Nvidia RTX 4070 Ti Review ft. Gigabyte Gaming OC

1. Introduction2. Meet the Gigabyte Gaming OC3. Testing Methodology4. 3DMark Fire Strike & Time Spy5. 3DMark DirectX Raytracing6. Assassin's Creed Valhalla7. Cyberpunk 20778. Days Gone9. Dying Light 210. Far Cry 611. Forza Horizon 512. God of War13. Horizon Zero Dawn14. Marvel's Spider-Man Remastered15. Red Dead Redemption 216. Resident Evil Village17. Total War: Warhammer III18. Performance Summary (12-Game Average)19. Cost Per Frame Analysis20. Ray Tracing: Cyberpunk 207721. Ray Tracing: Marvel's Spider-Man Remastered22. Ray Tracing: Metro Exodus Enhanced Edition23. Ray Tracing: Resident Evil Village24. Clock Speed25. GPU Temperatures26. Memory Temperatures27. Acoustics28. Power Draw: 1080p29. Power Draw: 1440p30. Power Draw: 2160p (4K)31. Performance per Watt: 1080p32. Performance per Watt: 1440p33. Performance per Watt: 2160p (4K)34. Total System Power Draw35. Overclocking36. Closing Thoughts37. View All Pages

While Nvidia may have officially announced the RTX 4070 Ti at yesterday's CES 2023 keynote, you wouldn't be wrong to think of that as a relaunch, following on from the now infamous RTX 4080 12GB ‘unlaunch'. That's because the RTX 4070 Ti we are reviewing today is the same exact GPU as the RTX 4080 12GB, but with a £150 price cut and, of course, a new name. Hitting the market at £799, what can we expect from this ‘new' graphics card, and is it worth buying? We test gaming performance, overclocking, power draw, thermals and more to find out…

As the third RTX 40-series card to hit the market, the Nvidia RTX 4070 Ti also brings with it a third new Ada Lovelace GPU – AD104. The smallest piece of 40-series silicon yet, RTX 4070 Ti offers 21% fewer CUDA cores than the RTX 4080, with 25% less VRAM, but for a price cut of around 33%. On paper it certainly looks like it will offer better value than its more expensive sibling, but today we will find out whether or not that plays out in practice.

As always, we have plenty of benchmarks and charts to go over, including a detailed look at power draw and efficiency. If you want to read this review as a single page, click HERE.

	RTX 4090	RTX 4080	RTX 4070 Ti	RTX 3080 Ti	RTX 3080	RTX 3070 Ti
Process	TSMC N4	TSMC N4	TSMC N4	Samsung 8N	Samsung 8N	Samsung 8N
SMs	128	76	60	80	68	48
CUDA Cores	16384	9728	7680	10240	8704	6144
Tensor Cores	512	304	240	320	272	192
RT Cores	128	76	60	80	68	48
Texture Units	512	304	240	320	272	192
ROPs	176	112	80	112	96	96
GPU Boost Clock	2520 MHz	2505 MHz	2610 MHz	1665 MHz	1710 MHz	1770 MHz
Memory Data Rate	21 Gbps	22.4 Gbps	21 Gbps	19 Gbps	19 Gbps	19 Gbps
L2 Cache	73729 KB	65536 KB	49152 KB	6144 KB	6144 KB	4096 KB
Total Video Memory	24GB GDDR6X	16GB GDDR6X	12GB GDDR6X	12GB GDDR6X	10GB GDDR6X	8GB GDDR6X
Memory Interface	384-bit	256-bit	192-bit	384-bit	320-bit	256-bit
Memory Bandwidth	1008 GB/Sec	716.8 GB/Sec	504 GB/Sec	912 GB/Sec	760 GB/Sec	608 GB/Sec
TGP	450W	320W	285W	350W	320W	290W

First, for a quick spec recap. We already mentioned that RTX 4070 Ti uses the AD104 die, measuring 295mm². Accordingly, transistor count is reduced to 35.8 billion, down from 45.9 billion with AD103. The fundamental building blocks are still the same of course, with the RTX 4070 Ti offering a total of 60 Streaming Multiprocessors (SMs), each housing 256 CUDA Cores, for a total of 7680. We also find 60 RT cores, 240 Tensor cores, 240 Texture Units, and 80 ROPs.

TSMC's N4 node has Nvidia cranking up the clock speed significantly this generation, with the RTX 4070 Ti sporting a 2610MHz rated boost clock. That's a 105MHz increase over the rated clock speed of the RTX 4080, but we would expect GPU Boost to push things further still.

The memory configuration is another area where AD104 has been cut-back significantly. The memory interface has been reduced to 192-bit, and even with 12GB GDDR6X running at 21Gbps, that brings total memory bandwidth down to 504 GB/s, lower than even the RTX 3070 Ti. That said, there has been a substantial upgrade to the L2 cache with the Ada architecture, with the RTX 4070 Ti now offering 49MB, compared to just 6MB for GA102.

Considering these cut-backs, power draw is naturally lower than the RTX 4080, with the 4070 Ti boasting a 285W TGP. This is something we focus on closely, using our updated GPU power testing methodology in this review, so read on for our most detailed power and efficiency testing yet.

The Gigabyte RTX 4070 Ti Gaming OC ships in a black box, with a robotic-style avatar visible on the front. On the back, Gigabyte highlights various key features of the card and cooler design.

Inside we find a few pieces of documentation, alongside a new dual 8-pin power adapter for the 12VHPWR connector. An anti-sag GPU support bracket is also included which screws into a couple of your motherboard stand offs, before attaching to the end of the card to keep it level.

As for the card itself, it's a very similar overall design to the RTX 4080 and RTX 4090 Gaming OC models we have already reviewed, just scaled down slightly. That means we find the same black plastic shroud that's made entirely from plastic. It looks fine but doesn't feel particularly premium in the hand, but it's not the end of the world.

There's also three 100mm fans, using Gigabyte's so-called ‘Unique Blade Fan' design. As expected from a Gigabyte card, the central fan spins in reverse relative to the others, which Gigabyte claims helps to reduce airflow turbulence and therefore increase air pressure down into the heatsink.

This RTX 4070 Ti design is a bit smaller than the 4080/4090 models however. The RTX 4080 version comes in at 343 x 150 x 75mm, whereas the 4070 Ti measures 336 x 140 x 58mm. It's still a big graphics card, no doubt, and it weighs in at a touch over 1.5kg, it's just not quite so monstrous as its higher-end siblings.

The front side of the card is home to the GeForce RTX branding, as well as the Gigabyte logo, one of the RGB zones on the card.

As for the backplate, this is made from metal and while it is a full-length design, there is a large cut-out towards the end of the card to allow airflow to pass directly through the heatsink. We can also note the dual-BIOS switch position in the middle, offering a choice of the OC or Silent modes. Both are identical in terms of clock speed and power limits, with the fan curve being the only difference.

We can also note the 12VHPWR connector as expected, though at least the included dual 8-pin is a bit easier to manage than some of the triple of even quad 8-pin adapters we have seen in recent months.

Video outputs are standard for an Nvidia GPU, offering 3x DisplayPort 1.4 and 1x HDMI 2.1 connectors.

It's important to stress that, as there is no Founders Edition card for the RTX 4070 Ti, we were sampled with the Gigabyte Gaming OC model. While this has not increased the power limit from the reference figure of 285W, it does ship with a 30MHz factory overclock. For the purposes of our testing today however, where all other cards are tested at reference specifications, we have downclocked the card by 30MHz, using MSI Afterburner, so it operates at reference speeds.

Driver Notes

• All AMD GPUs (except RX 7900 series) were benchmarked with the Adrenalin 22.10.3 driver.
• All Nvidia GPUs (except RTX 4080 and RTX 4070 Ti) were benchmarked with the 526.61 driver.
• RTX 4070 Ti was benchmarked with the 527.62 driver supplied to press.
• RTX 4080 was benchmarked with the 526.72 driver supplied to press.
• RX 7900 XTX and RX 7900 XTX were benchmarked with the Adrenalin 22.40.00.57 driver supplied to press.

Test System:

We test using a custom built system powered by MSI, based on Intel’s Alder Lake platform. You can read more about this system HERE and check out MSI on the CCL webstore HERE.

CPU	Intel Core i9-12900K
Motherboard	MSI MEG Z690 Unify
Memory	32GB (2x16GB) ADATA XPG Lancer DDR5 6000MHz CL 40-40-40
Graphics Card	Varies
SSD	2TB MSI Spatium M480
Chassis	MSI MPG Velox 100P Airflow
CPU Cooler	MSI MEG CoreLiquid S360
Power Supply	Corsair 1200W HX Series Modular 80 Plus Platinum
Operating System	Windows 11 Pro 21H2
Monitor	MSI Optix MPG321UR-QD
Resizable BAR	Enabled for all supported GPUs

Comparison Graphics Cards List

• AMD RX 7900 XTX 24GB
• AMD RX 7900 XT 20GB
• Sapphire RX 6950 XT Nitro+ 16GB
• AMD RX 6900 XT 16GB
• AMD RX 6800 XT 16GB
• AMD RX 6800 16GB
• Nvidia RTX 4090 FE 24GB
• Nvidia RTX 4080 FE 16GB
• Gigabyte RTX 3090 Ti Gaming 24GB
• Gigabyte RTX 3090 Eagle 24GB
• Gigabyte RTX 3080 Eagle 10GB
• Nvidia RTX 3070 Ti FE 8GB

All cards were tested at reference specifications.

Software and Games List

• 3DMark Fire Strike & Fire Strike Ultra (DX11 Synthetic)
• 3DMark Time Spy (DX12 Synthetic)
• 3DMark DirectX Raytracing feature test (DXR Synthetic)
• Assassin's Creed Valhalla (DX12)
• Cyberpunk 2077 (DX12)
• Days Gone (DX11)
• Dying Light 2 (DX12)
• Far Cry 6 (DX12)
• Forza Horizon 5 (DX12)
• God of War (DX11)
• Horizon Zero Dawn (DX12)
• Marvel's Spider-Man Remastered (DX12)
• Metro Exodus Enhanced Edition (DXR)
• Red Dead Redemption 2 (DX12)
• Resident Evil Village (DX12)
• Total War: Warhammer III (DX11)

We run each benchmark/game three times, and present mean averages in our graphs. We use FrameView to measure average frame rates as well as 1% low values across our three runs.

Fire Strike is a showcase DirectX 11 benchmark for modern gaming PCs. Its ambitious real-time graphics are rendered with detail and complexity far beyond other DirectX 11 benchmarks and games. Fire Strike includes two graphics tests, a physics test and a combined test that stresses the CPU and GPU. (UL).

3DMark Time Spy is a DirectX 12 benchmark test for Windows 10 gaming PCs. Time Spy is one of the first DirectX 12 apps to be built the right way from the ground up to fully realize the performance gains that the new API offers. With its pure DirectX 12 engine, which supports new API features like asynchronous compute, explicit multi-adapter, and multi-threading, Time Spy is the ideal test for benchmarking the latest graphics cards. (UL).

Looking first at 3DMark, initial indicators are positive. The RTX 4070 Ti sits between the RTX 3090 Ti and RX 6900 XT in Fire Strike, remaining 4% faster than the 3090 Ti in Time Spy while only falling behind in the 4K Time Spy Extreme benchmark.

Real-time ray tracing is incredibly demanding. The latest graphics cards have dedicated hardware that’s optimized for ray-tracing. The 3DMark DirectX Raytracing feature test measures the performance of this dedicated hardware. Instead of using traditional rendering techniques, the whole scene is ray-traced and drawn in one pass. The result of the test depends entirely on ray-tracing performance. (UL).

As a test of pure ray tracing performance, the 3DMark DXR feature test is always an interesting benchmark. The RTX 4070 Ti outpaces the RTX 3090 Ti handily here, with a 9% performance boost, though it trails the RTX 4080 significantly, coming in 21% slower.

Assassin's Creed Valhalla is an action role-playing video game developed by Ubisoft Montreal and published by Ubisoft. It is the twelfth major instalment and the twenty-second release in the Assassin's Creed series, and a successor to the 2018's Assassin's Creed Odyssey. The game was released on November 10, 2020, for Microsoft Windows, PlayStation 4, Xbox One, Xbox Series X and Series S, and Stadia, while the PlayStation 5 version was released on November 12. (Wikipedia).

Engine: AnvilNext 2.0. We test using the Ultra High preset, DX12 API.

As ever, we start our game benchmarks with Assassin's Creed Valhalla. At 1440p, the RTX 4070 Ti is sandwiched by the AMD 6900 XT and 6950 XT, with the new Ada-based GPU averaging 115FPS. This makes it 4% faster than the RTX 3090 Ti, while it trails the RX 7900 XT by 12%.

At 4K, it still holds out fairly well, averaging just over 70FPS, putting it level with the RTX 3090 Ti and 6950 XT. That makes it a good 20% slower than the RTX 4080, and it's 11% behind the 7900 XT.

Cyberpunk 2077 is a 2020 action role-playing video game developed and published by CD Projekt. The story takes place in Night City, an open world set in the Cyberpunk universe. Players assume the first-person perspective of a customisable mercenary known as V, who can acquire skills in hacking and machinery with options for melee and ranged combat. Cyberpunk 2077 was released for Microsoft Windows, PlayStation 4, Stadia, and Xbox One on 10 December 2020. (Wikipedia).

Engine: REDengine 4. We test using the Ultra preset, DX12 API.

As for Cyberpunk 2077, 1440p performance is strong, with the RTX 4070 Ti averaging 92FPS. This puts it within touching distance of the RTX 3090 Ti, being just 4% slower, while it outpaces the 6950 XT by a couple of frames.

Up at 4K however, we do see the 4070 Ti drop off slightly. It's now 10% slower than the 3090 Ti, when it was just 4% slower at 1440p, while it's 19% behind the RX 7900 XT, when the margin was 14% at 1440p.

Days Gone is a 2019 action-adventure survival horror video game developed by Bend Studio and published by Sony Interactive Entertainment for the PlayStation 4 and Microsoft Windows. As part of Sony's efforts to bring more of its first-party content to Microsoft Windows following Horizon Zero Dawn, Days Gone released on Windows on May 18, 2021. (Wikipedia).

Engine: Unreal Engine 4. We test using the Very High preset, DX11 API.

Next is Days Gone, where at 1440p the RTX 4070 Ti delivers over 150FPS on average, putting it between the RTX 3090 and 3090 Ti, though it's just 6% slower than the RX 7900 XT.

It falls off again at 4K however, where it's not matching the RTX 3080 Ti, coming in 12% slower than the RTX 3090 Ti. 85FPS is still pretty good at this resolution, but the relative performance is strongest at 1440p.

Dying Light 2: Stay Human is a 2022 action role-playing game developed and published by Techland. The sequel to Dying Light (2015), the game was released on February 4, 2022 for Microsoft Windows, PlayStation 4, PlayStation 5, Xbox One, and Xbox Series X/S. (Wikipedia).

Engine: C-Engine. We test using the High preset, DX12 API.

Dying Light 2 tells a similar story. We see solid performance at 1440p, with the 4070 Ti delivering over 130FPS, so it's within 7% of the RTX 3090 Ti.

Stepping up to 4K sees it slide down the chart however, where it is 17% slower than the 3090 Ti, and not much faster than the RTX 3080.

Far Cry 6 is a 2021 action-adventure first-person shooter game developed by Ubisoft Toronto and published by Ubisoft. It is the sixth main instalment in the Far Cry series and the successor to 2018's Far Cry 5. The game was released on October 7, 2021, for Microsoft Windows, PlayStation 4, PlayStation 5, Xbox One, Xbox Series X/S, Stadia, and Amazon Luna. (Wikipedia).

Engine: Dunia Engine. We test using the Ultra preset, HD Textures enabled, DX12 API.

As for Far Cry 6, as we've mentioned a few times, the 1440p results aren't particularly useful as most of the GPUs tested are CPU-limited at 1440p, which is why we can see the 4070 Ti coming in just 5% slower than the RTX 4080.

Up at 4K though, performance is decent enough. We get just under 85FPS on average, putting performance between the 3080 Ti and 3090, but 19% behind the R 7900 XT.

Forza Horizon 5 is a 2021 racing video game developed by Playground Games and published by Xbox Game Studios. The twelfth main instalment of the Forza series, the game is set in a fictionalised representation of Mexico. It was released on 9 November 2021 for Microsoft Windows, Xbox One, and Xbox Series X/S. (Wikipedia).

Engine: ForzaTech. We test using the Extreme preset, DX12 API.

Forza Horizon 5 is a tricky one for Nvidia GPUs at the minute, with some choppier than expected 1% lows, though the 4070 Ti does manage to come in 10% faster than the RTX 3090 Ti at 1440p which is the biggest victory for the Ada GPU we will see all day.

Stepping up to 4K still sees the card deliver almost 100FPS on average, though the 1% lows are noticeably better on the likes of the RX 6900 XT and 7900 XT.

God of War is an action-adventure game developed by Santa Monica Studio and published by Sony Interactive Entertainment (SIE). It was released worldwide on April 20, 2018, for the PlayStation 4 with a Microsoft Windows version released on January 14, 2022. (Wikipedia).

Engine: Sony Santa Monica Proprietary. We test using the Ultra preset, DX11 API.

God of War is another with odd 1% low performance at 1440p, where we can see the RTX 4070 Ti trails the 3090 Ti by 8%, while it's about 20% slower than the RTX 4080.

At 4K, things even out slightly, with the 4070 Ti now 10% slower than Nvidia's last-gen flagship, the RTX 3090 Ti. It's not miles behind the RX 7900 XT though, with just a 13% deficit.

Horizon Zero Dawn is an action role-playing game developed by Guerrilla Games and published by Sony Interactive Entertainment. The plot follows Aloy, a hunter in a world overrun by machines, who sets out to uncover her past. It was released for the PlayStation 4 in 2017 and Microsoft Windows in 2020. (Wikipedia).

Engine: Decima. We test using the Ultimate Quality preset, DX12 API.

As for Horizon Zero Dawn, this game is pretty typical of overall performance at 1440p, with the 4070 Ti hovering around the RX 6950 XT and RTX 3090 Ti.

It does again drop off slightly at 4K though, where it's just 7% faster than the RTX 3080, and 13% slower than the 3090 Ti.

Marvel's Spider-Man Remastered is a 2018 action-adventure game developed by Insomniac Games and published by Sony Interactive Entertainment. A remastered version of Marvel's Spider-Man, featuring all previously released downloadable content, was released for the PlayStation 5 in November 2020 and for Microsoft Windows in August 2022. (Wikipedia).

Engine: Insomniac Games Proprietary. We test using the Very High preset, DX12 API.

Spider-Man Remastered is more positive for the 4070 Ti though. At 1440p it delivers about 150FPS, making it just 6% slower than the 3090 Ti, and it's within a 10% margin of the 7900 XT.

It also doesn't fall off as much at 4K, and in fact closes the gap on the 7900 XT to just 8%. Averaging almost 90FPS at 4K is pretty good going, too.

Red Dead Redemption 2 is a 2018 action-adventure game developed and published by Rockstar Games. The game is the third entry in the Red Dead series and is a prequel to the 2010 game Red Dead Redemption. Red Dead Redemption 2 was released for the PlayStation 4 and Xbox One in October 2018, and for Microsoft Windows and Stadia in November 2019. (Wikipedia).

Engine: Rockstar Advance Game Engine (RAGE). We test by manually selecting Ultra settings (or High where Ultra is not available), TAA, DX12 API.

Performance from the RTX 4070 Ti is very strong in Red Dead Redemption 2, at least at 1440p – here it delivers almost 120FPS on average, putting it level with the RTX 3090 Ti and RX 6950 XT.

It does fall off a bit at 4K, coming in 8% slower than the 3090 Ti, but it's still averaging over 70FPS. Relative to the RTX 4080, we're looking at 22% less performance.

Resident Evil Village is a survival horror game developed and published by Capcom. The sequel to Resident Evil 7: Biohazard (2017), players control Ethan Winters, who is searching for his kidnapped daughter; after a fateful encounter with Chris Redfield, he finds himself in a village filled with mutant creatures. The game was announced at the PlayStation 5 reveal event in June 2020 and was released on May 7, 2021, for Windows, PlayStation 4, PlayStation 5, Xbox One, Xbox Series X/S and Stadia. (Wikipedia).

Engine: RE Engine. We test using the Max preset, with V-Sync disabled, DX12 API.

Resident Evil Village follows a very similar trend to RDR2. At 1440p, the RTX 4070 Ti basically matches the RTX 3090 Ti, while it's about 15% slower than the RX 7900 XT.

At 4K, the 4070 Ti drops back a touch, where it's 9% slower than the 3090 Ti and 19% behind the RX 7900 XT, though it still averages 120FPS.

Total War: Warhammer III is a turn-based strategy and real-time tactics video game developed by Creative Assembly and published by Sega. It is part of the Total War series, and the third to be set in Games Workshop's Warhammer Fantasy fictional universe (following 2016's Total War: Warhammer and 2017's Total War: Warhammer II). The game was announced on February 3, 2021 and was released on February 17, 2022.(Wikipedia).

Engine: TW Engine 3 (Warscape). We test using the Ultra preset, with unlimited video memory enabled, DX11 API.

Lastly, we close our benchmarks with a look at Total War: Warhammer III. At 1440p, the RTX 4070 Ti is able to deliver over 115FPS on average, making it 7% slower than the RTX 3090 Ti.

Once more we see performance drop off as we increase the resolution to 4K, with the 4070 Ti now 15% slower than the 3090 Ti. It's also only just 7% faster than the RTX 3080, which isn't a particularly impressive performance uplift considering the price.

Here we present frame rate figures for each graphics card, averaged across all 12 games on test today. These figures can disguise significant variations in performance from game to game, but provide a useful overview of the sort of performance you can expect at each resolution tested.

On average then, at 1440p the RTX 4070 Ti is within touching distance of the RTX 3090 Ti, as it's barely 2% slower across the 12 games tested, while it's just 2% faster than than the RX 6950 XT. That makes it 13% slower than the RX 7900 XT, and 18% faster than the RTX 3080 10GB.

As we saw a number of times today, the 4070 Ti does drop off at 4K as a result of its limited memory bandwidth, with less throughput than even the 3070 Ti. We can see the card is now 9% slower than the RTX 3090 Ti, and 14% faster than the RTX 3080 10GB. It's also 14% slower than the RX 7900 XT, and 22% behind the RTX 4080.

The MSRPs

Using the average frame rate data presented earlier in the review, here we look at the cost per frame using the UK MSRP launch prices for each GPU.

As for overall value then, the cost per frame data at 1440p, based on MSRP pricing, comes in at £5.71, and that makes it very similar to the RX 7900 XT and RX 7900 XTX. It's still not a step forward from the previous generation, but it's a significant improvement of about 23% compared to the RTX 4080.

Up at 4K, we see a similar story, though due to the 4070 Ti's weaker performance at this resolution, it's now a 19% improvement in cost per frame compared to the RTX 4080, but does come in slightly more expensive per frame than the likes of the RTX 3080 10GB.

Current retail pricing

Current retail pricing does tell a different story however, and there's not much difference between the 4070 Ti and the RX 6800 which sits at the top of the chart. The value proposition is basically the same as the RTX 3080, but an 18% improvement over the RTX 4080.

Likewise, at 4K, cost per frame matches the RTX 3080 – after all, the 4070 Ti is 14% faster than the RTX 3080 10GB, but 14% more expensive based on current pricing, so there's no real progression there in terms of value for money. It is, at least, better than any other 40-series card so far, and is within 4% of the RX 7900 XT.

Here we test Cyberpunk 2077, using the Ray Tracing: Ultra preset. DLSS/FSR are disabled.

Ray tracing performance in Cyberpunk 2077 scales very similarly to the rasterised numbers we saw earlier. At 1440p we're looking at almost identical results to the RTX 3090 Ti, so you would want to enable DLSS for a smoother experience, but the 4070 Ti is still 18% ahead of AMD's fastest, the RX 7900 XTX.

Up at 4K, we're not looking at playable frame rates using Ultra settings in Cyberpunk, so quickly moving on…

Here we test Marvel's Spider-Man Remastered, with the in-game ray tracing effects set to Very High, Object Range set to 6. DLSS/FSR are disabled.

Spider-Man Remastered is next, where the 4070 Ti can average 110FPS with ray tracing enabled at 1440p, though this does also introduce a fair CPU bottleneck at this resolution.

Up at 4K, the 4070 Ti can still average over 60FPS, and it's basically dead level with the RX 7900 XT despite being the cheaper card.

Here we test Metro Exodus Enhanced Edition, with the in-game ray tracing effects set to Ultra. DLSS/FSR are disabled.

Metro Exodus Enhanced Edition also sees the RTX 4070 Ti edging-out the RX 7900 XTX, though with worse 1% low performance overall.

Stepping up to 4K however sees the AMD GPUs fall off even more, and the 4070 Ti is about 6% slower than the RTX 3090 Ti.

Here we test Resident Evil Village, this time testing with the in-game ray tracing effects set to High. DLSS/FSR are disabled.

Lastly, for Resident Evil Village and its relatively light ray traced effects, the 4070 Ti is within 4% of the RX 7900 XTX at 1440p, while it's level with the RTX 3090 Ti.

As for 4K performance, it does back off slightly but is still on-par with the 7900 XT, though it is now 9% slower than the RTX 3090 Ti.

Here we present the average clock speed for each graphics card while running Cyberpunk 2077 for 30 minutes. We use GPU-Z to record the GPU core frequency during gameplay. We calculate the average core frequency during the 30 minute run to present here.

Looking at the RTX 4070 Ti's clock speed under load, we have three data points shown on the above chart. As mentioned earlier, we downlocked the Gaming OC by 30MHz so it would match reference specifications, where we saw the card boost to 2790MHz. Using the default OC BIOS, with the 30MHz factory OC, resulted in the card running exactly 30MHz faster, averaging 2820MHz over our thirty minute stress test.

Interestingly, the GPU clock speed proved to be absolutely locked in at 2820MHz during our stress test. Usually the GPU frequency bounces around somewhat as the chip hits its power limit and drops back, but the 4070 Ti does not appear to be power limited at all, so we get a completely flat line when looking at the frequency chart.

For our temperature testing, we measure the peak GPU core temperature under load. A reading under load comes from running Cyberpunk 2077 for 30 minutes.

As for thermal performance, considering the size of the Gaming OC and the vapour chamber, it's not really a surprise to see excellent results here. We don't yet know how this will compare to other AIB models, but with the OC BIOS keeping GPU temperatures below 60C, it's hard to imagine much of an improvement will be had.

For our temperature testing, we measure the peak memory temperature under load. A reading under load comes from running Cyberpunk 2077 for 30 minutes.

Memory thermals are also very low indeed, with the OC BIOS hitting just 54C, while the Silent BIOS – which runs a slower fan curve – still only saw the GDDR6X memory peak at 62C.

We take our noise measurements with the sound meter positioned 1 foot from the graphics card. I measured the noise floor to be 32 dBA, thus anything above this level can be attributed to the graphics cards. The power supply is passive for the entire power output range we tested all graphics cards in, while all CPU and system fans were disabled. A reading under load comes from running Metro Exodus Enhanced Edition for 30 minutes.

Speaking of fan curves, the Gaming OC's Silent BIOS is well worth using if you prefer a lower-noise experience, as the fans produced just 36dBa of noise. That's compared to 40dBa using the OC BIOS, where the three fans spun up to about 1800rpm, which still isn't awfully loud, but it's not super quiet either. There was no noticeable coil whine either.

Here we present power draw figures for the graphics card-only, on a per-game basis for all twelve games we tested at 1080p. This is measured using Nvidia's Power Capture Analysis Tool, also known as PCAT. You can read more about our updated power draw testing methodology HERE.

Per-Game Results at 1080p:

Click to enlarge.

12-Game Average at 1080p:

As we've come to expect from the Ada Lovelace GPUs, real-world power draw when gaming comes in well below the 285W TGP. Granted, this is hardly a 1080p card and there are a few occasions where the GPU isn't fully utilised at this resolution, but the average power draw of 222.6W is still significantly below TGP.

Here we present power draw figures for the graphics card-only, on a per-game basis for all twelve games we tested at 1440p. This is measured using Nvidia's Power Capture Analysis Tool, also known as PCAT. You can read more about our updated power draw testing methodology HERE.

Per-Game Results at 1440p:

Click to enlarge.

12-Game Average at 1440p:

Up at 1440p, power draw does increase as expected, up to an average of 242W, though that's still about 45W below TGP. Even then, power draw is only 15W, or 6%, lower than the RTX 4080, despite the 4070 Ti being significantly slower, so that will have some implications for efficiency which we will get to shortly.

Here we present power draw figures for the graphics card-only, on a per-game basis for all twelve games we tested at 2160p (4K). This is measured using Nvidia's Power Capture Analysis Tool, also known as PCAT. You can read more about our updated power draw testing methodology HERE.

Per-Game Results at 2160p (4K):

Click to enlarge.

12-Game Average at 2160p (4K):

Lastly, up at 4K, we see a new average power draw of 255.5W. Certain games do push more than that – including Cyberpunk, Days Gone and Resident Evil Village (as shown above), but on average, that's 25W lower power draw than the RTX 4080, or just a 9% reduction.

Using the graphics card-only power draw figures presented earlier in the review, here we present performance per Watt on a per-game basis for all twelve games we tested at 1080p.

Per-Game Results at 1080p:

Click to enlarge.

12-Game Average at 1080p:

In terms of overall efficiency, at 1080p we can see performance per Watt that is equal to the RTX 4090, but a good 11% behind the RTX 4080. Still, efficiency is better than AMD's closest challenger, the RX 7900 XT, by a 22% margin.

Using the graphics card-only power draw figures presented earlier in the review, here we present performance per Watt on a per-game basis for all twelve games we tested at 1440p.

Per-Game Results at 1440p:

Click to enlarge.

12-Game Average at 1440p:

Looking at 1440p, overall efficiency is slightly worse – in relative terms – than what we saw at 1080p. The 4070 Ti is now slightly less efficient than the RTX 4090 for instance, while it now offers 13% lower performance per Watt than the RTX 4080. It's still 12% more efficient than the RX 7900 XT, but usually we'd expect the lower-power GPUs to be more efficient than the high-power SKUs (like the 4080 and 4090), and that isn't the case here.

Using the graphics card-only power draw figures presented earlier in the review, here we present performance per Watt on a per-game basis for all twelve games we tested at 2160p (4K).

Per-Game Results at 2160p (4K):

Click to enlarge.

12-Game Average at 2160p (4K):

Lastly, at 4K, overall efficiency of the RTX 4070 Ti is now a match for the RX 7900 XTX, with performance per Watt coming in 12% lower than the RTX 4090, and 14% below the RTX 4080. It's perhaps surprising but this is the least efficient RTX 40-series GPU released so far.

We measure system-wide power draw from the wall while running Cyberpunk 2077 for 30 minutes. We do this at 1080p, 1440p and 2160p (4K) to give you a better idea of total system power draw across a range of resolutions, where CPU power is typically higher at the lower resolutions.

In terms of total system power draw, we measured a peak of 470W in Cyberpunk 2077 when running at 1080p, with less power drawn at 1440p and 4K. Nvidia officially recommends a 700W power supply.

For our manual overclocking tests, we used MSI Afterburner. Our best results are as below.

Overclocking the RTX 4070 Ti was surprisingly fun, and I got more out of it than I thought I would. For starters, I maximised the power limit to 119% (340W) and was able to successfully add 230MHz to the GPU. Increasing the voltage slider also helped, with the real-world clock speed increasing by about 45MHz.

Overclocking the memory, however, was absolutely nuts. I kept pushing it and it remained stable even at +2000MHz, which is as far as the slider will go in MSI Afterburner. I have no idea if I just hit the silicon lottery with my sample or if other 4070 Ti's will be able to do the same, but it was quite fun to keep pushing and pushing until I could go no further.

As a result of this extra frequency, the 4070 Ti was able to complete a 30-minute stress test, averaging 3090MHz – a pretty sizable increase over stock.

As a result of this overclock, we actually saw pretty decent gains – with a 10% boost in Cyberpunk 2077, 13% in Horizon Zero Dawn, and 12% in Resident Evil Village. This is the most success I've had overclocking Ada Lovelace, so it could be well worth doing if your sample is as capable as mine.

Power draw did also increase as a result of this overclock, by 13% – up to 272.8W. This scales pretty closely with the overall increase in performance so it's not like the overclock has improved efficiency, but I'd say the performance gain is worth it.

As the third RTX 40-series graphics card to hit the market, the RTX 4070 Ti has taken over from the ‘unlaunched' RTX 4080 12GB, offering the same core design and specs, but with a £150 price cut and a xx70-series moniker, rather than the xx80- SKU that it was originally intended to be.

All told, I have to say that decision was undoubtedly a wise move on Nvidia's part. Had this GPU released at £949, as it was originally announced, it would have been very underwhelming and clearly inferior to the RX 7900 XTX. Even now at £799 it's not going to blow your socks off, but the overall package is more appealing at the lower price-point.

That's certainly the case when looking at overall performance, as this is a capable gaming GPU. I’d say it is best suited for high refresh-rate 1440p gaming, where it’s basically a match for the RTX 3090 Ti and is about 13% slower on average than the RX 7900 XT. It could also be a solid entry level 4K card, though it performs worse, relatively speaking, at the UHD resolution due to its pretty narrow memory interface; while it roughly matches the 3090 Ti at 1440p, it’s 9% slower at 4K. It can do a job and will still hit 60FPS in most titles, but the memory bandwidth limitation is much less of a factor at 1440p where the 4070 Ti is delivering 120FPS at ultra settings more often than not.

Ray tracing performance scales similarly, too. Granted, you won't be enabling RT Ultra settings at 4K, but even in Cyberpunk 2077 the 4070 Ti was able to average 45FPS at 1440p with the eye candy cranked up to the limit, and that's without touching DLSS. Less demanding ray-traced titles, including Spider-Man Remastered, run significantly better, with performance generally equal to, if not ahead of, the more expensive RX 7900 XT.

In lieu of a Founders Edition design, today we have tested the Gigabyte Gaming OC model, and found it to be an excellent third-party design. It runs very quiet and very cool, with a choice of BIOS modes to boot and a fairly understated overall aesthetic.

Power draw of the 4070 Ti, however, is slightly higher than I would have hoped considering the performance levels. To be clear, it almost always drew less power than its 285W TGP in our testing, but the difference in power between itself and the RTX 4080 was never particularly significant.

Considering the performance difference between those two cards is fairly wide, at around 20%, the RTX 4070 Ti is actually the least efficient 40-series GPU so far, which is slightly odd given the lower-power SKUs are usually the more efficient. Overall efficiency is not bad by any means and it’s still a good chunk better than the 30-series, but it's slightly concerning that we’re not trending in the right direction as we head down the 40-series stack. For the full set of charts detailing power and efficiency, head to page 28 of this review.

All things considered, I do think the RTX 4070 Ti is a step in the right direction for the RTX 40-series as a whole. Coming in about 20% slower than the RTX 4080, but 33% cheaper, clearly makes the 4070 Ti a much better value, and it works out at 23% cheaper per frame at 1440p and 19% cheaper per frame at 4K.

However, there will still be those pointing out that, compared to the 30-series, the 4070 Ti is still very much a continuation in terms of value and cost per frame, and that's equally valid. Compared to the RTX 3080 10GB for instance, the 4070 Ti is 14% faster at 4K, but 14% more expensive based on current retail prices. Against the RTX 3070 Ti, the 4070 Ti is 46% more expensive comparing MSRP prices, but 43% faster at 1440p.

Clearly, it wasn't hard for the RTX 4070 Ti to look good against the RTX 4080 as it was never a very good value to begin with, but those hoping for a real generational improvement in cost per frame will be left disappointed. That said, there are other features to consider including DLSS 3 Frame Generation, which is definitely a value add, while ray tracing is generally more viable at 1440p with the 4070 Ti than previous 30-series cards.

At the end of the day, the Nvidia RTX 4070 Ti is not a bad product. It’s not particularly amazing and I can't imagine many will get overly excited about it, but if you were already going to drop £700 on an RTX 3080, I’d say stepping up to the 4070 Ti makes sense considering the DLSS 3 advantage, as well as the improvements to ray tracing and overall efficiency versus the 30-series.

As a final point however, I do worry that Nvidia is cutting back too much too soon, particularly in terms of the 192-bit memory interface. After all, if this is what is being offered at the £799 price point, who knows what we’ll end up with once the RTX 4060 and even RTX 4050 eventually arrive…

The RTX 4070 Ti will be available from Overclockers UK with an MSRP of £799.

Discuss on our Facebook page HERE.

Pros

• Strong performance that's roughly on par with the RTX 3090 Ti.
• Best value 40-series GPU so far.
• Great ray tracing capabilities for 1440p gaming.
• DLSS 3 is a promising new technology and a value-add for the 40-series.
• Our sample overclocked very well.
• Gigabyte Gaming OC runs quiet and cool.

Cons

• Least efficient 40-series card yet.
• No real improvement in cost per frame versus 30-series.
• Significantly more expensive than previous xx70 SKUs.

KitGuru says: The RTX 4070 Ti may not be a revolution in cost per frame, but it's a solid enough card that we'd take over the RTX 3080.

Become a Patron!

Be sure to check out our sponsors store EKWB here