After taking a break from the release of NVIDIA's GeForce RTX 20 Super Card and the new AMD Radeon RX 5000 series cards, the video card market is back up just in time for the all-important season for holiday shopping. Seeking to fill the bulk of the market, AMD is already delivering its OEM series of Radeon RX 5500 cards – with retail to follow soon – and meanwhile, NVIDIA is making its own machinations for the market.
Launches the first of the current hardware launches for this holiday tweak, NVIDIA, which announces its GeForce GTX 1660 Super. This is a relatively minor, but no less interesting, version of the GTX 1660 family that adds 1660 (vanilla) SKUs with faster GDDR6 memory for improved performance. Together with the GeForce GTX 1
The fourth Super card to be launched by NVIDIA, the GTX 1660 Super is a little more difficult to read than the high quality cards released last summer. In terms of performance, the card is just shy of the GTX 1660 Ti and priced just north of the GTX 1660 (vanilla). But unlike other Super Cards, this release is not the same kind of transparent pricing as the previous release; NVIDIA is not releasing any cards this time around, and commercial racks will be shared across the three GTX 1660 series cards in the coming months. However, this latest card plays at least the same basic role as a mid-range kicker – even if the original GTX 1660 only launched 7 months ago.
|NVIDIA GeForce GTX 1660 Series|
|GTX 1660 Ti|| Full TU116
w / 12Gbps GDDR6
|GTX 1660 Super (New)|| Cut-down TU116
w / 14Gbps GDDR6
|GTX 1660|| Cut-down TU116
w / 8Gbps GDDR5
The replacement factor here, as often is the case factor, is AMD. The company has already announced part of its RX 5500 family of products. However, it's unclear what to expect from the performance cards, especially since AMD created it for the NVIDIA lower-end GTX 1650. So, though at face value, the GTX 1660 Super looks like an NVIDIA counter on the upcoming AMD launch, beyond the AMD and NVIDIA testing labs, it remains to be seen where things are going.
However, what is clear is that this may also be the actual price reduction for the NVIDIA Ti GeForce GTX 1660. Set to launch at $ 50 less than NVIDIA's fastest GTX 1660 card, the GTX 1660 Super is only at least slower than the GTX 1660 Ti, delivering about 96% of its performance – almost beyond the regular our variations for launch and execution. When NVIDIA began shortening the GTX 1660 Ti to create the original GTX 1660, the only major change there was the replacement of the GDDR6 for the cheaper GDDR5. Now that they have effectively eliminated this – and then some – much of the performance that NVIDIA shaved on the GTX 1660 is back in the lower-priced Super Card. As a result, while the GTX 1660 Super offers nothing in terms of performance we've never seen before, it brings this level of Turing performance down to a lower price point and, hopefully, NVIDIA, something more enticing for GTX Owners of 10 and 9 series, which were delayed by the relatively high prices of the latest cards.
|Comparison of specifications of NVIDIA GeForce|
|GTX 1660 Ti||GTX 1660 Super||GTX 1660  GTX 1060 6GB|
|CUDA Cores||1536||1408||1408  1280|
|Boost Clock  1770MHz||1785MHz||1785MHz||1708MHz|
|Clock 14Gbps GDDR6||8Gbps GDDR5||8Gbps GDDR5 (X)|
|Bus width||bit||192 bit||192 bit  192 bit|
|Single Precision Perf.||5.5 TFLOPS||5 TFLOPS||5 TFLOPS||4.4 TFLOPs|
|TU116 )|| TU116
|Number of transistors||6.6B||] 6.6B||6.6B||4.4B  Architecture||Turing||Turing||Turing||Pascal|
|Manufacturing process||TSMC 12nm "FFN"||TSMC 12nm "FFN"||TSMC 12nm "FFN"||TSMC 16nm|
|Starting Price||$ 279  $ 229||$ 219|| MSRP: $ 249
FE: $ 299
And what exactly is GeForce GTX 1 660 Great? In short, get the GTX 1660 (vanilla), replace the 8Gbps GDDR5 with the 14Gbps GDDR6, dial a TDP with a few watts to meet the increased power requirements, and you have the GTX 1660 Super.
On the GPU front, nothing has changed since the GTX 1660. This is the same TU116 GPU in the same partial-shutdown configuration with the same clock speeds. True, I'm not even sure NVIDIA GPU bindings have changed for this product, or if they just had partners kill the TU116-300 GPUs on a GDDR6-related PCB. Either way, the net result is that paper performance has not changed since the GTX 1660, as the GPU itself can process just as many pixels, textiles and data elements as before.
Instead, what's changed is memory support, and that's where the vast majority of GTX 1660 Super's achievements lie. While the original GTX 1660, shipped with 8Gbps GDDR5, is both a cost-cutting and card-segregation tool of the full-fledged GTX 1660 Ti, the GTX 1660 Super gains the full width of the memory band it can handle. NVIDIA has given it 6GB of 14Gbps GDDR6, the fastest class of memory they currently have. This gives the GTX 1660 Super 336GB / sec the bandwidth, 75% more bandwidth than the original GTX 1660, or as much as the RTX 2060. And that wide bandwidth is almost entirely responsible for the 12% performance gains that we see above the original map.
In fact, the GTX 1660 Super actually has more memory bandwidth than the higher-end GTX 1660 Ti. This card comes with a slightly slower (and cheaper) 12Gbps GDDR6, which gives the GTX 1660 Super a bandwidth advantage, further eroding the little lead that the GTX 1660 Ti holds. After all, the TU116 is almost overwhelmed by the GDDR6, which is why the GTX 1660 Super does not outperform the GTX 1660 Ti despite its advantage; at this performance level, the additional memory bandwidth is not as important as the additional two SMs that receive the GTX 1660 Ti.
Finally, as a member of the GTX 1660, the latest card receives a very similar, but not quite identical 125W TGP. This small 5W increase over the GTX 1660 and 1660 Ti (as well as the GTX 1060/960 families) and seems to be mostly here to meet the larger power needs that come from shipping with fast GDDR6 memory – otherwise The GTX 1660 Super loses a bit of clock speed compared to the original GTX 1660. Plus it's a small enough change that PCBs should have no problem meeting increased power needs with their existing GTX 1660 Ti designs, meaning the GTX 1660 Super maps can easily be built with current parts, the only drawback In this case, the higher TDP and lower performance compared to the GTX 1660 Ti means that the GTX 1660 Super is not as energy efficient as the card, which would otherwise be redundant.
Blackout state of NVIDIA's product stack  If it looks like NVIDIA has released a lot of GeForce graphics cards in the last year, that's because it is. For the 13 months that return to the initial launch of the Turing architecture and the GeForce RTX 20 series of cards, NVIDIA has already released 12 different GeForce SKUs. And even throwing away the RTX 2080 and RTX 2070, which were officially discontinued but still sitting on some store shelves, NVIDIA's current product suite is still 10 cards deep.
|NVIDIA GeForce 20/16 Series (Turing) Product Stack  RTX 20 Series||GTX 16 Series|
|RTX 2080 Ti||GTX 1660 Ti|
|RTX 2080 Super||GTX 1660 Super|
|RTX 2080 (Discontinued)||GTX 1660||] RTX 2070 Super||GTX 1650 Super|
|RTX 2070 (Discontinued)||GTX 1650|
|RTX 2060 Super|
|] RTX 2060|
For comparison, the GeForce GTX 10 Series of product stacks were only 8 cards deep at its strongest point, and this was after the GTX 1070 Ti was released about 16 months later.
To be fair to NVIDIA, part of this is because the stack of GeForce products goes higher in price. than before, especially at $ 1100 RTX 2080 Ti. However, NVIDIA has released more cards at a faster rate than at any point in the past half decade. In particular, Super Cards have added significantly to this number, with NVIDIA shipping being what represents a mid-generation refresh for their line-up even a year after its launch. Compared to the GTX 9 series and the GTX 10 series, where NVIDIA had different two-year product cycles and launched very little filler between them, the RTX 20 and GTX 16 series essentially brought us back to the annual card release of more than two years. GPU cycles.
In the end, this means that there are a very large number of NVIDA video card options on the market, with NVIDIA offering 5 SKUs under $ 300. at any cost. As a sales strategy, market saturation is definitely a viable method of maximizing sales – making sure that there is always a price card that anyone is willing to spend – but that is something that goes wrong in NVIDIA's product suite, especially now with three different GTX 1660 cards.
Product Positioning and Competition
Switching gears, let's talk about the hardware behind today's launch. Because the new GTX 1660 Super cards are mid-range kicks on the main card, there are no reference cards here. Instead, it's a pure virtual start, with all the cards on the market being affiliate designs.
At this point, partners have been sending cards based on TU116 for more than 6 months so that they have been able to redirect their GTX 1660 family designs. This means that they are hitting the ground using their existing card designs for the GTX 1660 Super. So far I have not seen any new designs specifically for the Super card and I do not expect any. So these cards will look and appear similar to GTX 1660 cards already on the market.
Unsurprisingly, the GTX 1660 Super occupies the same common slot in NVIDIA product packages as the lighter GTX 1660 cards.
Which means that for system builders and upgrades, NVIDIA sends the card to the same general audience as before. That the GTX 1660 Super offers better price performance is a definite improvement and that will probably ensure that NVIDIA stays ahead of the upcoming AMD cards.
However, upgrades are likely to become a little more pretentious; Turing is not a whole generation of performance-based Pascal cards from the GTX 10 series, and the GTX 1660 Super does not change that. So while it's about 35-40% faster than the GTX 1060 6GB, the GTX 1660 Super will be much more appealing to GTX 960 owners and the like, where the new card delivers more than twice as much performance.
Meanwhile, as I mentioned at the beginning of this article, while NVIDIA is launching the new GTX 1660 Super card today, they will not be withdrawing any of their other cards afterwards. So, on paper, the GTX 1660 Super will be positioned very closely between the GTX 1660 Ti and the GTX 1660 (vanilla).
At current prices, the $ 229 GTX 1660 Super is $ 50 cheaper than the GTX 1660 Ti. And with only a card that is only about 4% behind NVIDIA's top 1660 cards, I would argue that the GTX 1660 Ti is redundant at this point; The GTX 1660 Ti is not forward enough to provide much in value. Instead, the next real step after the GTX 1660 Super is the GTX 2060, which is 22% faster but works out to about $ 100 more expensive.
Otherwise, coming from the other direction, the vanilla GTX 1660 is still regularly priced around $ 219. With a mere $ 10 price difference between the GTX 1660 Super and the Super Card, the Super Card is the obvious choice right now, as 12% more performance is more than $ 10. Nevertheless, NVIDIA insists that they and their partners will continue to sell the vanilla GTX 1660. So it will be interesting to see what NVIDIA actually does, both in terms of card supply and pricing. The GTX 1660 should be cheaper in order to have a future; and if that happens, then a suitable $ 200 spoiler could be obtained.
It should also be noted here that NVIDIA does not make packages or other promotional offers for any of the GTX 16 cards, so what
Finally, the competitive situation for the GTX 1660 Super is a mixed bag. AMD is still doing what it can to put the Radeon RX 590 as their closest competitor to the GTX 1660 Super. The notoriously awkward, late-cycle Polaris card was originally designed to beat the GTX 1060 6GB in performance (while throwing energy efficiency out the window), which certainly did. However, that means it has outpaced the GTX 1660 Super by about 20% (and with much less power), so AMD instead uses the card to stake its claim up to $ 199 price. So for buyers who absolutely cannot exceed $ 200, this is an option; but at the moment, the GTX 1660 Super is not facing much competition.