Intel UHD Graphics 615 vs Intel UHD Graphics 620 vs Intel Iris Plus Graphics 655
Intel UHD Graphics 615► remove from comparison
The Intel UHD Graphics 615 (GT2) is a low-end integrated graphics unit, which can be found in the Y-series of the Amber-Lake generation. This "GT2" version offers 24 Execution Units (EUs) clocked at up to 1000 MHz (depending on the CPU model). Due to its lack of dedicated graphics memory or eDRAM cache, the UHD 615 has to access the main memory (2x 64bit DDR3L/LPDDR3). Compared to the old HD Graphics 615 in Kaby-Lake-Y CPUs, the UHD 615 is the same GPU with slightly different clock speeds (depending on the model) and slight power/performance improvements due to the improved 14nm++ process .
The 3D performance of the UHD 615 depends on the CPU model and the cooling / TDP-setting of the laptop. Furthermore, the used main memory (single channel, DDR3(L), amount) is influencing performance. On average the UHD graphics 615 should be very similar to the Kaby-Lake HD Graphics 615 due to the same architecture and similar speeds. That means that only low demanding games like Farming Simulator 17 or Rocket League are playable in lowest detail settings.
The revised video engine in the HD 615 and UHD 615 now supports H.265/HEVC Main10 profile in hardware with 10 bit colors. Furthermore, Googles VP9 codec can also be hardware decoded. The UHD 615 should support HDCP 2.2 and therefore Netflix 4K. HDMI 2.0 however is still only supported with an external converter chip (LSPCon).
The UHD Graphics 615 can be found in Y processors with 5 W TDP (can be manipulated by the laptop manufacturers, and up 0,5 Watt from the previous generation) and is therefore suited for very thin, mostly passively cooled 2-in-1 notebooks and tablets.
Intel UHD Graphics 620► remove from comparison
The Intel UHD Graphics 620 (GT2) is an integrated graphics unit, which can be found in various ULV (Ultra Low Voltage) processors of the Kaby Lake Refresh generation (8th generation Core). Compared to the similar named Intel HD Graphics 620 in the 2016 Kaby Lake CPUs, only the name changed from HD to UHD. It is clocked with up to 1.150 MHz and therefore 100 MHz higher than the old HD Graphics 620 (max. 1.050 MHz) and offer a comparable performance. Due to its lack of dedicated graphics memory or eDRAM cache, the HD 620 has to access the main memory (2x 64bit DDR3/DDR4).
The exact performance of the UHD Graphics 620 depends on various factors like L3 cache size, memory configuration (DDR3/DDR4) and maximum clock rate of the specific model and as is similar to the Intel HD Graphics 620 and Intel HD Graphics 630. Compared to the average HD Graphics 620, the UHD graphics profits from the quad-core CPU a bit in our gaming benchmarks. Therefore, should be positioned between a GeForce 910M and 920M dedicated graphics card from Nvidia. For games, the drivers however are not that fast updated and tested as with GeForce graphics cards. In our tests, the UHD 620 is most suited for casual games like Team Fortress 2, Rocket League or Farming Simulator 17. See below for a list of modern games tested with various models.
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).
The UHD Graphics 620 can be found in mobile processors specified at 15 W TDP and is therefore suited for compact laptops and ultrabooks. The power consumption during playing The Witcher 3 (with an external monitor) was comparable to a Intel HD Graphics 620 based laptop and on the lowest end of all of our measurements. Dedicated GPUs, like the GeForce 940MX, needed a lot more power.
Intel Iris Plus Graphics 655► remove from comparison
The Intel Iris Plus Graphics 655 (GT3e) is a processor graphics card announced in September 2017. As the successor to the Intel Iris Graphics 650 (Kaby Lake), the Iris Plus Graphics 655 is used for 28-Watt Coffee Lake-U models. The biggest difference is the doubled eDRAM-cache at 128 MB. There should not be any big changes compared to the Iris Graphics 650.
The so called GT3e version of the Kaby Lake GPU still has 48 Execution Units (EUs), which can reach up to 1200 MHz (previously 1100 MHz) depending on the model. Besides eDRAM cache, the Iris 655 can also access the system memory (2x 64 Bit DDR3L-2133/DDR4-2400) via processor interface.
Compared to the Iris Plus 640 from the 15-Watt models, the Iris 655 only has a slightly higher maximum clock as well as almost twice the TDP, which allows a better utilization of the Turbo Boost potential.
The exact performance of the Iris Plus Graphics 655 depends on the CPU model, because maximum clock as well as the size of the L3 cache can differ a bit. The system memory (DDR3/DDR4) will influence the performance as well.
The fastest chips are high-clocked Core i7 models like the Core i7-8559U. Depending on the game, the Iris Plus 655 will probably be on par with a dedicated GeForce 930M or GeForce 940MX and can handle smooth gameplay in modern titles in low up to medium settings.
The reworked video engine now fully supports hardware decoding of H.265/HEVC videos. Contrary to Skylake, however, Kaby Lake can now also decode H.265/HEVC Main 10 with a 10-bit color depth as well as Google's VP9 codec. The video output is possible via DP 1.2/eDP 1.3 (up to 3840 x 2160 @60 Hz), whereas HDMI is also supported in the older 1.4a standard. An HDMI 2.0 output can be added via converter from DisplayPort. The GPU can drive up to three displays simultaneously.
The Iris Plus Graphics 655 is used for 28-Watt processor and therefore small/medium-sized notebooks.
|Intel UHD Graphics 615||Intel UHD Graphics 620||Intel Iris Plus Graphics 655|
|Gen. 9.5 Series|
|Codename||Kaby-Lake GT2||Kaby-Lake-Refresh GT2||Kaby Lake GT3e|
|Architecture||Gen. 9 Amber Lake||Gen. 9.5||Gen. 9.5 Kaby Lake|
|Pipelines||24 - unified||24 - unified||48 - unified|
|Core Speed||300 - 1000 (Boost) MHz||300 - 1150 (Boost) MHz||300 - 1200 (Boost) MHz|
|Memory Bus Width||64/128 Bit|
|API||DirectX 12_1, OpenGL 4.4||DirectX 12_1, OpenGL 4.4||DirectX 12_1, OpenGL 4.5|
|Power Consumption||5 Watt|
|technology||14 nm||14 nm||14 nm|
|Date of Announcement||30.08.2018||01.09.2017||01.09.2017|
|Theoretical Performance||883 TFLOPS FP16, 442 TFLOPS FP32, 110.4 TFLOPS FP64|
|Displays||HDMI 1.4, DisplayPort 1.2|
|Predecessor||HD Graphics 620|