Just a day ahead of their official launch, the test results of Nvidia’s GeForce RTX 4070, RTX 4060, and RTX 4050 on a mainstream laptop have been leaked – showing their lowered performance but higher efficiency than their predecessors.
This is the result of Nvidia using entry-level dies with limited core count and memory configurations for these cards – which produced fair improvements to their GeForce RTX 30 Series counterparts. Here are the results;
Performance Tests of GeForce RTX 40 Series
As we’re expecting the next generation of Nvidia’s GeForce cards to perform great, the company slightly disappoints us with the leaked test results. The chips – GeForce RTX 4070, RTX 4060, and RTX 4050 were tested against a Lenovo high-end laptop with a 13th Gen Raptor Lake-HX chip that has 150W TGP and a 205W TPP.
And for results, the new-gen GPUs are tested in various AAA and eSports titles at 1080p & 1440p resolutions, resulting as below;
Nvidia’s GeForce RTX 4050 offered a 37.3% performance improvement over the RTX 3050 Ti, while the RTX 4060 offered a 19.3% improvement over the RTX 3060, and surprisingly, the high-end chip – RTX 4070 performed 4% slower than the RTX 3070 Ti – by the averages in the 3DMark benchmark.
This was intriguing, and as expected, considering how Nvidia is cutting the costs of these new chips – while also overcharging them! What’s more intriguing was that both RTX 4070 and RTX 4060 capped out around 110W in terms of performance, while the RTX 4050 capped out around 90W – lower than their peak numbers as listed by Nvidia.
Well, the platform design they’re hosted on – a standard 29×29 package codenamed as “PCB2”, features a 5+1 phase design with 140W max power – which can also be used for RTX 4090 and RTX 4080 GPUs which feature a 40×40 SKU with a 7+2 phase power design. Though they’re conservative and cost-effective for Nvidia, they aren’t much excitement for the users, especially the high-end RTX 4070.