NVIDIA GeForce GTX 1070 Ti (Desktop) vs Intel HD Graphics 6000 vs NVIDIA GeForce GTX 1070 SLI (Desktop)

NVIDIA GeForce GTX 1070 Ti (Desktop)

The Nvidia GeForce GTX 1070 Ti is the second fastest Pascal based graphics card in November 2017. It is a high-end desktop graphics card based on a cut down GP104 chip as the GTX 1080 and manufactured in 16 nm FinFET at TSMC. Compared to the GTX 1080, it uses the more available GDDR5 graphics memory, and compared to the GTX 1070 it offers more shaders (2432 versus 1920). More details on the Pascal architecture including all features and comparisons to Maxwell, can be found here.

The performance of the GTX 1070 Ti was in our tests right between the GTX 1070 and GTX 1080. However, the powerefficiency was worse than the GTX 1070 and near the GTX 1080. Nvidia also specifies the 1070 Ti and 1080 at 180W.

Intel HD Graphics 6000

The Intel HD Graphics 6000 (GT3) is an integrated Broadwell graphics card revealed in Q1 2015. It can be found in several ULV SoCs (15 W TDP) such as the Core i5-5250U or i7-5650U. The so-called GT3 GPU offers 48 EUs (Execution Units) and therefore somewhat more shader power than the previous HD 5000 (40 EUs). Depending on the specific CPU, the maximum GPU frequency varies between 950 and 1000 MHz.

Architecture and Features

Broadwell features a GPU based on the Intel Gen8 architecture, which has been optimized in various aspects compared to the previous Gen7.5 (Haswell). Inter alia, the shader arrays called "subslice" have been reorganized and now offer 8 Execution Units (EUs) each. Three subslices form a "slice" for a total of 24 EUs. Combined with other improvements such as larger L1 caches and an optimized frontend, the integrated GPU has become faster and more efficient than its predecessor.

The HD Graphics 6000 represents the top-end version of the Broadwell GPU family and consists of two slices with 48 EUs. Beyond that, there is also a low-end (GT1, 12 EUs), a mid-range (GT2, 24 EUs) and a high-end variant (GT3e, 48 EUs + eDRAM).

All Broadwell GPUs support OpenCL 2.0 and DirectX 12 (FL 11_1). The video engine can now decode H.265 using both fixed function hardware as well as available GPU shaders. Up to three displays can be connected via DP 1.2/eDP 1.3 (max. 3840 x 2160 @ 60 Hz) or HDMI 1.4a (max. 3840 x 2160 @ 24 Hz). HDMI 2.0, however, is not supported.

Performance

Depending on the specific CPU, the maximum GPU frequency varies between 950 and 1000 MHz. Due to the low TDP, however, the average clock in 3D applications will be significantly lower. Overall, the HD Graphics 6000 outperforms the previous HD 5000 by 20 - 25 percent and offers a performance slightly below a dedicated GeForce 820M.

Games as of 2014/2015 will usually run fluently only at (very) low settings.

Power Consumption

Utilizing a new 14 nm process, the Broadwell ULV chips are specified at just 15 W TDP and therefore suited for thin ultrabooks. The TDP is flexible and can be further reduced (9.5 W), which has a significant impact on performance.

NVIDIA GeForce GTX 1070 SLI (Desktop)

The Nvidia GeForce GTX 1070 SLI for desktops is a combination of two GTX 1070 graphics cards in SLI mode. Each card renders one frame at a time (AFR mode) but depends for most games on a good profile in the driver. Therefore, the performance can range from no gains over a single GTX 1070 to up to 90% faster performance. In most games with SLI support, the GTX 1070 SLI is therefore faster than a single GTX 1080. However, sometimes the combination also suffers from micro-stuttering.

The technical details of the card are identical to a single GTX 1070 (see for more information) and the power consumption is doubled (2x 150 = 300 Watt).