Nvidia will announce the following technology of GeForce graphics card in some unspecified time in the future. This we all know. What we will not say for certain is simply what that subsequent technology will appear like. We’re not utterly at nighttime, nonetheless. Latest leaks claiming perception into Nvidia’s mammoth machinations counsel we’re eyeing up a graphics card of herculean proportions.
On the phrase of a dependable leaker over on X, kopite7kimi. They declare the cardboard will provide 32 GB of GDDR7 reminiscence, throughout a 512-bit reminiscence bus. As soon as up-and-running, stated card will gobble up 600 watts—proper on the most output accessible from a single 12v2x6 connector and PCIe connection.
Furthermore, the reported RTX 5090 will include 21,760 CUDA cores—that is 33% extra cores than the RTX 4090.
The GB202-300-A1 would have to be completely huge to accommodate that many CUDA cores, too. If you happen to scaled up the AD102 GPU discovered contained in the RTX 4090 to the dimensions of Nvidia’s greatest datacentre chip, Hopper’s GH100, you’d in all probability match simply sufficient SMs inside it to make the RTX 5090. That makes for a die dimension round 814mm2, or conveniently simply over 33% bigger than the AD102 at 609mm2.
Clearly, this proposed graphics card would make mincemeat of 4K. It’d even lend credence to Nvidia’s claims of 8K gaming, too. Although my preliminary response is cut up between emotions of awe, surprise, pleasure, and some extent of worry.
Is the RTX 5090 going to be the most costly gaming graphics card but?
The preliminary leak has led many onlookers to anticipate such a factor, together with KitGuru’s Leo Waldock asking whether or not we might get any change from our $2,500 down cost.
Fortunately, Leo, I dare say that feels an excessive amount of, even for this monster card.
Let’s break it down into easy arithmetic. The RTX 4090 began out at $1,599 within the US, although it was comparatively costlier in most of Europe. We’ll take the US MSRP for ease, anyhow. With 33% extra cores, lending to a presumed 33% or extra fps acquire in video games, the RTX 5090 is already coming near $2,149. Maybe $2,099 on a superb day.
Would Nvidia go so simple as matching the worth to the efficiency like that? Nvidia has typically priced its graphics playing cards roughly instantly consistent with their efficiency uplift, even to the purpose of some frustration with the RTX 40-series. That stated, the RTX 4090 was a a lot better deal than most. Costly, sure, however a superb deal for what you bought.
The RTX 4090 trounced the RTX 3090 Ti. I imply, actually trounced it. And when the worth tag was a mere $1,599, that is truly a major worth discount versus the RTX 3090 Ti at $1,999.
This tells us two issues: firstly, that Nvidia does not thoughts pricing a graphics card at a whopping $2,000; secondly, that Nvidia is not essentially pricing graphics playing cards in tandem with efficiency.
So, possibly we’re truly a $1,899/$1,999 card within the RTX 5090. That’d be a fairly good situation, and something lower than that might be a superb one. What provides me severe hesitation, nonetheless, is simply how costly the RTX 4090 is true now.
You possibly can chalk up the RTX 4090’s inflated price ticket to its AI means, ban in China, or assumed finish of life standing, nevertheless it’s by no means actually been a card that is bought at its MSRP. You’d have been fortunate to pay simply $1,599 for one—they have been upwards of $1,650, if not $2,000, at frequent instances all through its two-year lifespan.
So, the cardboard that is doubtlessly two thirds of what this proposed RTX 5090 can handle is already promoting for $2,000, because the market calls for, and Nvidia’s earlier historical past suggests it is not an organization that likes being minimize out of its personal revenue.
If the demand is there for extreme-high-end graphics playing cards such because the RTX 4090, and to such an extent for board companions and retailers to fortunately cost upwards of $2,000 for one, that is going to provide any business-savvy product supervisor at Nvidia a vivid concept: ‘If they will cost $2,000 for the RTX 4090, why cannot we cost $2,200 for the RTX 5090?’
We noticed this form of considering again in the course of the cryptocurrency mining craze, which is what seemingly led to such extraordinarily excessive, however unsustainable, costs for playing cards such because the RTX 3090 Ti and later the unlaunched RTX 4080 12GB. I keep in mind saying again then I certain hope these inflated costs do not act as an excuse to maintain costs excessive as soon as the craze is over, and so they did, for a bit. However then Nvidia noticed sense and realised it needed to convey its costs inside purpose for us mere mortals.
However the level stands: the roof has been blown off graphics card costs previously half decade, and we’re barely on the mercy of the market to inform us what to spend on these playing cards. The super-duper overclocked Asus Strix or equal RTX 5090 is bound going to sail previous $2,000 whether or not that is the MSRP for the Founders Version or not. For now, we do not know what Nvidia wish to worth the RTX 5090 at, nor will we discover out for fairly a while.
The RTX 5090 is at the moment rumoured to be introduced at CES 2025 and deliberate costs can, and really probably will, change between every now and then.
