NVIDIA, a renowned graphics card manufacturer, has begun production of the new GeForce RTX 4070. However, instead of using the specially designed AD104 graphics processor for this model, the company opted for the more powerful AD103 chip, previously exclusive to high-end versions of older graphics cards like the RTX 4070 Ti and RTX 4080.

For the RTX 4070, a portion of the chip's blocks are disabled to adjust performance levels, a decision that somehow reduced production costs.

RTX 4070Image: EnsiplayThis practice is not new in the industry, and similar graphics cards are most commonly found in Asian markets. One RTX 4070 buyer from MSI discovered that instead of the expected AD104, their card utilized a modified AD103. This was confirmed with a special program, which currently cannot correctly identify the new graphics card due to changes in its configuration.

Even with disabled blocks, the version of the card with the AD103 chip outperforms the one with the native AD104. For example, it boasts a larger cache memory and has significant overclocking potential.

It remains unclear how widespread (geographically) this practice is and which specific versions of the RTX 4070 will be equipped with modified AD103 chips. It's possible that this modification may not extend beyond the Asian market.

Main image: Ensiplay

Comment on this