3Dwarrior
SOCIAL:
Filed under: Alchemist GPU, Alchemist Graphics Card, arc, ARC Alchemist GPU, ARC Graphics, ARC Graphics Card, Featured, graphics card, Hardware, Intel, Intel Alchemist, Intel Alchemist GPU, INtel AMD, Intel ARC, Intel ARC GPU, Intel ARC Graphics, Intel ARC Graphics Cards, Intel Graphics Card, Intel NVIDIA, Intel NVIDIA AMD, intel xe, Intel Xe GPU, Intel Xe HPG, Intel Xe-HPG GPU, News, Report, Sticky, Xe, Xe GPU, Xe Graphics Card, Xe HPG, Xe-HPG GPU

Intel’s Flagship ARC Graphics Card With Xe-HPG Alchemist GPU To…

Intel's Flagship ARC Graphics Card With Xe-HPG Alchemist GPU To Tackle AMD RX 6700 XT & NVIDIA RTX 3070

Intel ARC graphics playing cards based on the Alchemist Xe-HPG GPUs are all established for launch subsequent 12 months and primarily based on the technical specs, we could be wanting at quite competitive functionality figures in opposition to AMD and NVIDIA GPUs.

Intel’s Flagship ARC Graphics Playing cards With Xe-HPG Alchemist GPU To Be Extremely Aggressive Towards NVIDIA GA104 & AMD Navi 22

The initially Intel ARC graphics playing cards will be driven by the Alchemist GPUs based mostly on the Xe-HPG architecture. Intel has so considerably confirmed that the initial discrete graphics cards will hit retail by Q1 2022 and will be centered on the TSMC 6nm approach node. Intel also specific the specs of Alchemist GPUs and the main making blocks which involve the Xe-Main.

Intel ARC Xe-HPG Alchemist GPU – The Creating Blocks

So rounding up what we realized, the Intel Xe-HPG Alchemist GPU features a Xe-Core which is the elementary DNA of the 1st Gen ARC lineup. The Xe-Main is a compute block that is composed of sixteen Vector Engines (256-bit per motor) and sixteen Matrix Engines (1024-little bit for every motor). Each Vector Engine is composed of 8 ALUs so, in full, we are wanting at 128 ALUs for every Xe-Main. Every single Matrix Engine block is also referred to as an XMX block which will tackle tensor functions in both FP16 and INT8 modes. The Xe-Main more attributes its own focused L1 cache.

Intel fuses four Xe-Cores alongside one another to kind a Render Slice which is composed of four Ray Tracing Models, 4 Sampler Units, Geometry/Rasterize/HiZ engines, and two Pixel Backend blocks with eight units on every. These Render Slices are set with each other to kind the primary GPUs. The flagship is composed of an 8 Render Slice configuration which functions 32 Xe-Cores, 512 Vector Engines, and 4096 ALUs. There will be unique configurations with two, four, six Render Slices but we are concentrating on the flagship part in this report.

Intel ARC Alchemist vs NVIDIA GA104 & AMD Navi 22 GPUs

GPU Title Alchemist DG-512 NVIDIA GA104 AMD Navi 22
Architecture Xe-HPG Ampere RDNA 2
Procedure Node TSMC 6nm Samsung 8nm TSMC 7nm
Flagship Product ARC (TBA) GeForce RTX 3070 Ti Radeon RX 6700 XT
Raster Engine 8 6 two
FP32 Cores 32 Xe Cores forty eight SM Models forty Compute Models
FP32 Models 4096 6144 2560
FP32 Compute ~16 TFLOPs 21.7 TFLOPs twelve.4 TFLOPs
TMUs 256 192 one hundred sixty
ROPs 128 96 sixty four
RT Cores 32 RT Units forty eight RT Cores (V2) 40 RA Units
Tensor Cores 512 XMX Cores 192 Tensor Cores (V3) N/A
Tensor Compute ~131 TFLOPs FP16
~262 TOPs INT8
87 TFLOPs FP16
174 TOPs INT8
25 TFLOPs FP16
fifty TOPs INT8
L2 Cache TBA 4 MB three MB
Extra Cache 16 MB Intelligent Cache? N/A ninety six MB Infinity Cache
Memory Bus 256-bit 256-bit 192-bit
Memory Potential sixteen GB GDDR6 8 GB GDDR6X 16 GB GDDR6
Launch Q1 2022 Q2 2021 Q1 2021

Intel ARC Xe-HPG Alchemist GPU – Comparing It To NVIDIA’s GA104 & AMD’s Navi 22

A rundown of the requirements and comparison has been designed by 3DCenter which offers us an thought of the theoretical performance that Intel’s new GPU will have to supply. So right off the bat, Intel’s ARC Xe-HPG Alchemist flagship will offer a lot more TMUs and ROPs than the NVIDIA and AMD competitors. The main count at 4096 is greater than AMD’s Navi 22, Navi 21 (RX 6800) but decrease when compared to NVIDIA’s GA104. NVIDIA is utilizing a dual FP32 numbering methodology and really should theoretically be 3072.

Intel’s ARC Alchemist GPUs have reduced ray tracing models than the opposition but we will not know just how their Ray tracing implementation is effective. For illustration, though Navi 22 features a lot more RT cores than the GA106 Ampere GPUs, the hardware-degree integration in just NVIDIA’s RT cores is superior in all regards to AMD’s implementation. So the remaining performance would rely upon Intel’s components-stage integration and application-stage optimization for ray tracing apps.

A major lead that Intel could have above the competitors, particularly NVIDIA given that AMD lacks in this section, is AI support in supersampling technologies. Intel has now showcased an outstanding demo of its XeSS technology and dependent on the anticipated numbers, Intel GPUs could outperform NVIDIA’s Tensor Main implementation (DLSS) with its XMX architecture. Intel is also expected to element a little but handy video game cache on its GPUs and will be outfitted with better VRAM capacities of up to sixteen GB (GDDR6) throughout a 256-bit bus interface. This would be two times as substantially memory as NVIDIA’s RTX 3070 and RTX 3070 Ti so they may have to prepare a refresh to counter it.

Finally, the theoretical FP32 compute overall performance is computed with an envisioned peak clock level of two GHz. That is the most probably state of affairs for TSMC’s 6nm method node offered how nicely clocks scale on TSMC’s 7nm system node. Based on that, the Intel Xe-HPG Alchemist GPU could offer you all around sixteen-17 TFLOPs of Compute power. This is somewhat lessen FLOPs than what NVIDIA’s GA104 produces but it ought to be famous that not all FLOPs should be measured equally as gaming architecture operates very various as opposed to datacenter chips.

Primarily based on these early requirements, we are seeking at an Intel graphics card that could stop up becoming quicker than AMD’s Radeon RX 6700 XT and NVIDIA’s RTX 3070 with relieve. To thrust its 1st Gen graphics playing cards additional into the client segment, Intel could very likely supply incredibly aggressive costs against recognized giants like AMD and NVIDIA. And alongside with a solid suite of software program-level optimizations, they could have a earn-win in their arms which will only be pushed forward with long run generations of ARC GPUs.

Intel ARC Alchemist vs NVIDIA GA104 & AMD Navi 22 GPUs

GPU Title Alchemist DG-512 NVIDIA GA104 AMD Navi 22
Architecture Xe-HPG Ampere RDNA two
Course of action Node TSMC 6nm Samsung 8nm TSMC 7nm
Flagship Product ARC (TBA) GeForce RTX 3070 Ti Radeon RX 6700 XT
Raster Motor 8 six two
FP32 Cores 32 Xe Cores forty eight SM Units forty Compute Models
FP32 Units 4096 6144 2560
FP32 Compute ~16 TFLOPs 21.seven TFLOPs 12.four TFLOPs
TMUs 256 192 one hundred sixty
ROPs 128 96 64
RT Cores 32 RT Units forty eight RT Cores (V2) 40 RA Units
Tensor Cores 512 XMX Cores 192 Tensor Cores (V3) N/A
Tensor Compute ~131 TFLOPs FP16
~262 TOPs INT8
87 TFLOPs FP16
174 TOPs INT8
25 TFLOPs FP16
50 TOPs INT8
L2 Cache TBA 4 MB 3 MB
Supplemental Cache sixteen MB Good Cache? N/A 96 MB Infinity Cache
Memory Bus 256-bit 256-little bit 192-little bit
Memory Capacity sixteen GB GDDR6 eight GB GDDR6X 16 GB GDDR6
Start Q1 2022 Q2 2021 Q1 2021

The put up Intel’s Flagship ARC Graphics Card With Xe-HPG Alchemist GPU To Deal with AMD RX 6700 XT & NVIDIA RTX 3070 by Hassan Mujtaba appeared to start with on Wccftech.