Intel HD Graphics 620 vs AMD Radeon RX Vega 11 vs AMD Radeon RX Vega M GL / 870

Intel HD Graphics 620

The Intel HD Graphics 620 (GT2) is an integrated graphics unit, which can be found in various ULV (Ultra Low Voltage) processors of the Kaby Lake generation. This "GT2" version of the Kaby Lake GPU is clocked at up to 1050 MHz (depending on the CPU model). Due to its lack of dedicated graphics memory or eDRAM cache, the HD 620 has to access the main memory (2x 64bit DDR3/DDR4).

Performance

The exact performance of the HD Graphics 620 depends on various factors like L3 cache size, memory configuration (DDR3/DDR4) and maximum clock rate of the specific model but is very similar to the old HD Graphics 520 GPU. That means only some less demanding recent games like Farming Simulator 17 run fluently in low detail settings (see below for gaming benchmarks). Compared to dedicated graphics cards, the HD620 should be positioned between a GeForce 910M and 920M from Nvidia. For games, the drivers however are not that fast updated and tested as with GeForce graphics cards.

Features

The revised video engine now supports H.265/HEVC Main10 profile in hardware with 10 bit colors. Furthermore, Googles VP9 codec can also be hardware decoded. The first models do not support HDCP 2.2 and therefore Netflix 4K. This was added with the new models in 2017 and beginning of May 2017 also the older CPU models like the i3-7100U are now shipped with a new S-Spec Code that supports HDCP 2.2. HDMI 2.0 however is still only supported with an external converter chip (LSPCon).

Power Consumption

The HD Graphics 620 can be found in mobile processors specified at 15 W TDP and is therefore suited for compact laptops and ultrabooks.

AMD Radeon RX Vega 11

The AMD Radeon RX Vega 11 is an integrated GPU in the Ryzen desktop APUs (e.g. Ryzen 5 2400G) that were launched early 2018. The graphics card is based on the Vega architecture (5th generation GCN) and has all 11 CUs (=704 shaders) clocked at up to 1240 MHz (Ryzen 5 2400G). The GPU benefits from fast dual-channel DDR4-2400 RAM (contrary to DDR4-2133 single-channel, which is also possible).

The Vega architecture offers some improvements over the Polaris generation and now supports DirectX 12 Feature Level 12_1. More information is available in our dedicated article about the Raven Ridge architecture.

According to a first 3DMark Fire Strike that was published by AMD, the Vega 11 offers a similiar theoretical performance as the older dedicated Radeon R9 M280X. In games the missing dedicated graphics memory may however prove to be a bottleneck. Therefore, modern and demanding games will only be playable in lowest settings and the APU is mostly intended for lower demanding games like Overwatch, Dota 2, ...

Thanks to the 14nm process and clever power-saving features, the power consumption is comparatively low. The high TDP of 65 Watt of the APU however should lead to a much higher power consumption compared to the mobile APUs with 15 Watt TDP.

AMD Radeon RX Vega M GL / 870

The AMD Radeon RX Vega M GL or Radeon RX Vega 870 (according to Dell) is an integrated GPU in the slower Intel Kaby-Lake-G SoC. It combines a Kaby-Lake processor, a Vega graphics card and 4 GB HBM2 memory on a single package. The graphics card offers 20 CUs (1280 shaders) and is clocked from 931 - 1011 MHz.

Currently it looks like the GPU is a mixture of Polaris (Shaders?) and Vega (HBM memory controller at least). E.g. the graphics card is called Polaris 22 internally. Detailed information on the Vega architecture can be found in our dedicated article about the Raven Ridge architecture.

The performance of the Vega M GL should be between the Nvidia GeForce GTX 1050 and 1050 Ti according to benchmarks from Intel. Therefore, it should be well suited for 1080p and high resolution gaming.

The power consumption should be quite similar to a dedicated Vega Mobile GPU, but due to the small form factor, smaller laptops can be built with the Kaby-Lake-G SoC. The TDP is specified at 65 Watt for the whole package including GPU, CPU and HBM2 memory.