The Chinese company Innosilicon has introduced a graphics card called Fenhua N1, which does not seem to use the technologies of NVIDIA, AMD or Intel to move graphics. Are we facing a Chinese company with the ability to stand up to the big three or instead this hardware hides a secret and is not what it seems?
With the GPU market for PC gaming being one of the most profitable, it is normal for a superpower such as China to want to participate in it, since so far the big three are from the United States. To all this we must add the continuous boycotts and sanctions on China, so it is not surprising that graphics cards that are entirely Chinese are beginning to appear and therefore do not use chips from AMD, Intel or NVIDIA.
Although at the moment China’s interest in graphics processors is focused on their capabilities not for computing, but for the execution of parallel algorithms at high speed for the scientific and military world. Which are strategic for any nation in the world. So the country that Xi Jinping governs does not seem to be very interested in creating GPUs for gaming.
What is Innosilicon Fenhua N1?
Through a press release without much detail, the Innosilicon company has presented a graphics card that is not based on any GPU from Intel, AMD or NVIDIA but on a chip called Fenhua N1, which is designed to offer “high graphics performance in data centers ”and therefore we are talking about server hardware.
What’s so special about it? To start with the fact that it has video outputs HDMI 2.1 and eDP 1.4 And it doesn’t make sense on a compute-only card, since they can’t display graphics, which means that the Fenhua N1 can. Another thing that differs from HPC GPUs is the GDDR6X memory usage instead of the HBM2E and as you can see in the image it is a conventional graphics card that can also be connected to a fourth generation PCI Express x16 slot.
Although what confirms that we are facing a card capable of generating graphics is the fact that according to its press release it supports several APIs such as OpenGL, DirectX, OpenGL EX, Vulkan and OpenCL. So we would be facing the first graphics card for gaming, but this would be used at the beginning in servers, which is a way of talking about Cloud Gaming.
Its GPU, on the other hand, is of European origin
That is the big million dollar question, however, if we do a little searching we will find a joint press release between Innosilicon and Imagination Technologies from October 2020 where you can read the following:
Imagination Technologies has announced its cooperation with Innosilicon, a leading provider of mixed-signal intellectual properties and custom ASICs. Making use of the highly innovative Chiplet and GDDR6 technologies in Innosilicon they have integrated the newly launched BXT multicore GPU in a high performance for desktop and data center applications. The two companies are exploring a long-term strategy to bring even more powerful GPUs to market.
So the first GPU for gaming made in China does not use technology from that country, let’s not forget that Imagination is the owner and developer of PowerVR architectures, Tile Renderer-type graphics processors that have a long history in the world of graphics hardware. . The company was formerly known as Videologic and has a fairly important track record to its credit:
- The SEGA Dreamcast graphics processor.
- The Kyro family of graphics cards that competed against the first GeForce and Radeon in the early 2000s.
- Apple’s graphics hardware is based on Imagination technology, but little by little they changed the pieces until the British were left out.
- The graphics hardware of the PlayStation Vita.
Well, it seems that Imagination’s PowerVR technology has found a home, will PowerVR return to graphics cards for PC gaming?