX 1060.
The only potential minor blemish on the
spec sheet: memory capacity. The move to
GDDR6 memory greatly improves overall
bandwidth for the RTX 2060 versus the GTX
1060, but the 6GB capacity might not be
enough to run textures and other memory-
intensive graphics options at maximum
settings in all games if you’re playing at 1440p
resolution. Most 1440p-capable graphics
cards include 8GB of VRAM. But Nvidia’s
Brandon Bell told me via email that the
company chose the lesser capacity in order to
utilize cutting-edge GDDR6 memory, rather
than the much slower GDDR5 memory that
came on the GTX 1060:
“We could’ve certainly outfitted the RTX
2060 with 8GB of GDDR6, but hitting $349
wouldn’t have been feasible. And given that
today’s games don’t take advantage of the
additional memory, it made sense to price the
card more aggressively at the cost of a little
frame buffer size. Right now the faster
memory bandwidth is more important than
the larger memory size.
“Maybe in a year or two games will come
along that need the full 8GB of memory, but
we’re just not seeing that right now, and by
the time it does happen we’ll probably be on
another generation of GPUs beyond Turing
and GDDR6 memory prices will be lower than
they are today.”
Moving on, the RTX 2060 Founders
Edition strays from the path beaten by
Nvidia’s other RTX FE options. While the
Founders Edition versions of the RTX 2070,
2080, and 2080 Ti all come with a mild
overclock and premium pricing, the RTX
2060 Founders Edition sticks to the stock
specifications outlined above, and the card’s
$350 MSRP.
That’ll make it hard for entry-level models
by Nvidia’s board partners (like EVGA and