AliDropship is the best solution for dropshipping

Reflecting on the past, it’s hard to believe how uncomplicated it once was to purchase video cards. The surge in demand due to cryptocurrency enthusiasts, coupled with supply chain disruptions during the pandemic, has driven GPU prices to unprecedented heights, well above standard manufacturer rates. Ideally, I’d like to present NVIDIA’s RTX 5070 and RTX 5070 Ti as more affordable options compared to the exorbitantly priced $2,000 RTX 5090 and the $1,000 RTX 5080. Unfortunately, retailers and manufacturers have already pushed the prices of the RTX 5070 beyond their suggested retail prices. Currently, our review unit, the ASUS 5070 Ti Prime, is priced at $900 at Best Buy and $750 at Newegg (but we’ll see how long that lasts). Naturally, both retailers show the item as out of stock.

Although I can’t promise a set price for any RTX 5070 Ti model, I can assure you of this: these GPUs will undoubtedly perform well in 4K gaming, offering a far more economical alternative to the RTX 5080 and 5090. If you’re not in urgent need of an upgrade, it may be wise to wait a few months for pricing and stock availability to normalize.

Image for the large product module

ASUS

The NVIDIA RTX 5070 Ti stands out as a powerful GPU capable of delivering impressive 4K gaming performance, primarily due to its DLSS 4 capabilities. It makes an excellent upgrade choice for gamers seeking enhanced power and memory over the standard 5070.

Pros

  • Outstanding 4K performance
  • DLSS 4 multi-frame generation offers significant enhancements
  • Remains cool even under heavy load
Cons

  • Prices can vary significantly
  • Availability is a major concern
  • At $750, it’s still a significant investment for a mid-range GPU

$900 at Best Buy

With its specifications and anticipated pricing, the RTX 5070 Ti notably delivers the best balance of performance and value within NVIDIA’s range. It boasts 8,960 CUDA cores and 16GB of GDDR7 VRAM, which, although fewer than the 5080’s 10,752 CUDA cores, still offers ample memory. Conversely, the more affordable 5070 is limited to 12GB of VRAM, which might present challenges for 4K gaming.

Our ASUS 5070 Ti unit maintains a straightforward design, featuring three fans, a plastic casing, and a conventional heatsink layout. Gamers can select between performance and quieter BIOS modes, modifying how aggressive the fans operate. Its 2.5-slot design allows it to fit nicely into compact cases, although it is slightly larger than the RTX 5090 Founders card.

RTX 5090

RTX 5080

RTX 5070 Ti

RTX 5070

RTX 4090

Architecture

Blackwell

Blackwell

Blackwell

Blackwell

Lovelace

CUDA cores

21,760

10,752

8,960

6,144

16,384

AI TOPS

3,352

1,801

1,406

988

1,321

Tensor cores

5th Gen

5th Gen

5th Gen

5th Gen

4th Gen

RT cores

4th Gen

4th Gen

4th Gen

4th Gen

3rd Gen

VRAM

32 GB GDDR7

16 GB GDDR7

16 GB GDDR7

12 GB GDDR7

24 GB GDDR6X

Memory bandwidth

1,792 GB/sec

960 GB/sec

896 GB/sec

672 GB/sec

1,008 GB/sec

TGP

575W

360W

300W

250W

450W

The RTX 5070 Ti is also designed to fit into a broader range of gaming setups without requiring a power supply upgrade. Its peak power consumption is rated at 300 watts, which is lower compared to the 5080’s 360W and the 5090’s significant 575W. This means that the 5070 Ti can comfortably operate with an 850W PSU, avoiding the need for a costly shift to a higher-capacity 1,000W unit.

NVIDIA RTX 5070 Ti

Devindra Hardawar for Engadget

What sets this GPU apart is its comprehensive support for multi-frame generation found in DLSS 4, NVIDIA’s AI upscaling technology. This feature allows the graphics card to create up to three AI-generated frames for every frame it renders in real-time. This capability also enables NVIDIA to claim that the 5070 can rival the performance of the $1,599 RTX 4090. Although some may argue that these frames are merely “simulated” to improve benchmark figures, my experience with the RTX 5070 Ti alongside the 5090 demonstrates that they contribute to smoother gameplay overall.

In addition to multi-frame generation, various DLSS 4 features are gradually being introduced to earlier NVIDIA models. As noted in my 5090 review, “RTX 40 series cards will exhibit enhanced efficiency in single-frame generation, while RTX 30 and 20 series cards will also benefit from advancements in AI transformer models utilized for ray reconstruction (resulting in more stable ray tracing), Super Resolution (enhanced texture quality), and Deep Learning Anti-Aliasing (DLAA).”

NVIDIA RTX 5070 Ti

Devindra Hardawar for Engadget

Initially, it’s essential to note that the RTX 5070 Ti is only marginally faster than the 4070 and 4070 Ti across most benchmark tests. In 3DMark Timespy Extreme, the new card outperformed the 4070 Ti by 17 percent, and it was 21 percent faster than the 4070 Ti Super in the Speedway benchmark. The difference is even less pronounced in raw computing and rendering tasks, with the 5070 Ti achieving just an 8 percent lead over the 4070 Ti in the Geekbench 6 GPU benchmark.

NVIDIA RTX 5070 Ti

3DMark TimeSpy Extreme

Geekbench 6 GPU

Cyberpunk (4K RT Overdrive DLSS)

Blender

NVIDIA RTX 5070 Ti

12,675

238,417

153fps (4X frame gen)

7,365

NVIDIA RTX 5090

19,525

358,253

246fps (4X frame gen)

14,903

NVIDIA RTX 4070 Ti Super

11,366

220,722

75fps (1x frame gen)

7,342

NVIDIA RTX 4070

8,610

N/A

45fps (1x frame gen)

6,020

Ultimately, actual gaming performance is far more crucial than benchmark results. If you’re playing a title that supports DLSS 4, you’ll experience noticeable improvement. For instance, Dragon Age: The Veilguard maintained a solid 200fps at 4K with 4X multi-frame generation, ray tracing, and graphics settings maxed out. In comparison, the 4070 Ti typically averaged between 90fps and 100fps under the same conditions with DLSS 3.5 employing single-frame generation.

While I wouldn’t claim that the gameplay appeared twice as smooth on my Alienware 32-inch QD-OLED monitor, it certainly felt incredibly smooth during my testing. There were no noticeable upscaling artifacts, and those additional frames genuinely enhanced the experience. It’s worth mentioning that the RTX 5090 recorded 240fps in Dragon Age with the same settings. Although my CPU (a Ryzen 9 7900X) may have slightly limited performance, the RTX 5070 Ti still performed impressively close despite being a more cost-effective option.

Cyberpunk 2077 also performed beautifully at 4K in ray tracing overdrive mode with multi-frame generation, averaging 150fps. Although this is notably below the 5090’s remarkable 250fps, it’s still impressive for a game notorious for demanding an extensive hardware setup. Additionally, it managed to hit 230fps at 1440p with similar settings, which upscales excellently to 4K displays.

For titles that don’t utilize DLSS 4, such as Halo Infinite, the 5070 Ti continues to excel, averaging 140fps with maximum graphics settings and ray tracing enabled. Comparatively, the RTX 5090 hit an average of 180fps. For gamers fortunate enough to own a 240Hz 4K monitor, the 5070 Ti’s performance should satisfy even the most demanding users. If frame rates are more critical than resolution, it remains highly capable; I observed 220fps in Halo Infinite at 1440p and 320fps at 1080p.

During normal operation, the ASUS 5070 Ti maintained a temperature range of 30C to 35C, quickly peaking at 65C under load. While its cooling system isn’t as sophisticated as the 5090 Founder’s card, it effectively cooled the card below 40C in about 15 seconds.

NVIDIA RTX 5070 Ti

Devindra Hardawar for Engadget

In summary, the RTX 5070 Ti handled everything I tested without issues, and I hardly missed the 5090 (except for the pride factor). Although I did not get to test the RTX 5080, its high price point would make me hesitant to recommend it to most users.

The important question for gamers is whether you require the 16GB of VRAM and higher CUDA count provided by the RTX 5070 Ti. If you’re primarily gaming at 4K, having more than the 5070’s 12GB of RAM is beneficial, as games become increasingly complex each year. It’s likely that soon enough, 16GB of VRAM will be necessary for optimal 4K gaming. However, if you’re at 1440p, you should find that 12GB will be adequate for quite some time.

The biggest selling point of NVIDIA’s 50-series cards, specifically DLSS 4’s multi-frame generation, is primarily advantageous for 4K gaming. If you’re content with your 40-series card and not pushing a 4K, 240Hz monitor, there’s little reason to upgrade. However, if you own a 30-series or 20-series card, patience will indeed be rewarded.

As reiterated previously, it is worthwhile to wait a few months to observe how prices adjust. If you’re fortunate enough to acquire the RTX 5070 Ti for $750, it’s a great deal. However, anything above that threshold diminishes its appeal, especially as it creeps closer to the $1,000 MSRP of the 5090.

We’re still anticipating AMD’s next releases, the RDNA 4 Radeon 9070 and 9070 XT GPUs, which are positioned as direct competitors to the 5070 and 5070 Ti. AMD will also introduce AI-powered upscaling technology this year, suggesting that the gap between their offerings and NVIDIA’s might narrow. Nevertheless, NVIDIA holds a considerable advantage and may take time for AMD’s Fluid Motion Frames technology to catch up in multi-frame generation capabilities.

NVIDIA RTX 5070 Ti

Devindra Hardawar for Engadget

The RTX 5070 Ti has pleasantly surprised me in ways I hadn’t anticipated. While I expected it to slightly outperform the 4070 Ti Super, its multi-frame generation capability makes it significantly more adept at 4K gaming. Additionally, it proves to be quite future-proof, thanks to its 16GB of VRAM that matches the 5080.

Although the $549 RTX 5070 remains the most appealing option in NVIDIA’s new lineup, it’s refreshing to see a sensible choice available between that and the $1,000 5090. And yes, it still feels odd to refer to a $750 graphics card as “sensible.”

Source link

Sell anywhere with AliDropship