Tuesday, July 4th 2023

ICYMI, GeForce RTX 4060 Launched Last Week: A Summary of our Mammoth Coverage

Last week (June 29), NVIDIA launched the GeForce RTX 4060 "Ada" graphics card targeting a majority of PC gamers who still play at 1080p Full HD resolution. Reception to the new card was somewhat lukewarm, only 6% of our polled readers are interested in buying one (poll results). We crunched thousands of benchmark runs in our new 2023H2 test suite and managed to review ten different models of the 4060, six of which are available at the NVIDIA MSRP of $299, and we followed up with four more reviews of premium custom-design cards priced above the MSRP. Our results confirm that GeForce RTX 4060 is designed for maxed out gaming at 1080p; gameplay with ray tracing is possible at 25-50 FPS at 1080p, which can be increased by dialing down graphics settings, or through use of DLSS and DLSS 3 Frame Generation, in supported games. On the other hand, there were concerns with the small performance gains and high pricing, which makes many other options a viable choice.

We've put together a video presentation which compares all the cards tested, and that also serves as a really quick executive summary of the RTX 4060, and whether you should get one.

Add your own comment

5 Comments on ICYMI, GeForce RTX 4060 Launched Last Week: A Summary of our Mammoth Coverage

#1
wolf
Performance Enthusiast
I really have no issues with the GPU itself bar two things, weighted quite differently.
  • Name - it should have been the 4050, especially since the 3060 had 12GB of VRAM. It would genuinely represent a very nice uplift over the similarly configured 3050 too, although this doesn't matter nearly as much as;
  • Price - $299 is insulting at this performance level, while I think it should have cost $199, I can begrudgingly admit even $249 would have been a lot more reasonable. So even MSRP cards barely make sense, let alone the ludicrous notion of the behemoth variants costing more than MSRP.
Both major camps trying their dandiest to normalise the pricing we saw during COVID shortages, anyone else think next gen we might see minor price drops at certain tiers, yet remaining still above RTX 30/RDAN2 prices, and they'll pat themselves on the back for doing us a favour?
Posted on Reply
#2
mama
I like the roundup idea. Makes a quick review of the card type.
Posted on Reply
#3
Unregistered
wolfI really have no issues with the GPU itself bar two things, weighted quite differently.
  • Name - it should have been the 4050, especially since the 3060 had 12GB of VRAM. It would genuinely represent a very nice uplift over the similarly configured 3050 too, although this doesn't matter nearly as much as;
  • Price - $299 is insulting at this performance level, while I think it should have cost $199, I can begrudgingly admit even $249 would have been a lot more reasonable. So even MSRP cards barely make sense, let alone the ludicrous notion of the behemoth variants costing more than MSRP.
Both major camps trying their dandiest to normalise the pricing we saw during COVID shortages, anyone else think next gen we might see minor price drops at certain tiers, yet remaining still above RTX 30/RDAN2 prices, and they'll pat themselves on the back for doing us a favour?
The naming scheme is wrong from the top to bottom, the 4090 is more like a 4080 and so on.

Though I don't mind the naming as long as the price is correct, but none of the 4000 is priced properly.

Another problem is that people keep buying nVidia even though AMD (no defending AMD their RDNA3 pricing is as awful as nVidia's) has better offering now with RDNA2, a recent example people were defending DLSS, but I don't blame them as fake reviewers like infamous just buy it just or DF which are basically another marketing branche for nVidia.
#4
Crackong
Xex360The naming scheme is wrong from the top to bottom, the 4090 is more like a 4080 and so on.

Though I don't mind the naming as long as the price is correct, but none of the 4000 is priced properly.

Another problem is that people keep buying nVidia even though AMD (no defending AMD their RDNA3 pricing is as awful as nVidia's) has better offering now with RDNA2, a recent example people were defending DLSS, but I don't blame them as fake reviewers like infamous just buy it just or DF which are basically another marketing branche for nVidia.
I think the naming of 4090 is fine
But Nvidia got too greedia for 4090->4080, this 4080 should be a 4070ti
The real 4080 should be something around 12k cuda cores

Then
4070ti->4070
4070->4060
4060ti ->4050ti
4060->4050
Posted on Reply
#5
wolf
Performance Enthusiast
I think the lift from Ampere to Ada, massively aided by the node disparity, allowed Nvidia so much of an up tick they certainly got a bit excited punching cards up a teir for most of them.

The exception to some extent imo is the 4080, iirc it's the first time we see an xx103 GPU in a desktop card, and it performs so well they've slotted the 4080 there and left ad102 for the true ultra enthusiast, "I'm rich" level.

I don't think we're 'owed' certain gpu's to become certain products, but a few cards in the stack would definitely make more sense to folk if they kept to how it was done with Ampere and Turing. Like the 4060 is underwhelming because they went down one die designation, but compared to a 3050 instead it's damn impressive imo, just garbage prices.
Posted on Reply
May 29th, 2024 07:42 EDT change timezone

New Forum Posts

Popular Reviews

Controversial News Posts