After slashing the price by $200 you couldn’t possibly imagine that NVIDIA would actually decrease their overhead margins.
Most reviewers have primarily addressed the performance aspect of NVIDIA’s latest GeForce RTX 4080 SUPER graphics card. Or rather, they have highlighted the lackluster performance in comparison to the original graphics card.
Despite the revised uplift, consumers have only seen a marginal performance increase in the single-digit range, all while still facing a price premium compared to AMD’s Radeon RX 7900 XTX. The AMD option boasts an additional 8GB of VRAM and offers greater gaming performance and memory bandwidth.
Few tech tubers or review outlets have delved into the raw internals of the RTX 4080 SUPER. Many speculate that it is essentially identical to the original RTX 4080, as both share the same AD103 core and you’d be absolutely right on the money with such an asumption.
One small problem however, you see the RTX 4080 SUPER is supposed to provide the full AD103 core with “10240” CUDA cores as apposed to the “9728” CUDA cores present on the outgoing and unloved GeForce RTX 4080, an increase of just over 5% however the actual performance increase between the RTX 4080 and this new SUPER is far slimmer than that figure.
The only “saving grace” regarding the disappointing lemon would be the fact that NVIDIA allegedly slashed the MSRP price tag by two hundred bucks down to just $1000.
Except you’d be hard-pressed to actually find an RTX 4080 SUPER for sale on storefronts such as Newegg or Amazon for less than $1050, with the cheapest on Newegg being $1100 for a cheap and flimsy ZOTAC.
But hey, an fradulent MSRP of $999 does make the RTX 4080 SUPER seem a whole lot better than it actually is, a whole year later NVIDIA is finally catching up to the Radeon RX 7900 XTX in terms of pricing, except for the fact that you can regularly purchase one of those for $950 or even $930 for a limited time.
NVIDIA is also a step above the curve, the word “slimmer” doesn’t just describe the marginal price drop and performance increase but the word can also be used to describe the power delivery of the RTX 4080 SUPER as well.
According to Geekerwan, the new RTX 4080 SUPER certainly doesn’t feature a super exciting VRM layout compared to its predecessor, as NVIDIA has reduced the number of phases from 13 to 11 for the GPU and from 3 to 2 for memory.
Interestingly, there is also one missing phase near the power connector. Although not a major change, it’s an intriguing choice considering the RTX 4080 SUPER has more cores and faster memory. Additionally, it’s worth noting that the ASUS ROG STRIX sample, also tested by Geekerwan, features the same power layout as the original version.
Geekerwan suggests that this alteration to the PCB design enabled NVIDIA to reduce power consumption during gaming. Although the card maintains the same TDP value of 320W, NVIDIA claims that average gaming loads now consume only 246W, which is merely five watts less than the RTX 4080’s 251W.
These numbers are reported by NVIDIA on their official website.
I for one simply am not buying their excuses, this is just another showcase of NVIDIA’s brilliance when it comes to corporate greed, they’ve managed to slim off a marginal amount of the excess fat that is their horrid pricing however it certainly comes at a cost of them penny pinching when it comes to its power delivery.
While the slight decrease in the number of phases won’t impact performance during stock configuration, it’s a concerning tactic by NVIDIA. They are unwilling to provide consumers with the same componentry for a slightly reduced figure, effectively saving mere dollars per unit on a graphics card that costs over $1050.
But then again, this is the same company that attempted to sell you middle of the road product with only 53% of the maximum possible amount of CUDA cores for $1200 in the form of the RTX 4080 whereas just one generation prior they were giving you effectively more bang for your buck with the RTX 3070 which had an MSRP figure of just $500.
Excluding the fact that the overall die size of AD103 is effectively smaller than GA104 at 379mm2 versus 392mm2, despite the steep price hike associated with TSMC’s cutting edge N4 (5nm) processing node which sees a monumental increase in actual transistor density by over 100%.
There is absolutely no justification for a 140% markup on such a product. This entire graphics generation has been disappointing at best, with NVIDIA leading the charge in exploiting consumers with outrageously expensive hardware being marketed as products of a higher caliber, tier, or SKU than they should be, AMD has effectively followed suit in the pricing game, offering more competitive solutions for better perceptions of “value” across NVIDIA’s entire product stack, excluding the RTX 4090.
The blame for all this doesn’t solely fall on NVIDIA, but also on the millions of gamers who are willing to continually pay hundred more than the previous generation for roughly the same performance uptick, it’s simply mindshare gone mad and now all parties involved are reaping the rewards despite far fewer direct sales.
So I cannot blame NVIDIA for deciding to deciding to take a $10 pittance worth of VRM phases with the RTX 4080 SUPER, they clearly need the extra margin fluff and capital to facilitate upcoming 12VHPWR lawsuits as we close out the decade.
