The much-awaited entry of Intel in the discrete graphics segment is getting closer and closer as the company gets ready to launch its first Arc Alchemist GPUs for gamers, based on the brand new Xe-HPG architecture.
Intel Arc Alchemist GPU: First Major Discrete Graphics Architecture For All Kinds of Gamers
[Updated – 17/07/22]
When we talk about GPUs, only two names come to mind, AMD or NVIDIA. Both companies have been offering their latest GPUs for a diverse range of markets. These include desktops, mobility, workstations, and servers. Intel hasn’t had a profound presence within the graphics segment yet, they have the highest GPU share of all companies. Why’s that the case? Well, Intel has had GPUs, but only integrated ones that are found on almost all of their CPUs. Whether consumers use those GPUs is a whole different thing.
Intel is now aiming to change that, in fact, for the last couple of years and ever since the formation of their Visual Technologies Group which is headed by Raja Koduri, the company is set to release a brand new graphics architecture that is all set to enter the GPU world in 2022. Meet Arc, the brand new chapter in Intel’s graphics vision!
Intel Arc Alchemist GPU – So What Are We Going To Call Them?
So we know the brand name from Intel is Arc and the first graphics lineup is going to be called Alchemist and they are going to feature the Xe-HPG (High-Performance Graphics) graphics architecture. Now the Alchemist name sounds a little too geeky but it’s also a nice touch compared to all the Scientists & Star codenames that NVIDIA and AMD have been using. Also, Intel doesn’t stop at Alchemist and has even more codenames detailed that they plan on launching in the future. Keep on reading to know more about that.
To make things simple, this is how each company brands its current graphics lineup [Company][Branding][Identifer][Family]:
- Intel Arc A-Series
- NVIDIA GeForce 30
- AMD Radeon 6000
But that’s not all, we also have come to know what the products will be called. Intel is confirming today at least 5 products based on the Discrete Arc A-Series GPUs, all of which are aimed at the mobility segment first. So we know the nomenclature of these graphics chips will be something like the following:
- Intel Arc A300, A500, and A700 (performance-wise)
So the full branding would be something like (Note: following products are just used for naming comparisons and may not fall in the same performance category):
- Intel Arc A770M
- NVIDIA GeForce RTX 3080
- AMD Radeon RX 6800
So now that we know how Intel is going to call their Arc Alchemist graphics products, let’s move over to the specifications.
Intel Arc Alchemist GPU Architecture
The Intel ARC Alchemist GPUs are designed primarily for desktop and notebook platforms. From the information that we have currently gathered, the lineup will consist of two GPUs and each of those will have various SKUs, and each of those SKUs will be featured in a range of solutions for desktop graphics cards and mobility GPUs.
All Intel ARC Alchemist GPUs will utilize the TSMC 6nm process node and from what we already know, the Intel Xe-HPG Alchemist GPU features a Xe-Core which is the fundamental DNA of the 1st Gen ARC lineup. The Xe-Core is a compute block that is composed of 16 Vector Engines (256-bit per engine) and 16 Matrix Engines (1024-bit per engine).
Each Vector Engine is composed of 8 ALUs so, in total, we are looking at 128 ALUs per Xe-Core. The Xe-Core further features its own dedicated 192 KB L1 cache so that is a total of 6 MB of L1 cache on the complete chip.
2 of 9
Intel fuses four Xe-Cores together to form a Render Slice which is composed of 4 Ray Tracing Units, four Sampler Units, Geometry/Rasterize/HiZ engines, and two Pixel Backend blocks with 8 units on each. These Render Slices are put together to form the main GPUs. The flagship is composed of an 8 Render Slice configuration which features 32 Xe-Cores, 512 Vector Engines, and 4096 ALUs. There will be different configurations with 2, 4, 6 Render Slices but we are focusing on the flagship part in this report.
Intel’s Xe HPG architecture will be able to achieve 1.5x higher clock rates than Xe LP and also deliver 1.5x higher performance per watt. This means we are looking at clocks in the 2.1 GHz range considering the Xe LP discrete GPUs were clocked at 1.4 GHz. It also means that Intel will be able to squeeze additional power out of the architecture, should they want to, without increasing the power draw (or reducing the power draw while keeping performance constant.
2 of 9
Out of the two GPUs, the first one will be known as ACM-G10. This is the top SKU and will be featured in mainstream and high-end designs. The second GPU is known as ACM-G11. This is an entry-level SKU and as such, will be featured in entry-level and mainstream designs. Each GPU has its own SKUs with different specifications.
2 of 9
Each Matrix Engine block is also referred to as an XMX block which will handle tensor operations in both FP16 (128 FP16 ops/clock), INT8 (256 INT8 ops/clock), & INT4 (512 INT4/INT2 ops/clocks) mode. The XMX engines allow for up to 16 times the compute capabilities for INT8 inferencing. This helps boost the performance through the dedicated XeSS (Xe Super Sampling) technology.
Coming to the media engine, the Intel Arc Alchemist discrete GPUs are equipped with the latest tech which includes up to 8K60 12-bit HDR decode, up to 8k 10-bit HDR encode, VP9, AVC, HEVC, and AV1. Intel is partnering with top brands to leverage the AV1 capabilities of their Arc GPUs through the Alliance for Open Media. The new AV1 engine will provide 50x faster GPU encoding capabilities versus software mode. The tech will be integrated into several creation tools such as FFMPEG, Handbrake, Adobe Premiere Pro, DaVinci Resolve, and XSplit on launch.
2 of 9
In addition to that, the Intel Arc Xe Display Engine will support the latest display standards such as HDMI 2.0b, DisplayPort 1.4a, and 2.0 10G. Support for a variety of high resolution and high refresh rate modes is also going to be available which include:
- 2x 8k60 HDR
- 4x 4k120 HDR
- 1080p360
- 1440p360
2 of 9
As for adaptive syncing technology, Intel’s discrete Arc Alchemist GPUs will support both Adaptive-Sync and Speed Sync technology. The former will sync the display’s refresh rate to provide a smooth and tear-free experience while the latter will speed up the latest frame, delivering low latency, no tearing, and no cap. There’s also a third mode known as Smooth Sync which will blur distracting screen tears with a dithering filter.
2 of 9
Intel Xe-HPG Based Discrete Alchemist GPU Configurations:
| GPU Variant | Graphics Card Variant | GPU Die | Execution Units | Shading Units (Cores) | Memory Capacity | Memory Speed | Memory Bus | TGP |
|---|---|---|---|---|---|---|---|---|
| Xe-HPG 512EU | ARC A780? | Arc ACM-G10 | 512 EUs | 4096 | Up To 32/16 GB GDDR6 | 18 / 16 / 14 Gbps | 256-bit | ~225W (Desktops) 120-150W (Laptops) |
| Xe-HPG 384EU | ARC A580? | Arc ACM-G10 | 384 EUs | 3072 | Up To 12 GB GDDR6 | 16 / 14 Gbps | 192-bit | 150-200W (Desktops) 80-120W (Laptops) |
| Xe-HPG 256EU | ARC A550? | Arc ACM-G10 | 256 EUs | 2048 | Up To 8 GB GDDR6 | 16 / 14 Gbps | 128-bit | 60-80W (Laptops) |
| Xe-HPG 128EU | ARC A380? | Arc ACM-G11 | 128 EUs | 1024 | Up To 6 GB GDDR6 | 16 / 14 Gbps | 96-bit | ~75W (Desktops) |
| Xe-HPG 128EU | ARC A350? | Arc ACM-G11 | 128 EUs | 1024 | Up To 4 GB GDDR6 | 16 / 14 Gbps | 64-bit | 35-50W (Laptops) |
| Xe-HPG 96EU | ARC A330? | Arc ACM-G11 | 86 EUs | 768 | Up To 4 GB GDDR6 | 16 / 14 Gbps | 64-bit | ~35W (Laptops) |
Intel Arc Alchemist ‘ACM-G10’ GPU Specifications – The Top GPU
The top Arc Alchemist GPU is going to be the ACM-G10. It is said expected to measure 407mm2 which makes it larger than both NVIDIA’s GA014 and AMD’s Navi 22 chips. The ACM-G10-powered GPUs are going to compete against NVIDIA’s GeForce RTX 3070(Ti) and AMD Radeon RX 6700 XT.
NVIDIA packs in tensor cores and much bigger RT/FP32 cores in its chips while AMD RDNA 2 chips pack a single ray accelerator unit per CU and Infinity Cache. Intel will also have dedicated hardware onboard its Alchemist GPUs for Raytracing & AI-assisted super-sampling tech. The full die features 32 Xe Cores, 512 EUs, 4096 ALUs, a 256-bit bus interface, & up to 16 GB GDDR6 memory featuring clock speeds between 16-18 Gbps from Samsung.
The Xe-HPG Alchemist ACM-G10 chip is suggested to feature clocks of around 2.2 – 2.5 GHz though we don’t know if these are the average clocks or the maximum boost clocks. We have already spotted a flagship configuration with DG2-512 GPU running at 2.4 GHz, delivering nearly 20 TFLOPs of FP32 horsepower so we might very well end up with 2.5 GHz clocks in the final revision.
The TDP target for these chips seems to start at 120W for laptops and go all the way up to 300W for desktop parts. In either case, we can expect the final model to rock an 8+6 pin connector config, The reference model is also going to look very much like the drone marketing shot Intel put out during the ARC branding reveal.
Intel ARC ACM-G10 vs NVIDIA GA104 & AMD Navi 22 GPUs
| Graphics Card Name | Intel ARC A770 | NVIDIA GeForce RTX 3070 Ti | AMD Radeon RX 6700 XT |
|---|---|---|---|
| GPU Name | ACM-G10 | NVIDIA GA104 | AMD Navi 22 |
| Architecture | Xe-HPG | Ampere | RDNA 2 |
| Process Node | TSMC 6nm | Samsung 8nm | TSMC 7nm |
| Die Size | 406mm2 | 392mm2 | 335mm2 |
| Transistors | 21.7B | 17.4B | 17.2B |
| Transistor Density | 53.4M Transistors/mm2 | 44.4M Transistors/mm2 | 51.2M Transistors/mm2 |
| FP32 Cores | 32 Xe Cores | 48 SM Units | 40 Compute Units |
| FP32 Units | 4096 | 6144 | 2560 |
| Max Clock | ~2400 MHz | 1770 MHz | 2581 MHz |
| FP32 TFLOPs | ~20 TFLOPs | 21.75 TFLOPs | 13.21 TFLOPs |
| Memory Bus | 256-bit | 256-bit | 192-bit |
| Memory Capacity | 16 GB GDDR6 | 8 GB GDDR6X | 12 GB GDDR6 |
| Launch | Q3 2022 | Q2 2021 | Q1 2021 |
Intel Arc Alchemist ‘ACM-G11’ GPU Specifications – The Small GPU
The Intel ACM-G11 is going to be the smaller Arc GPU, aiming at the entry-level and mainstream PC platforms. The GPU is said to measure around 156mm2. This is also much smaller than the 200mm2 TU117 die that the chip will be competing against. The GA107 die size isn’t known yet but it is likely to be around 160-180mm2. It is a bigger chip compared to the recently released AMD Navi 24 which measures only 107mm2.
There are two configs that feature the full-fat SKU with 1024 cores, a 96-bit, and a 64-bit variant with 6 GB and 4 GB memory capacity, respectively. The cut-down variant will come with 96 EUs or 768 cores and a 4 GB GDDR6 memory featured across a 64-bit bus interface. The chip is expected to feature a clock speed of around 2.2 – 2.5 GHz and have a sub 75 Watt power consumption which means we will be looking at connector-less graphics cards for the entry-level segment.
This GPU will be very similar to the DG1 GPU-based discrete SDV board however Alchemist will have a more improved architecture design and definitely more performance uplift over the first-gen Xe GPU architecture. This lineup is definitely going to be aimed at the entry-level desktop discrete market based on the specifications.
Intel ARC ACM-G11 vs NVIDIA GA106 & AMD Navi 24 GPUs
| Graphics Card Name | Intel ARC A350? | NVIDIA GeForce RTX 2050 | AMD Radeon RX 6500 XT | Intel ARC A380? | NVIDIA GeForce RTX 3050 |
|---|---|---|---|---|---|
| GPU Name | ACM-G11 | NVIDIA GA107 | AMD Navi 24 | ACM-G11 | NVIDIA GA106 |
| Architecture | Xe-HPG | Ampere | RDNA 2 | Xe-HPG | Ampere |
| Process Node | TSMC 6nm | Samsung 8nm | TSMC 6nm | TSMC 6nm | Samsung 8nm |
| Die Size | 157mm2 | TBC | 107mm2 | 157mm2 | 276mm2 |
| FP32 Cores | 8 Xe Cores | 16 SM Units | 16 Compute Units | 8 Xe Cores | 24 SM Units |
| FP32 Units | 1024 | 2048 | 1024 | 1024 | 3072 |
| Memory Bus | 64-bit | 64-bit | 64-bit | 96-bit | 128-bit |
| Memory Capacity | 4 GB GDDR6 | 4 GB GDDR6 | 4 GB GDDR6 | 6 GB GDDR6 | 8 GB GDDR6 |
| TDP | 35-50W | 30-45W | ~75W | ~75W | ~75W |
| Launch | Q2 2022 | Q1 2022 | Q2 2022 | Q2 2022 | Q1 2022 |
Intel Arc A-Series Desktop Graphics Lineup
Intel’s Arc A-Series Desktop lineup is expected to include as many as 6 graphics cards based on the ACM-G10 and ACM-G11 Alchemist GPUs.
Intel Arc 7 ‘High-Performance’ Gaming Graphics Card Lineup
The Arc Alchemist lineup will include the flagship Arc A770 which will feature the full Arc ACM-G10 GPU with 32 Xe-Cores and a 256-bit bus interface. The Intel Arc A770 will feature both 16 GB and 8 GB flavors across a 256-bit bus interface and a TDP of 225W. It is expected to be positioned in the same performance category as the RTX 3060 Ti but will offer slightly better performance and even compete with the RTX 3070 in certain titles. The graphics card is expected to cost between the $349 to $399 US range.
The second part is the Intel Arc A750 which will also be equipped with an ACM-G10 GPU but house 24 Xe Cores (3072 ALUs), 24 ray tracing units 8 GB GDDR6 memory running across a 256 -bit bus interface, and a TDP target of 225W, same as the Arc A770. This GPU will aim for the GeForce RTX 3060 series mobility options. While Intel hasn’t shown the official performance of the flagship part yet, they have shown the Arc A750 graphics card to be up to 17% faster than the GeForce RTX 3060 in modern titles and will be doing so in the $279-$329 US range.
Both the Intel Arc A770 and Arc A750 graphics cards will come in Limited Edition flavors and also custom designs which will be available globally. The Arc A770 will be as high as the Alchemist line will go and if you were looking for more enthusiast variants, then you’d have to wait for the next-generation ‘Battlemage’ lineup.
Intel Arc 5 ‘Advanced’ Gaming Graphics Card Lineup
The Intel Arc 5 lineup is expected to include just one variant, for now, the Arc A550. The graphics card is expected to feature 16 Xe-Cores (2048 ALUs) and will also feature 8 GB of GDDR6 memory across a 128-bit wide bus interface. The graphics card is expected to compete against the RTX 3050 and will be aiming at the $200-$299 US segment with a TDP of 175W. It is likely that this variant will be one of the best sellers if it can be priced under the $250 US bracket and close to $200 US since that will put it close to the RX 6500 XT while offering better performance, and a finer feature set like AV1, XeSS, better raytracing capabilities to name a few.
Intel Arc 3 ‘Enhanced’ Gaming Graphics Card Lineup
As for the entry-level lineup, Intel is expected to have two offerings, the Arc A380, and Arc A310. The models are likely to feature 8 and 4 Xe-Cores with the top variant rocking 6 GB GDDR6 (96-bit) memory and the entry-level model featuring 4 GB GDDR6 memory (64-bit). The entry-level lineup will compete against NVIDIA’s GeForce GTX 1650 and GTX 1050(Ti) offerings. In addition, to the gaming variants, Intel will also have two Pro variants, the Arc Pro A50 and Arc Pro A40. The first could utilize the ACM-G10 GPU while the latter could utilize the ACM-G11 GPU.
Pricing of the Intel Arc A380 is already confirmed at $129-$139 US while the Arc A310 will enter the sub-$100 US market segment.
Intel Arc A-Series Desktop Graphics Card Lineup ‘Rumored’:
| Graphics Card Variant | GPU Variant | GPU Die | Execution Units | Shading Units (Cores) | Memory Capacity | Memory Speed | Memory Bus | TGP | Price |
|---|---|---|---|---|---|---|---|---|---|
| Arc A770 | Xe-HPG 512EU (TBD) | Arc ACM-G10 | 512 EUs (TBD) | 4096 (TBD) | 16 GB GDDR6 | 16 Gbps | 256-bit | 225W | $349-$399 US |
| Arc A770 | Xe-HPG 512EU (TBD) | Arc ACM-G10 | 512 EUs (TBD) | 4096 (TBD) | 8 GB GDDR6 | 16 Gbps | 256-bit | 225W | $349-$399 US |
| Arc A750 | Xe-HPG 384EU (TBD) | Arc ACM-G10 | 384 EUs (TBD) | 3072 (TBD) | 8 GB GDDR6 | 16 Gbps | 256-bit | 225W | $299-$349 US |
| Arc A580 | Xe-HPG 256EU (TBD) | Arc ACM-G10 | 256 EUs (TBD) | 2048 (TBD) | 8 GB GDDR6 | 16 Gbps | 128-bit | 175W | $200-$299 US |
| Arc A380 | Xe-HPG 128EU (TBD) | Arc ACM-G11 | 128 EUs | 1024 | 6 GB GDDR6 | 15.5 Gbps | 96-bit | 75W | $129-$139 US |
| Arc A310 | Xe-HPG 64 (TBD) | Arc ACM-G11 | 64 EUs (TBD) | 512 (TBD) | 4 GB GDDR6 | 16 Gbps | 64-bit | 75W | $59-$99 US |
Intel Arc A-Series Entry-Level Desktop Graphics Performance
So far, Intel has showcased the official performance numbers of two graphics cards, the Arc A750 and the Arc A380. The Intel Arc A380 seems to be competing against the AMD Radeon RX 6400 as it trades blows with it while the Arc A750 offers up to 17% better performance than the NVIDIA RTX 3060 graphics card.
Intel has started to share performance numbers and information regarding its high-end Arc Alchemist GPUs too. The first of these designs is the Intel Arc A750 Limited Edition which was also teased back at IEM 2022. The performance benchmarks shared by Intel include five games which include F1 2021, Cyberpunk 2077, Control, Borderlands 3, and Fortnite. All of these games were tested at 1440p resolution using the high preset and the Arc A750 graphics card was compared against NVIDIA’s GeForce RTX 3060 which it beat by delivering up to 17% performance.
Intel Arc A750 Limited Edition Performance Benchmarks in Games:
Intel Arc A750 Limited Edition Performance Demo In Cyberpunk 2077:
As you can see in the disclaimers below, both graphics cards were tested on the same system configurations comprising an Intel Core i9-12900K CPU and the latest Windows 11 version. In terms of drivers, the RTX 3060 used the 516.40 version while the Arc A750 used an engineering driver. You can also see a more detailed performance breakdown (FPS) in the slide below:
Intel Arc A750 Limited Edition Performance Disclaimers & Configurations Used:
All of the testing in the official benchmarks was done using an Intel Core i5 12600k with 32 GB of 3200 MHz DDR4 RAM and Windows 11 OS and a 4TB NVME SSD. Only the GPUs, ie the GTX 1650, RX 6400, and Intel Arc A380 were swapped between them. Testing was conducted almost a month back so it’s worth noting that driver performance would almost certainly have increased during this time:
As we can see, the Intel Arc A380 trades blows with the AMD RX 6400 and (less occasionally) with the NVIDIA GTX 1650. It actually beats the RX 6400 in Total War: Troy, Naraka Bladepoint, The Witcher 3, and F1 2021. Considering this is the official documentation, it’s actually pretty cool that Intel did not present a one-sided story about its upcoming GPU. Here is also where the story gets really interesting. FineWine
is a term that AMD fans and readers of this site would be very familiar with and was a popular term to describe AMD’s ongoing post-launch driver development back in the days when it used to be cash-strapped and was the underdog.
Compute workloads are once again a mixed bag when it comes to Intel Arc A380. It beats out both the GTX 1650 and RX 6400 handily in HandBrake and is slightly worse than the GTX 1650 in DaVinci Resolve.
Intel Arc A-Series Mobile Graphics Lineup
The Mobility lineup was the first to launch in the consumer segment with five new discrete chips within three different tiers. The Intel Arc 7 and Arc 5 lineup will utilize the ACM-G10 GPU while the Arc 3 lineup will utilize the ACM-G11 GPU. All of these GPUs are equipped with the same feature set which includes support for XeSS, DirectX XII Ultimate, XMX acceleration, Xe Media Engine, and PCIe Express 4.0 along with a host of other features.
Intel Arc 7 ‘High-Performance’ Gaming GPU Lineup
The Intel Arc 7 lineup is going to utilize the flagship ACM-G10 GPU and will feature two variants, the Arc A770M and the Arc A730M. The top-end variant for mobility platforms, the Arc A770M, will be equipped with the full ACM-G10 configuration, utilizing 32 Xe-Cores for 4096 ALUs, 32 ray tracing units, a graphics clock of 1650 MHz, up to 16 GB GDDR6 memory that operates across a 256-bit wide bus interface and a TDP target of 120-150W. This is comparable to the GeForce RTX 3070 Ti Max-Q variant.
The second part is the Intel Arc A730M which will also be equipped with an ACM-G10 GPU but house 24 Xe Cores (3072 ALUs), 24 ray tracing units, a graphics clock of 1100 MHz, 12 GB GDDR6 memory running across a 192-bit bus interface and a TDP target of 80-120W. This GPU will aim for the GeForce RTX 3060 series mobility options.
Intel Arc 5 ‘Advanced’ Gaming GPU Lineup
The Intel Arc 5 lineup will only feature one variant for now that makes use of the ACM-G10 GPU. It will feature 16 Xe Cores (2048 ALUs), 16 Ray tracing units, a 900 MHz graphics clock, 8 GB GDDR6 memory running across a 128-bit bus interface, and a TDP range of 60-80W. This chip should aim for the RTX 3060 Max-Q graphics chip.
Intel Arc 3 ‘Enhanced’ Gaming GPU Lineup
Lastly, we have the Intel Arc 3 lineup which is the entry-level and power-optimized family making use of the ACM-G11 GPU. The lineup features the Arc A370M which utilizes the full GPU config & 8 Xe Cores (1024 ALUs), 8 ray tracing units, 1550 MHz graphics clock, 4 GB of 64-bit GDDR6 memory, and a TDP range of 35-50W. This chip would be tackling the GeForce RTX 3050 series.
The second option is the Intel Arc A350M which features 6 Xe cores (768 ALUs), 6 ray tracing units, 1150 MHz graphics engine clock, a 4 GB 64-bit bus interface, and a TDP range of 25-35W which is going to aim the entry-level MX500 series options from NVIDIA.
Intel Arc A-Series Mobility GPU Lineup:
| Graphics Card Variant | GPU Variant | GPU Die | Execution Units | Shading Units (Cores) | Memory Capacity | Memory Speed | Memory Bus | TGP |
|---|---|---|---|---|---|---|---|---|
| Arc A770M | Xe-HPG 512EU | Arc ACM-G10 | 512 EUs | 4096 | 16 GB GDDR6 | 16 Gbps | 256-bit | 120-150W |
| Arc A730M | Xe-HPG 384EU | Arc ACM-G10 | 384 EUs | 3072 | 12 GB GDDR6 | 14 Gbps | 192-bit | 80-120W |
| Arc A550M | Xe-HPG 256EU | Arc ACM-G10 | 256 EUs | 2048 | 8 GB GDDR6 | 14 Gbps | 128-bit | 60-80W |
| Arc A370M | Xe-HPG 128EU | Arc ACM-G11 | 128 EUs | 1024 | 4 GB GDDR6 | 14 Gbps | 64-bit | 35-50W |
| Arc A350M | Xe-HPG 96EU | Arc ACM-G11 | 96 EUs | 768 | 4 GB GDDR6 | 14 Gbps | 64-bit | 25-35W |
Intel Arc A-Series Entry-Level Mobility GPU Performance
Intel has been tight-lipped about the performance of its Arc Alchemist graphics but they are now starting to share some performance metrics of their entry-level parts. We also know through our own sources where these GPUs will land. The ACM-G10 is expected to be an NVIDIA GA104 & AMD Navi 22 competitor while the ACM-G11 is expected to compete against NVIDIA GA106 / GA107 and AMD Navi 24 GPUs.
The flagship model should end up with around 18.5 TFLOPs FP32 compute which is 40% more than the RX 6700 XT but 9% lower than the NVIDIA RTX 3070. In terms of performance positioning, the top 512 EU variant is said to compete against the RTX 3070 / RTX 3070 Ti, the 384 EU variant is said to compete against the RTX 3060 / RTX 3060 Ti on desktops. On the laptop side, the 512 EU might be just as fast as the RTX 3080, 384 EU variant around RTX 3070 level and the 256 EU will end up against the RTX 3060.
Performance for the entry-level ACM-G11 GPU is expected to land between the GeForce GTX 1650 and GTX 1650 SUPER but with raytracing capabilities. This performance level, if priced correctly, will be highly competitive against similar solutions from NVIDIA and AMD.
Intel has shown its entry-level Arc A370M GPU offering a smooth 60 FPS+ gaming experience at 1080p across a variety of games versus the integrated Intel Iris Xe graphics. Then in another slide, they show competitive titles can deliver over 90 FPS at 1080p when running on the Arc A370M series.
Intel also highlights the content acceleration features integrated within Arc GPUs with the A370M providing a 2.4x boost over the integrated Iris Xe GPU. Intel hasn’t shared any performance figures versus the competition yet.
2 of 9
Intel Arc Alchemist GPU Gaming Features
Intel knows that making a GPU is just one part of the whole graphics ecosystem and they are also creating a wide portfolio of features that will go into power their Arc GPU lineup. Raja Koduri and Lisa Pearce have highlighted that drivers play a crucial role in graphics development & they are collaborating with a list of developers and studios to optimize their GPUs for current and next-gen AAA games and creative applications.
2 of 9
Intel’s DG2 ARC Alchemist will offer the first true gaming drivers and that is its own initiative. NVIDIA and AMD have worked for years to optimize their current suite of gaming drivers and to get Intel’s first discrete gaming GPUs on the same level as the competition is certainly a big deal. A range of driver-specific optimizations including Resizable BAR support is also headed to Linux for Intel’s Arc Alchemist GPUs.
Technologies that will be incorporated by Intel Arc GPUs include:
- XeSS (Xe Super Sampling) AI-Assisted Super Sampling Technology
- XeGTAO (Ground-Truth Ambient Occlusion)
- Xe DeepLink Technology
- XeXMX with Xe-Cores
- Hardware-Accelerated Raytracing
- Intel GameDev Boost with 1oneAPI Tool Kit
- Resizable-Bar Support for Desktops & Laptops
Intel has shown several games that will utilize XeSS in their recent demos such as 505 Game’s Death Stranding, IO Interactive’s Hitman 3, and Exor Studio’s The Riftbreaker, all of which are fairly new AAA titles. Intel showcased both games running on an undisclosed Xe-HPG ARC Alchemist GPU. Both games were compared on 1080p resolution and 4K XeSS upscaling. Although the videos are of 1080p quality, you can still see that 4K XeSS really helps enhance the visual quality in both titles.
Intel ARC Alchemist 4K XeSS Super Resolution Demo In Hitman 3:
2 of 9
Intel ARC Alchemist 4K XeSS Super Resolution Demo In Riftbreakers:
2 of 9
The upscaled images are both sharp, less blurry and show the textures on various objects in more detail. We have a very in-depth interview with the principal engineer of XeSS here which talks about how the company plans on expanding the technology in their future updates. You can also see an internal 4K XeSS demo here along with further XeSS tech details here.
Intel Arc Alchemist XeSS Technology (GDC 2022 Demo)
Intel XeSS is a super-sampling technique that leverages Machine Learning to reconstruct a low-resolution frame into a high-resolution frame, running at a fraction of the cost of rendering at a higher resolution such as 4K. The core principle of XeSS is to take advantage of ML through the integrated XMX AI acceleration hardware that is featured on the Xe-Cores for Arc Alchemist GPUs. XeSS will be fully compliant with DirectX 12 and uses the Intel Vector Shading language-based NN (Neural Networking) running on Intel’s Arc SIMD architecture.
Compared to native resolution (4K), XeSS 4K with XMX takes less than half the cost to render a scene. The technology is also DP4a compliant which means it can run on GPUs without AI or ML acceleration engines such as XMX but offer similar performance and image quality.
In terms of image quality, Intel XeSS will eliminate all kinds of ghosting with minimal artifacts (such as shimmering) when compared to other upscaling methods such as TAAU. It also comes with its own built-in sharpening engine and will be replacing TAA. Intel has stated that XeSS can achieve much higher scaling ratios without compromising quality than Temporal Supersampling or Spatial upscaling.
Intel also shared a brand new XeSS Rens demo which was running on an Intel Arc Alchemist GPU that was running at a fixed frequency. The demo was run on 1440p with raytracing and 4K with raytracing enabled. The demo was run at each resolution on 5 different XeSS presets that range from Ultra Performance, Performance, Balanced, Quality, and Ultra Quality. The Ultra performance mode offers up to 2.53x performance boost over native while Ultra Quality offers a 27% boost in performance over native at 4K resolution.
Intel Arc Alchemist Raytracing Technology (GDC 2022 Demo)
Next up, Intel talked about its raytracing approach and how they are making it better than AMD’s and NVIDIA’s approach. Intel will be bypassing SIMD divergency for hit shaders, textures and using a set of HW sorting for rays and threads to maximize uptime of each lane. This would allow Intel to essentially automatically accelerate raytracing on hardware.
In a slide showcasing performance metrics on a pre-production Intel Arc Alchemist silicon, the GPU offers 0.775x performance in RayQuery (relative) vs DXR1.0. Intel states that this is a bigger perf hit than competitor 1 (NVIDIA) while competitor 2 (AMD) takes an even bigger hit. Intel also offers their own conclusion and fixes to why this performance hit is seen on their GPUs.
In addition to XeSS support, Intel also showed a brief demo of raytracing running in Metro Exodus on its own ARC Alchemist GPUs. Another key technology Intel talked about is XeGTAO which is the brand new Gound-Truth Ambient Occlusion method, an advanced form of screen-space ambient occlusion, that delivers greater accuracy for higher image quality.
Intel ARC Alchemist Raytracing Demo In Metro Exodus:
2 of 9
Intel XeSS Is Backwards Compatible With DG1 ‘Xe-LP’ & 11th Gen CPUs
Intel has also confirmed that the XeSS technology will be backward compatible with both Xe-LP-based DG1 GPUs & iGPUs on the 11th Gen Tiger Lake GPUs.
Intel also wants to leverage the workstation and content creation market with its Xe-HPG GPUs in applications such as 3DSMax where they can give NVIDIA’s Quadro and AMD Radeon PRO graphics cards some tough competition. It is specifically stated that Intel’s ARC GPUs can offer great graphics performance within content creation and development applications.
Intel also emphasized the driver release and how it plans to release them on regular basis and new releases every time a major title is launched. The company has gone on a major hiring spree for its graphics division, acquiring renowned names from the industry over the last couple of weeks. Since we mentioned drivers, Intel will also have integrated overclocking tools within their latest driver suite available when Arc Alchemist GPUs hit the shelves.
Intel ARC Alchemist Internal XeGTAO Demo:
2 of 9
Intel also announced that devs will have access to their XeSS technology through the DevMesh program and anyone from indie to AAA developers can submit a form from the official Intel DevMesh site. Intel is touting up to 2x FPS with their XeSS Super Resolution technology so it will be great to have more options for gamers & the tech also will be workable on both NVIDIA and AMD GPUs.
Moving away from GPUs for a bit, Intel also discussed how they can leverage their hybrid design introduced in 12th Gen Alder Lake CPUs within game engines. Intel and IO Interactive have been optimizing both the GPU and CPU sides of things. It is stated that developers can leverage back-ground tasks such as AI acceleration, Character Animation, Physics, Collisions, audio-processing, and more, leaving the performance cores with their leading single-threaded performance to be available for the more demanding tasks.
Intel DeepLink Demo Using ARC Alchemist & Iris Xe GPUs:
2 of 9
The last most interesting thing shared by Intel was their DeepLink technology which will help Xe-HPG ARC Alchemist GPUs work alongside Iris Xe to boost performance in creation applications. A demo showcasing a standalone ARC Alchemist GPU and another system with DeepLink where the same ARC Alchemist GPU is working alongside an Iris Xe GPU integrated on Intel’s CPUs is shown. The DeepLink system ended up with 40% faster transcoding in Handbrake as it was able to utilize more performance out of the Iris Xe chip.
There Intel Arc 3 series is available starting now in various laptop designs while the Arc 5 and Arc 7 GPUs will be hitting mobile segment in early Summer so around May or June. Expect more info on these parts in the coming months.
Intel Arc Alchemist General Performance Expectations
Intel has been tight-lipped about the performance of its Arc Alchemist graphics. It is likely that the company will be releasing more benchmarks as they get super close to launching. But despite that, we know through some leaks where these GPUs will land. The DG2-512 is expected to be an NVIDIA GA104 & AMD Navi 22 competitor while the DG2-128 is expected to compete against NVIDIA GA106 / GA107 and AMD Navi 24 GPUs.
The flagship model should end up with around 18.5 TFLOPs FP32 compute which is 40% more than the RX 6700 XT but 9% lower than the NVIDIA RTX 3070. In terms of performance positioning, the top 512 EU variant is said to compete against the RTX 3070 / RTX 3070 Ti, the 384 EU variant is said to compete against the RTX 3060 / RTX 3060 Ti on desktops. On the laptop side, the 512 EU might be just as fast as the RTX 3080, 384 EU variant around RTX 3070 level and the 256 EU will end up against the RTX 3060.
Performance for the entry-level DG2-128 GPU is expected to land between the GeForce GTX 1650 and GTX 1650 SUPER but with raytracing capabilities. Based on initial leaks, the DG-128 has shown up to be around the same tier as GeForce RTX 3050 Ti while the DG2-512 has shown up on par with NVIDIA’s GeForce RTX 3070 Ti. This performance level, if priced correctly, will be highly competitive against similar solutions from NVIDIA and AMD.
Intel Arc Alchemist GPU Price & Availability
So that brings us to the next segment of Intel Arc Alchemist GPUs, and that is pricing & availability. One big advantage that Intel could have over AMD and NVIDIA is that with these cards, they might enter the sub-$250 US market which has not seen a worth-it product for some time. The recent AMD Radeon RX 6500 XT for $199 US has ended up as a failure in multiple perspectives while the RTX 3050 from NVIDIA for $249 US is better on the features/performance side but pricing is again going to be questionable.
Now Intel would also be bound by the same chip procurement issues that other companies are facing by relying on TSMC’s 6nm process node but the 6nm node itself isn’t that much in demand as compared to the 7nm process. So getting those wafers mostly is all about how much dollars you can flex at TSMC and Intel has really wide pockets. With that in mind, they could potentially have a bigger supply than AMD and one that can even challenge NVIDIA if they manage to price it right around the same level or lower. Lower will be a huge win for Intel but that remains to be seen. As one of Raja’s strategies, the top part may actually end up with an MSRP close to $399 US while the entry-level GPUs will end up with a top-bin price of $99-$249 US.
But availability is the more concerning thing. This is Intel’s first major discrete GPU launch for both desktop PCs and laptops. Due to driver and software delays, Intel launched its first Arc GPUs in Q2 2022 (a delay from its original Q1 2022 release window) and only in the Chinese market segment. The mobility lineup has only started making its way in the US and EU markets while the desktop cards, that have only received the entry-level Arc A380, are so far China-exclusive. The rest of the lineup is expected to launch by the end of Summer and the company hopes to bring more added optimizations through its drivers for current-generation titles based on the DirectX 12 and Vulkan APIs.
The graphics cards will be available in both reference ‘Limited Edition’, ‘Standard’, and custom models with GUNNIR prepping both Arc A380 and high-end variants in non-reference designs. Other manufacturers are also expected to offer their designs once we get closer to the proper discrete graphics card launch of Arc GPUs.
Intel Arc GPU Roadmap & Future Products
Each generation of ARC GPU will be named after fictional character classes from various games. The first is Alchemist and is the naming scheme attached to the first generation of Xe-HPG GPUs. According to Intel, the Alchemist name was derived from various fantasy games, including Final Fantasy XIV & Dungeons and Dragons. In those games, the Alchemist is able to craft powerful potions from basic herbs, elements, and other crafting materials.
The Alchemist will be the start of Intel’s ARC graphics journey as the company also revealed the codenames for its next-generation Xe-HPG lineups to be Battlemage, Celestial, and Druid. These characters were also embedded in the presentation slide during Architecture Day 2021. The site reports that the Battlemage name has been derived from the Elder Scrolls world, Celestial comes from several fantasies including the recent Marvel Eternals movie while Druid has been a common character in the fantasy RPG/RTS universe. As you can tell, these are all fantasy and mythological characters which sound cool, and it’s good to see Intel is thinking out of the box & going with more gaming-inspired names while their competitors use GPU naming conventions based on Scientists and Stars.
We have also seen rumors of the 5th Gen Arc GPU codename which is expected to be called ‘Elasti’ and expected for a 2026 release.
Currently, these characters serve as an icon for each of their respective ARC GPU generation but Intel is asking the community for input on what else could be done with them. Intel hints that they may offer ARC branded goodies such as T-Shirts with these characters and that sounds like a cool idea. One suggestion could be to feature these characters in future technology demos such as the recent XeSS showcase. With that said, you do get to see some really high-res images that you could use as wallpaper on the desktop and mobile devices to showcase your support for Intel’s ARC GPU family.
2 of 9
Intel ARC Gaming GPU Lineup
| GPU Family | Intel Xe-HPG | Intel Xe2-HPG | Intel Xe3-HPG | Intel Xe Next | Intel Xe Next Next |
|---|---|---|---|---|---|
| GPU Products | ARC Alchemist GPUs | ARC Battlemage GPUs | ARC Celestial GPUs | ARC Druid GPUs | ARC E*** GPUs |
| GPU Segment | Mainstream / High-End Gaming (Discrete) | Mainstream / High-End Gaming (Discrete) | Mainstream / High-End Gaming (Discrete) | Mainstream / High-End Gaming (Discrete) | Mainstream / High-End Gaming (Discrete) |
| GPU Gen | Gen 12 | Gen 13? | Gen 14? | Gen 15? | Gen 16? |
| Process Node | TSMC 6nm | TBA | TBA | TBA | TBA |
| Specs / Design | 512 EUs / 1 Tile / 1 GPU | TBA | TBA | TBA | TBA |
| Memory Subsystem | GDDR6 | TBA | TBA | TBA | TBA |
| Launch | 2022 | 2023? | 2024? | 2025? | 2026? |
Overall, Arc Alchemist marks the start of a new era for Intel and its graphics team, bringing the next-gen GPUs for gamers, content creators, and power users across the globe. We can’t wait to see their GPUs in action this quarter!
The post Intel Arc Alchemist ‘Xe-HPG’ GPUs Specs, Performance, Price & Availability – Everything You Need To Know by Hassan Mujtaba appeared first on Wccftech.
Powered by WPeMatico
