Nvidia GeForce RTX 4070 Ti evaluation: Hobbled and extremely expensive

This evaluation will be brief and sweet. Nvidia’s brand-new $799 GeForce RTX 4070 Ti is a technically hobbled, dramatically overpriced graphics card that costs $100 more than the higher-tier RTX 3080 did last generation. In exchange for that 14 percent greater cost, you normally get 14 percent greater efficiency at 4K resolution, and around 20 percent at 1440 p.

Depending on how you take a look at it, the GeForce RTX 4070 Ti uses stagnant-at-best worth at comparable rates to last generation, or a painfully high mark-up over last-gen’s currently pricey RTX 3070 Ti. It’s a beastly card that can fall the previous RTX 3090 Ti flagship in numerous video games, particularly at 1440 p resolution or when DLSS 3 swings into action, however at this chest-clutching rate, we just can’t advise it. We advise preventing it outright unless there are massive cost decreases.

GeForce RTX 4070 Ti: Specs, functions, and style

Brad Chacos/IDG

The specifications for the GeForce RTX 4070 Ti should not shock anybody, given that it’s the ” unlaunched” $899 RTX 4080 12 GB with a brand-new name. Players ( and us) called shenanigans on Nvidia at first providing this scaled-down GPU the exact same 4080 branding as the far more effective 16 GB RTX 4080, and soon prior to AMD exposed its Radeon RX 7900- series graphics cards— which would’ve made this GPU appearance silly at the very same $899 rate as the 7900 XT– Nvidia canceled its initial launch strategies. It’s back with a lower price that’s still unreasonably, annoyingly high for what’s being used.

Here’s a wrap-up of the RTX 4070 Ti’s core specifications:

• Graphics Processing Clusters: 5
• Texture Processing Clusters: 30
• Streaming Multiprocessors: 60
• CUDA Cores: 7680
• Tensor Cores: 240 (4th Gen)
• RT Cores: 60 (3rd Gen)
• Texture Units: 240
• ROPs: 80
• Boost Clock: 2610 MHz
• Memory Clock: 10500 MHz
• Memory Data Rate: 21 Gbps
• L2 Cache Size: 49152 KB
• Total Video Memory: 12 GB GDDR6X
• Memory Interface: 192- bit
• Total Memory Bandwidth: 504 GB/sec
• Texture Rate: 626 Gigatexels/sec
• Connectors: 3 x DisplayPort, 1 x HDMI
• Max Display Resolution: 4K at 240 Hz or 8K at 60 Hz with DSC
• Power Connectors: 2x PCIe 8-pin cable televisions OR 300 w or higher PCIe Gen 5 cable television (adapter consisted of in box)
• Minimum Power Supply: 700 Watts
• Power: 12 W Idle, 20 W AV1 Video Playback, 226 W Average Gaming Power (AGP), 285 W Total Graphics Power (TGP)
• Maximum GPU Temperature: 90 ° C
• PCI Express Interface: Gen 4

Yes, the 4070 Ti’s core count and memory capability are dramatically decreased versus the 4080– Nvidia made the ideal call (ultimately) with its name modification. Now that the name will not plant confusion among potential purchasers, the more worrying tweak revolves around the 4070 Ti’s memory bus.

Wider bus lanes let you move information better. Think about it like heavy traffic; a four-lane highway can keep traffic moving much faster than a single-lane dirt roadway. Bus width is a crucial aspect for general memory bandwidth, particularly at greater resolutions where textures and ray tracing can demolish memory quick

The GeForce RTX 4070 Ti loads a pipsqueak 192- bit memory bus. By contrast, the RTX 3080— which, once again, was a higher-tier card that cost $100 less than the 4070 Ti– wielded a 320- bit bus, which is normal for a high-end GPU. Even the already-pandemic-overpriced $600 RTX 3070 Ti had a 256- bit bus. You needed to go all the method to the $330 RTX 3060 to discover an Nvidia GPU with a 192- bit bus last generation.

The brand-new 4070 Ti deals 504 GB/s of overall memory bandwidth; the 3080 hit 760 GB/s and the 3070 Ti provided 608 GB/s, while the 3060 and its far slower GDDR6 non-X memory struck a proper 360 GB/s.

Brad Chacos/IDG

It matters. Compared to the RTX 3080, the 4070 Ti’s uplift in 4K video gaming is measurably slower than its 1440 p uplift. You can still play 4K video games with whatever cranked and still struck 60 frames per second without concern the majority of the time, however deliberately hobbling the 4K efficiency of a freaking $799 graphics card leaves a sour taste in the mouth. The 4070 Ti is tangibly slower than all other new-gen GPUs and the RTX 3080 in Counter-Strike: Global Offensive at 4K and 1440 p resolutions, and I would not be shocked if the neutered bus width elements into that. It would be more appropriate if this GPU costs numerous dollars less.

Fortunately, the RTX 4070 Ti is far more effective than the 3080 was, ranked for 285 watts of overall graphics power compared to the 3080’s 320 W. It’s quicker, too. Nvidia’s brand-new Ada Lovelace GPU architecture impresses there.

The customized Asus Tuf 4070 Ti we’re examining today– Nvidia isn’t releasing a Founders Edition variation of this GPU– still utilizes the brand-new 16- pin 12 VHPWR adapter, nevertheless, with an adapter consisted of in package. ( Make sure you plug it in comfortably) It’s likewise a big 3.25- slot graphics card in spite of the lower power draw. Asus otherwise did a great task with Tuf, which includes a striking commercial style, a dual-BIOS switch (we check at default), and not one, however 2 HDMI 2.0 b adapters to accompany 3 DisplayPorts. Discovering double HDMI ports on a GPU is unusual nowadays, and Asus just charges a $50 premium for this variation of the Tuf– however, once again, the RTX 4070 Ti is hugely overpriced to start with.

Brad Chacos/IDG

Nvidia keeps its lead in regards to using well-rounded functions and abilities that AMD merely can’t match. The Radeon RX 7900 duo is tangibly faster than the 4070 Ti in regards to raw rasterization (read: conventional video gaming) efficiency, however Nvidia’s GPU holds an enormous ray tracing efficiency benefit, reinforced by AI-powered DLSS and DLSS Frame Generation innovations. They’re increasingly remarkable. Usually speaking, Nvidia likewise thwomps AMD in CUDA assistance for artificial intelligence jobs (duh), general material development expertise, and with other life-enhancing functions like Nvidia Broadcast, Reflex, and Shadowplay. If you do more with your GPU than merely play video games, or if you think about peak ray tracing efficiency a top priority, target an Nvidia GPU.

Let’s get to the video games.

Our test system

We evaluate graphics cards on an AMD Ryzen 5900 X PC utilized specifically for benchmarking GPUs. We now check with PCIe Resizable BAR (likewise called Smart Access Memory on Ryzen systems) active, as many contemporary video gaming PCs launched in the last 4 years support the performance-boosting function, either natively or by means of a motherboard firmware upgrade. Nvidia likewise advises turning on the optional “Hardware-accelerated GPU scheduling” choice in Windows to let the RTX 40- series stretch its legs to the maximum, so we’ve made that fine-tune. The majority of the hardware was offered by the makers, however we bought the storage ourselves.

• AMD Ryzen 5900 X, stock settings
• AMD Wraith Max cooler
• MSI Godlike X570 motherboard
• 32 GB G.Skill Trident Z Neo DDR4 3800 memory, XMP active
• Corsair HX1500 i power supply(and optional $2012 VHPWR 600 cable television for Nvidia GPUs)
• 1TB SK Hynix Gold S31 SSD x2

We’re comparing Nvidia’s $799 GeForce RTX 4070 Ti versus all the contemporary GPUs launched by both Nvidia and AMD so far, in addition to last generation’s RTX 3080, 3080 Ti, and 3090 and Radeon RX 6900 XT. Time restrictions avoided us from benchmarking the $600 GeForce RX 3070 Ti, the 4070 Ti’s apparent predecessor, however offered this brand-new card’s jacked-up cost point the RTX 3080 is a more sincere contrast anyhow.

In addition to switching on PCIe Resizable BAR by default, we’ve likewise transferred to a brand-new set of video games in our screening suite for this brand-new generation. We evaluate a range of video games covering different engines, categories, supplier sponsorships (Nvidia, AMD, and Intel), and graphics APIs (DirectX 9, 11, DX12, and Vulkan), to attempt to represent a complete variety of efficiency capacity. Each video game is evaluated utilizing its in-game criteria, peace of mind examined by Nvidia’s FrameView tool, at the greatest possible graphics presets unless otherwise kept in mind, with VSync, frame rate caps, real-time ray tracing or DLSS impacts, and FreeSync/G-Sync handicapped, in addition to any other vendor-specific innovations like FidelityFX tools or Nvidia Reflex. We’ve likewise allowed temporal anti-aliasing (TAA) to press these cards to their limitations.

We run each standard a minimum of 3 times and list the typical outcome for each test. Beyond esports, we’ve restricted our standards to 4K and 1440 p resolutions, as those are the natural suitable for these juggernaut GPUs (regardless of the 4070 Ti’s paltry bus width).

We’re going to reveal these standards without specific video game commentary. In basic, take note of just how much quicker the 4070 Ti is compared to the $100 more affordable 3080, and how the 4070 Ti gets more speed on the 3080 at 1440 p than 4K. Keep in mind how the likewise overpriced $899 Radeon RX 7900 XT outshines the $7994070 Ti by a substantial quantity throughout the board. We didn’t have time to completely benchmark the 4070 Ti’s ray tracing expertise, however consisted of a chart proving Cyberpunk 2077 keeping up RT allowed (with DLSS and FSR both on, and off) to validate that, yep, Nvidia’s GPU architecture continues to blow AMD’s ray tracing efficiency out of the water. It isn’t even close.

Gaming efficiency criteria

Brad Chacos/IDG

Brad Chacos/IDG

Brad Chacos/IDG

Brad Chacos/IDG

Brad Chacos/IDG

Brad Chacos/IDG

Brad Chacos/IDG

Brad Chacos/IDG

Brad Chacos/IDG

Brad Chacos/IDG

GeForce RTX 4070 Ti: Power and thermals

We evaluate power draw by looping the F1 22 criteria at 4K for about 20 minutes after we’ve benchmarked whatever else (to heat up the GPU) and keeping in mind the greatest reading on our Watts Up Pro meter, which determines the power usage of our whole test system. The preliminary part of the race, where all contending automobiles are onscreen at the same time, tends to be the most requiring part.

This isn’t a worst-case test; this is a GPU-bound video game performing at a GPU-bound resolution to determine efficiency when the graphics card is sweating hard. If you’re playing a video game that likewise hammers the CPU, you might see greater total system power draws. Consider yourself alerted.

Brad Chacos/IDG

Despite being faster than the RTX 3080, the 4070 Ti takes in far less power both when idle and under load. Even if Nvidia isn’t moving the needle in regards to worth with the 4070 Ti, its effectiveness is no doubt improving.

We evaluate thermals by leaving GPU-Z open throughout the F1 22 power draw test, keeping in mind the greatest optimum temperature level at the end.

Brad Chacos/IDG

The custom-made Asus Tuf cooling style provides definitely freezing temperature levels under complete load. Overclockers might have a field day with this specific design. You hear some fan sound to accomplish those remarkable temperature levels, however it isn’t extremely loud or frustrating. I have no grievances about the Asus Tuf’s total thermal style.

Should you purchase the GeForce RTX 4070 Ti?

Please do not. It’s a monster of a card that reveals what Nvidia’s excellent software and hardware engineers can– however not that this eye-watering rate.

In a vacuum, the $799 GeForce RTX 4070 Ti’s rasterization efficiency underwhelms compared to the $899 Radeon RX 7900 XT other than in ray tracing efficiency, however Nvidia’s plentiful software application and function offerings– Studio chauffeurs, DLSS and Frame Generation, Broadcast, Reflex, CUDA, et cetera— can make that space much easier to swallow if you utilize your GPU for more than simply video gaming. In truth, the RTX 4070 Ti– like the RTX 4080 and Radeon 7900 XT prior to it– bring cost that make their worth stagnant at finest and borderline insulting at worst.

Brad Chacos/IDG

At $799, the GeForce RTX 4070 Ti expenses $100 more than last generation’s sensational RTX 3080– and it’s a lower-tier GPU. It’s $200 more than the already-overpriced RTX 3070 Ti was. In exchange for that 14 percent cost walking over the 3080, you normally get a stagnant 14 percent more efficiency than the 3080 at 4K, or approximately 20 percent much faster speeds at 1440 p. Nvidia cut the 4070 Ti’s memory bus back to RTX 3060 levels which slows it down at 4K. That’s flat-out inappropriate in a $799 graphics card

Gamers remain in a bad location. Nvidia and AMD’s flagships, the GeForce RTX 4090 and Radeon RX 7900 XTX, deal sensational efficiency gains over their predecessors for comparable rates. Whatever under those bring downright uncomfortable MSRPs compared to previous generations. Nvidia and AMD should still believe we’re in the middle of a pandemic with roaring cryptocurrency worth. We’re not. Miners were the ones purchasing graphics cards for agonizing, outrageous rates, not players. These graphics card expenses are ridiculous.

Nvidia’s RTX 4080 expenses $500 more than the 3080 did. Gross. Instead of taking the minute, AMD priced its feature-lacking Radeon RX 7900 XT at $899, when spec-wise it most likely must’ve been called the Radeon RX 7800 XT and priced a couple hundred dollars lower. AMD’s thirst for 7900 XT earnings margins likewise enabled Nvidia to reestablish the “unlaunched” devastating RTX 4080 12 GB with a brand-new name and a cost tag simply $100 less than in the past. It’s still very unattractive. You should not need to pay $800 dollars for what’s being actively marketed as a 1440 p graphics card in 2023, particularly with its 4K efficiency being hobbled by the below average memory bus.

Bah. It’s like WOPR stated in WarGames: “An odd video game. The only winning relocation is not to play.”

Part of the reason that Nvidia and AMD are keeping costs so high is to move their stocked last-gen GPUs now that cryptocurrency has actually crashed, at thoroughly handled high rates. Nvidia in specific keeps attempting to reveal the RTX 40- series as pricey alternatives on top of the 30- series lineup, instead of the generational replacement it genuinely is– a worryingly anti-consumer service choice. Buy a pre-owned graphics card rather if you aren’t going all-in on a 4090 or a 7900 XTX. You can discover likewise carrying out GPUs like the Radeon RX 6800 XT and 6900 XT or the RX 3080 for numerous dollars less than Nvidia and AMD are charging for these hugely pricey brand-new offerings. We have pointers to assist you prevent purchasing an utilized lemon

And I’m stating that regardless of being a man who generally advises that you prevent utilized graphics cards

Because paying $800 for a hobbled, significantly overpriced 1440 p-focused graphics card in 2023 ain’t it. The launch of this brand-new generation mostly mirrors the RTX 20- series launch, where stagnant gen-on-gen worth resulted in awful sales and the intro of an RTX 20 Super series quickly after to assist start matters … though that launch accompanied the release of AMD’s outstanding, sanely priced Radeon RX 5000- series. This generation, AMD appears material to ride the coattails of Nvidia’s extreme rates. Flagship GPUs should not provide the very best price-per-dollar worth, and the rest of your stack should not provide efficiency that scales nearly linearly with the prices of last-gen designs.

Discrete graphics cards sales simply struck a 20- year low Unless something extreme modifications, like a modern-day intro of Super series equivalents, releases like the GeForce RTX 4070 Ti and 4080 will likely just intensify the discomfort.

Not simply pass, however difficult pass. If you need to purchase a next-gen GPU around this rate point, the Radeon RX 7900 XT is likewise atrociously priced however it’s significantly faster and includes 20 GB of memory, 8GB more than the 4070 Ti. If you require Nvidia functions and should purchase a 4070 Ti, I extremely advise the Asus Tuf design.

But we’re in the middle of an unusual video game. Please do not purchase either. Not at these costs.