This article is more than 1 year old
Nvidia's RTX 4060 and 4060TI are actually priced like mid-tier cards
Maybe Jensen finally got the memo about the whole GPU shortage being over
Nvidia appears to have come to its senses with the launch of the RTX 4060 and 4060TI on Thursday.
While the company's prices for its flagship 4090 and 4080 cards, which retailed for a whopping $1,599 and $1,199, respectively, when they launched in October, the chipmaker won't charge a premium for its mid-tier graphics cards.
The 4060 and slightly higher specced 4060TI will retail for $299 and $399 respectively. If you recall, the RTX 3060 actually had an MSRP of $329 when it launched, not that it ever sold for that on account of crypto-fueled price jacking. As a result, Nvidia's entry-level 40-series part is actually cheaper, in theory. We'll have to wait and see what board partners actually charge when the cards hit store shelves.
The cards themselves still target 1080p gaming — that much hasn't changed — but unlike older cards they should be able to push much higher frame rates, especially if Nvidia's deep learning super sampling is turned on. The tech, which is now in its third-generation, uses deep learning to upscale lower resolutions to achieve higher frame rates.
One of the reasons behind its apparent stagnation may be down to vRAM. Both cards feature 8GB of GDDR6 memory. While this used to be plenty for AAA games, it may not be enough to satisfy the latest titles at the highest settings. And although this is a mid-tier card, vRAM requirements have crept up steadily over the past few years. Many new and upcoming games — Jedi Survivor, for example — are now calling for a minimum of 8GB of vRAM.
For those worried about vRAM bottlenecks, Nvidia says it will offer a 16GB version of the 4060TI for $499 alongside the 4060 in July.
Even though vRAM has remained unchanged and actually gone backwards in the case of the lower end 4060, the two cards have received a rather substantial boost to their cache architecture. The 4060 now features 24MB of L2 cache, up from just three last generation, and the TI-variant is now packing 36MB.
- To quench AI's thirst, the way we build, operate datacenters needs to change
- With AI hype driving up server GPU prices, will cloud costs rise next? We reckon so
- Nvidia CEO pay falls ten percent in FY23 on missed sales targets
- A lone Nvidia GPU speeds past the physics-straining might of a quantum computer – in these apps at least
Don't expect the larger cache to contribute much in terms of generation-on-generation gains, however. Nvidia didn't share much about the performance of these cards, but if its claims are to believed, you can expect the 4060 TI to perform about 1.7x faster than its predecessor, when using its frame generated DLSS 3.0. When Nvidia's AI upscaler is turned off, the card is about 1.6x faster than the two generations old 2060 Super. If you've already got a 3060 or 3060TI, these cards probably aren't for you.
The 4060 TI will be available starting May 24, and will be getting the "Founders Edition" treatment. Meanwhile the 16GB version of the TI and entry-level 4060 will launch sometime in July. ®