Gamers Nexus
More sources:
AnandTech
TechSpot
Tom's Hardware
PC Perspective
Guru3D
TechPowerUp: 2060 Super / 2070 Super
PC World
Tom's Hardware Germany (translated): 2060 Super / 2070 Super
Videos:
AMD’s biggest ally for the RTX launch was NVIDIA, as the company crippled itself with unimplemented features and generational price creep out the gate. With RTX Super, NVIDIA demonstrates that it has gotten the memo from the community and has driven down its pricing while increasing performance. Parts of the current RTX line will be phased-out, with the newer, better parts coming into play and pre-empting AMD’s Navi launch. The 2070 Super is priced at $500, $50 above the 5700 XT, and the on-paper specs should put it about equivalent with an RTX 2080 in performance; it’s even using the TU-104 RTX 2080 die, further reinforcing this likely position. The 2060 Super sees a better bin with more unlocked SMs on the GPU, improving compute capabilities and framebuffer capacity beyond the initial 2060. Both of these things spell an embarrassing scenario about to unfold for AMD’s Radeon VII card, but we’ll have to wait another week to see how it plays-out for the yet unreleased Navi RX cards. There may be hope yet for AMD’s new lineup, but the existing lineup will face existential challenges from the freshly priced and face-lifted RTX Super cards. Today, we’re reviewing the new RTX Super cards with a fully revamped GPU testing methodology.
Unless AMD has some last minute driver magic, then we can pretty much conclude once and for all that nothing changed since AMD's Navi presentation and that Nvidia knows exactly how AMD's cards perform and just priced theirs accordingly to fit snugly in both line-ups like Hardware Canucks says (fourth video in the OP):
View attachment 277498
Now, reviewers are all abiding by AMD's NDA and some are reserving an eventual Nvidia Super recommendation based on the relative comparison to AMD's Navi cards, so we'll be waiting for those reviews as well, but as I said, unless AMD has some trick up its sleeve, we're just playing the NDA waiting game to know what we already do from how the pricing of both line-ups stacks up.
And maybe Steve from Hardware Unboxed let it slip when he said that "we're desperately in need of AMD to step it up a notch... or two... maybe three.".
At 15:58:
https://youtu.be/dGpzf0uHuCc?t=958
Is this the part where I say "back in my day" the flagship was around $500?GeForce RTX 2070 sells for as little as $500
Having bought an RTX 2080 in November, I obviously am not thrilled by these Super Variants being out, but at the same time I'm not totally bummed as
1. RTX 2080 is still legitimately, well outside of margin of error, faster than any RTX 2070 Super. 2080 also still has more RT cores and Tensor cores as well.
2. With 2100mhz sustained on air, and Samsung GDDR6 that can clock over 8300mhz, at the very least I've got a decent chip.
Now if the 2070 Super matched the 2080, then yeah, I'd be pretty super heated pissed. Thankfully, that's not the case.
The one's I really feel sorry for are the one's that bought 2060's and 2070's, especially the 2060, with the Super getting 8Gb VRAM.
Either way, everybody who bought RTX got screwed big time. Something better is always around the corner. I'm not beating myself up too badly though, as the 2080 was basically the only choice I had when I built my brand new rig in November, coming from an 860M (Maxwell).
https://bit-tech.net/reviews/tech/graphics/msi-geforce-rtx-2070-armor-review/12/ said:It’s also considerably quieter, with the large fans never going past 1,200 RPM. The MSI card is certainly the bigger card, but it seems to put its size to good use in reducing noise output.
https://www.kitguru.net/components/graphic-cards/dominic-moass/msi-rtx-2070-armor-8g-review/15/?PageSpeed=noscript said:The clock speed, however, impressed me most – the card averaged 1962MHz with our overclock applied, which is almost 210MHz faster than stock. This pushed it ahead of the Gaming Z, and that was clearly reflected in our performance tests shown on the previous page.
Moving on into operating frequencies, temperatures, fan speeds and a little overclocking. Out of the box the RTX 2060 Super Founders Edition ran at 70C with a fan speed of just 1650 RPM, so it was virtually silent. The GPU maintained an average operating frequency of 1815 MHz after an hour-long loop of the F1 2018 benchmark.
Overclocked our sample was able to sustain 2010 MHz after the hour long test and this saw the operating temperature increase to 75C with a fan speed of 1900 RPM, so a pretty typical 11% overclock for a Turing based GPU.
Moving on to the 2070 Super and we see out of the box the Founders Edition model peaked at just 68 degrees with a fan speed of 1600 RPM, and this allowed it to maintain an operating frequency of 1890 MHz.
Overclocked our sample maintained an operating frequency of 1980 MHz. During the first few minutes of the test it was running at over 2 GHz but after about 5 minutes began throttling down and by the end of the hour sustained 1980 MHz which is a mere 5% overclock. The card was only running at 71C with a fan speed of 1700 RPM, so it was incredibly quiet.