NVIDIA has pretty much won CES 2024 with their announcement of some SUPER average graphics cards that completely fuck over existing owners.

AMD’s presentation gave us the wonderful Radeon RX 7600 XT that’s built upon the same Navi 33 core as the standard RX 7600 with barely any actual performance improvement, with any actual performance gains stemming from its enlarged 16GB memory buffer when under high resolution workloads or with ray tracing enabled.

It was highly speculated that NVIDIA would announce a trio of “SUPER” revised models of their RTX 4000 series lineup, with the RTX 4080 and 4070 Ti both seemingly having their productions ceased as they’re both due for replacement, while an additional RTX 4070 SUPER model would be released alongside the standard RTX 4070 graphics card.

Such plans have now come into fruition as NVIDIA officially unveiled the RTX 4080 Super, RTX 4070 Ti SUPER and the RTX 4070 SUPER in what would be described as products that provide enhanced performance for the same price points as the outgoing models.

Except the RTX 4080 SUPER, that’ll be priced to compete against the Radeon RX 7900 XTX at $999, most likely due to the fact that hardly anybody was stupid enough to pay $1200 for one.

Starting off with the most improved model, that being the RTX 4070 SUPER, this was a card that was very much needed from NVIDIA to combat AMD’s immensely popular Radeon RX 7800 XT graphics card which provided greater performance than NVIDIA’s GeForce RTX 4070 for less.

As previously specified, the RTX 4070 SUPER will be having an increased core count, now featuring 7168 CUDA cores versus the 5888 CUDA cores present on the non-SUPER model, a ~20% increase in core count is a surefire way for the company to combat what would otherwise be the best GPU of the generation, the RX 7800 XT.

Contrary to previous rumors the RTX 4070 SUPER unfortunately will not be receiving a substantial increase in its L2 cache allotment, but nevertheless despite having the same memory configuration, that being 12GB of GDDR6X memory strapped to a 192-bit memory interface, the severe increase in core count alone should see the RTX 4070 SUPER performing above AMD’s offering in an out of the box configuration.

You could overclock your Radeon RX 7800 XT and probably match or surpass the RTX 4070 SUPER but such an argument is quite a stretch as most consumers simply do not overclock their graphics cards anymore.

Consumers will be getting the RTX 4070 SUPER for the same price as what the GeForce RTX 4070 previously was, that being $599 which still makes AMD’s Radeon offering compelling as they’re currently going for $510 on Newegg but I can see why majority of consumers would rush in to buy the RTX 4070 SUPER over AMD’s offering given that it’s from a supposedly “reputable” and “premium” brand as NVIDIA and stock performance figures would undoubtedly favor team green.

The RTX 4070 SUPER will be available to purchase on January 17th, if you were in the market for a mid-range product around the $600 range this might be a very compelling choice.

However those who’ve previously bought the RTX 4070 for $599 you’ve just been thrown under the bus, a normality when it comes to NVIDIA, however given that the RTX 4070 is the only product mentioned today that isn’t being discontinued, retailers will now be dropping their prices for existing RTX 4070 models, most likely down to $549 because anything less than a $50 price drop would impede upon their profit margins too much.

If you’re looking for performance, the RTX 4070 Ti SUPER, despite its stupid name is possibly to be the most intriguing card among the RTX 4000 series SUPER GPUs announced.

NVIDIA is boosting the memory to 16GB of GDDR6X memory, the same configuration featured on the outgoing RTX 4080, which makes sense given how the RTX 4070 Ti SUPER is leveraging the AD103 graphics core.

The GDDR6X memory is a little slower, at 21Gbps but a 256-bit memory bus does put it inline with the 4080, especially when compared to the standard AD104 based RTX 4070 Ti which came with just 12GB of GDDR6X strapped to a 192-bit memory bus.

If there was a card I’m thankful is now gone it would certainly be the RTX 4070 Ti, those who’ve been suckered into buying one of those for $800 will most certainly be kicking themselves as this brand new SUPER graphics card comes with 8448 CUDA cores versus 7680.

A ten percent increase in core count sure, but the additional memory bandwidth would put this model very close for comfort in terms of gaming performance versus the now discontinued RTX 4080 graphics card, despite essentially replacing the outgoing model’s price tag at $799 this is quite a serious offering made to compete against AMD’s somewhat neglected Radeon RX 7900 XT which starts from $780 on Newegg.

In terms of gaming performance, the RTX 4070 Ti is hard to gouge, for a more clearer picture we’re going to have to wait until reviews go live, but spitballing TechPowerUp figures, I can probably see the RTX 4070 Ti SUPER perform on par or fractionally lower than the RX 7900 XT for basically the same price, this particular card will be launching on January 24th.

If you were in the market for solid performance from your preferred brand of green, with a requirement of a tolerable 16GB of VRAM, the 4070 Ti SUPER seems like a solid option.

With the introduction of the RTX 4070 SUPER and RTX 4070 Ti SUPER, it certainly feels as if AMD are on the backfoot now and would need to respond to NVIDIA’s new offerings with compelling price cuts, mainly for the Radeon RX 7900 XT and now 7900 XTX.

Why would AMD need to drop the price of their flagship Radeon RX 7900 XTX? It’s quite simple, the dung of NVIDIA’s latest announcements, that being the RTX 4080 SUPER certainly doesn’t boast any redeemable qualities in terms of enhanced specifications, with a full fat AD103 core at its disposal it only features 10240 CUDA cores.

We’ve gone over this before, a substantially miniscule core increase of around 5%, with the same 16GB configuration of GDDR6X memory tied to a 256-bit memory interface, albeit this time around the memory modules themselves are marginally faster.

23Gbps vs 22.4Gbps GDDR6X memory.

NVIDIA is marketing this as the product with the world’s fastest memory which is essentially true in terms of frequency but hardly bandwidth, as the RTX 4080 SUPER should in theory provide roughly 2.7% more effective bandwidth over the outgoing model.

SUPER lousy if you asked me, the only real main point of this model besides “you should just buy the 4070 Ti Super” is that this new revision has had its price tag slashed by $200 now being priced as a marginally less inflated $1000 price point.

Launching on January 31st, the best case scenario for the RTX 4080 SUPER is that it’ll provide ~5% increased gaming performance over the now discontinued RTX 4080 but on average it’ll likely be less than that.

Don’t expect the RTX 4080 SUPER to surpass AMD’s Radeon RX 7900 XTX, however performance isn’t the actual concern more rather it’s now priced at an equal $999 or figuratively it is anyway and if consumers were given the option to buy either AMD Radeon or NVIDIA GeForce for an identical price, nine times out of ten they would likely purchase a GeForce graphics card.

All things considered NVIDIA’s RTX 4000 SUPER lineup is a price cut that isn’t actually a price cut, consumers who’ve purchased outgoing models have been royally fucked over big time as retailers will be quick to dump their remaining stock of RTX 4070 Ti and RTX 4080 GPUs for slightly lower prices.

Even with the price drop of the RTX 4080, with both 70-class GPUs retaining their prices, this entire generation of products from both NVIDIA and AMD remain to be woeful at the best of times, now is a considerably bad time to be in the market for a brand new graphics card from a value perspective.

It’s always nice to gawk at shiny new toys that are slightly faster than before, but they’re still demanding absurdly high prices that no sensible person aught to be paying. In my eyes these are just graphics cards that are slightly less overpriced than they were before.

If AMD wanted to be a little more serious than their dogshit attempt at a “7600 XT” in actually attempting to capture some fanfare, sales and market share they have no other option than to considerably drop their prices across their mid-range to high-end offerings, but regardless NVIDIA are now offering better performance than before for the same outrageous figure, where the only substantial “increase” of value comes in the form of being slightly less scammed in buying the now $1000 RTX 4080 Super as opposed to what was effectively a $1200 lemon that hardly anybody wanted.

The ball is firmly in NVIDIA’s court at the moment, if AMD wants to salvage something they need to drop their prices.