Chinese Fenghua GPU Aims for GeForce RTX 3060 Compute Performance
The Fenghua GPU is rather powerful, though it can't match high-end GPUs.
When Xindong/Innosilicon introduced its Fenghua 1 (Fantasy 1) discrete graphics processing unit (GPU) in mid-November, the company unveiled a rather impressive set of features for the GPU. This week, Innosilicon published performance specifications for the chip. While the Fenghua 1 cannot compete against the fastest offerings in our best graphics cards list, and likely lands quite a way down from the top of the GPU benchmarks hierarchy, it aims to provide comparable performance modern mid-range GPUs from AMD and Nvidia.
The Xindong/Innosilicon's Fenghua 1 (Fantasy 1) GPU is reportedly based on Imagination Technologies' PowerVR architecture. While we can only speculate about the exact microarchitecture used, the graphics processors are quite capable as they support contemporary application programming interfaces for graphics and compute, including DirectX, Vulkan, OpenGL, OpenCL, OpenGL ES, Caffe 1.0, TensorFlow 1.1.2, and ONNX. Since this is a PowerVR-based GPU, it comes with Android, Linux, and Windows software stacks. The GPU is made using a 12nm fabrication process.
There are two Fantasy 1 graphics cards at present: the single-chip Type A and the dual-chip Type B board. The single-chip Fantasy 1 boasts compute performance of around 5 FP32 TFLOPS for graphics and 25 INT8 TOPS for AI/ML, which is comparable to the (theoretical) performance of Nvidia's GeForce RTX 2060 GPU. Meanwhile, the dual-chip Fantasy 1 doubles that with FP32 compute performance of approximately 10 FP32 TFLOPS and 50 INT8 TOPS for AI/ML. This is slightly higher compared to performance numbers offered by Nvidia's GeForce RTX 3060.
Note that those figures represent theoretical compute performance, and as we've seen from AMD and Nvidia in the past, scaling via dual-GPU designs can double compute while creating a host of other hurdles when it comes to real-time graphics processing. Besides driver support, dual GPUs typically have to contain two copies of everything in memory — one copy for each GPU and its direct attached VRAM. Syncing data for frames between the GPUs further complicates things, often requiring game-specific support.
Xindong/Innosilicon's Fantasy 1 Graphics Cards
Row 0 - Cell 0 | Type A | Type B |
Number of GPUs | 1 | 2 |
FP32 Performance | 5 FP32 TFLOPS | 10 FP32 TFLOPS |
INT8 Performance | 25 TOPS | 50 TOPS |
Pixel Rate | 160 GPixel/s | 320 GPixel/s |
Video Decoding | 4x4Kp60, 16x1080p60, 32x720p30 | 8x4Kp60, 32x1080p60, 64x720p30 |
Number of users | 16 1080p users | 32 1080p users |
The Fantasy 1 Type A card can be equipped with 4GB, 8GB or 16GB of GDDR6 or GDDR6X memory, and the Fantasy 1 Type B card should double that with support for up to 32GB of memory. Both graphics cards support a PCIe Gen4 host interface as well as DisplayPort 1.4, eDP 1.4, and HDMI 2.1 interfaces.
While the raw compute performance of the Fantasy 1 Type B looks impressive, it likely won't directly translate over to games. The developer never mentions multi-GPU technologies akin to AMD's CrossFire or Nvidia's SLI — neither of which are properly supported on modern GPUs in contemporary games. It looks like the Fantasy 1 Type B is aimed mostly at datacenters. Keeping in mind that the GPU fully supports GPU virtualization as well as PCIe SR-IOV, GPU computing in datacenters and virtual desktop infrastructure (VDI) are among the applications it was designed for.
Perhaps the most surprising part about the Fantasy 1 GPU is power consumption. The typical power consumption of one Fantasy 1 Type A card in 'a multi-channel cloud environment' is supposedly about 50W, which is considerably lower than the TDP of Nvidia's GeForce RTX 2060 (TU106). Obviously, Xindong/Innosilicon's claims have to be independently tested, but they do sound impressive.
Xindong/Innosilicon is currently sampling its Fantasy 1 graphics cards with interested parties. However, it is unclear when these GPUs will be available commercially. Besides availability and performance, we would also need pricing details to determine whether they can compete with the likes of AMD Radeon, Nvidia GeForce, and Intel Arc — or more likely, against the datacenter variants of those GPUs.
Stay On the Cutting Edge: Get the Tom's Hardware Newsletter
Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.
Anton Shilov is a contributing writer at Tom’s Hardware. Over the past couple of decades, he has covered everything from CPUs and GPUs to supercomputers and from modern process technologies and latest fab tools to high-tech industry trends.
-
EyadSoftwareEngineer China, is really unstoppable, it is really impressive they achieved RTX3060 performance level!Reply -
gargoylenest that is a very good news. This is the kind of things that could bring down some MSRP levels to current card level considering that they are independent (I think...) from usual silicon foundries limitations.Reply -
VforV If the performance is real, then this is indeed good news.Reply
We need even the 4th player in GPU market, the more the merrier I'd say.
Those that don't like the fact that is a Chinese owned and Chinese made GPU, can simply not buy it, but 4 GPU manufacturers competing on prices... that would be amazing! All consumers will benefit, no matter which company one is a fan of, because of that competition. -
renz496 VforV said:If the performance is real, then this is indeed good news.
We need even the 4th player in GPU market, the more the merrier I'd say.
Those that don't like the fact that is a Chinese owned and Chinese made GPU, can simply not buy it, but 4 GPU manufacturers competing on prices... that would be amazing! All consumers will benefit, no matter which company one is a fan of, because of that competition.
sadly they will not going to live long. and consume only want competition so they can get cheap price. they did not care if the competing company will survive in the future or not. -
VforV
What makes you think they will fail?renz496 said:sadly they will not going to live long. and consume only want competition so they can get cheap price. they did not care if the competing company will survive in the future or not.
They have their own Chinese internal market if external one is not good for them anyway... and that market alone is huge.
Nah, they won't fail. They just need to prove and deliver that performance. I'm pretty sure their prices will not be high as we see with phones and cars, they all are lower than the other (western) brands. -
gargoylenest actually, I wouldnt expect such a big gap in pricing. You can compare to Xiaomi or Huawei flagship equipment's (phones, tv, laptop, etc.) high end tech still have a minimal price. But they should help a lot by bringing more availability and competitionReply -
JarredWaltonGPU Just a few things to consider.Reply
This is still on 12nm (e.g. Turing-era tech of 2018)
Supposedly the GPU only needs 50W
The compute is supposed to be nearly as high as a 3060 (for the dual-GPU 100W variant)
Dual-GPU means crappy scaling in gaming performance
Actually extracting 10 TFLOPS of usable FP32 performance might be far more difficult than on an AMD or Nvidia GPU
Because of the above, I'm betting actual performance of the 5 TFLOPS card will probably look a lot more like a GTX 1660 — at best. Probably more in the line of a GTX 1650, but perhaps with better performance in compute-only workloads. -
renz496
We have several big names competing in discrete gpu market before. Now there is only two left. And even between the two one of them just exist for the sake of being there even if they have competitive parts. Intel has been trying for several times. Good for intel they are the type that can burn billions, failed and yet still lives to see the day. Others like Qualcomm, Imagination and ARM can't do that. I heard Imagination try to make some sort of coming back when they licence their gpu to Innosilicon. They can built good gpu no doubt but from the way i see how they do things for years they hate to do the software aspect of it. That's why in the past only Apple end up being their real client since Apple develop their own drivers. Other SoC maker end up ditching Imagination and go with ARM instead because of this. This innosilicon gpu? The driver aspect most likely being handle completely by Innosilicon instead of Imagination. So they need to create real devrel with game developer if they want to market their gpu globally. And to do this they need money. To get money they need people to buy their gpu. Now we come to the most cruel aspect: the consumer itself. For most people they want the best for themselves. They will not going to keep buying bad product of a certain company so they can improve in the future. Want my money? Then earn it. That's why even in duopoly situation we have extremely unbalance market for years. 70/30 for almost a decade and now it was 80/20.VforV said:What makes you think they will fail?
They have their own Chinese internal market if external one is not good for them anyway... and that market alone is huge.
Nah, they won't fail. They just need to prove and deliver that performance. I'm pretty sure their prices will not be high as we see with phones and cars, they all are lower than the other (western) brands. -
VforV
All you said is valid, but... times are different now.renz496 said:We have several big names competing in discrete gpu market before. Now there is only two left. And even between the two one of them just exist for the sake of being there even if they have competitive parts. Intel has been trying for several times. Good for intel they are the type that can burn billions, failed and yet still lives to see the day. Others like Qualcomm, Imagination and ARM can't do that. I heard Imagination try to make some sort of coming back when they licence their gpu to Innosilicon. They can built good gpu no doubt but from the way i see how they do things for years they hate to do the software aspect of it. That's why in the past only Apple end up being their real client since Apple develop their own drivers. Other SoC maker end up ditching Imagination and go with ARM instead because of this. This innosilicon gpu? The driver aspect most likely being handle completely by Innosilicon instead of Imagination. So they need to create real devrel with game developer if they want to market their gpu globally. And to do this they need money. To get money they need people to buy their gpu. Now we come to the most cruel aspect: the consumer itself. For most people they want the best for themselves. They will not going to keep buying bad product of a certain company so they can improve in the future. Want my money? Then earn it. That's why even in duopoly situation we have extremely unbalance market for years. 70/30 for almost a decade and now it was 80/20.
They can simply sell in the internal Chinese market which is huge and also just to miners (which don't need gaming drivers) and be very successful until they get bigger and better to go international, if they want that.
That's one way of not failing in the current GPU world, not like 10-20 years ago when it was a very different world...