Recent from talks
Nothing was collected or created yet.
Intel Arc
View on Wikipedia
| Release date | March 30, 2022[1] |
|---|---|
| Manufactured by | TSMC |
| Designed by | Intel |
| Codenames |
|
| Architecture | Intel Xe |
| Cores | Up to 32 Xe cores |
| Transistors | Up to 21.7 billion |
| Fabrication process | TSMC N5 TSMC N6 |
| Cards | |
| Entry-level | Arc 3 |
| Mid-range | Arc 5 |
| High-end | Arc 7 |
| API support | |
| OpenCL | 3.0[a] |
| OpenGL | 4.6 |
| Vulkan | 1.3 |
| DirectX | 12 Ultimate |
| History | |
| Predecessor | |
Intel Arc is a brand of graphics processing units (GPUs) developed by Intel, representing the company’s line of discrete GPUs for gaming, content creation, and professional applications. Arc GPUs are designed by Intel and manufactured under contract by TSMC. The brand also includes supporting graphics software and driver technologies, and is sold alongside Intel Graphics Technology, the company's line of integrated graphics processors, found in most of its processors.
Intel Arc competes with Nvidia's GeForce and AMD's Radeon products.[2] The first generation, the Arc A-series, launched in 2022 with laptop GPUs debuting in March and desktop models such as the A750 and A770 following later that year.[3][4]
The Arc Pro series for workstation use was introduced in August 2022,[5] followed by the second-generation Battlemage (B-series) GPUs, announced in December 2024. The first model, the B580, was released later that month.[6]
Etymology
[edit]According to Intel, the brand is named after the concept of story arcs found in video games.[7] Each generation of Arc is named after character classes sorted by each letter of the Latin alphabet in ascending order. They begin with A, then B, then C, and so on. The first generation is named Alchemist, while Battlemage, Celestial and Druid are the respective names for the second, third and fourth Arc generations.[8]
Graphics processor generations
[edit]Alchemist
[edit]
Developed under the previous codename "DG2", the first generation of Intel Arc GPUs (codenamed "Alchemist") released on March 30, 2022.[1][9] It comes in both add-on desktop card and laptop form factors. TSMC manufactures the die, using their N6 process.[10]
Alchemist uses the Intel Xe GPU architecture, or more specifically, the Xe-HPG variant. Alchemist supports hardware-based ray tracing, XeSS or supersampling based on neural networks (similar to Nvidia's DLSS and AMD's FSR), and DirectX 12 Ultimate.[1][11] Also supported are DisplayPort 2.0 and overclocking. AV1 fixed-function hardware encoder is included in Alchemist GPUs as part of the Intel Quick Sync Video core.[12]
Intel confirmed ASTC support has been removed from hardware starting with Alchemist and future Arc GPU microarchitectures will also not support it.[13]
Arc Alchemist does not support SR-IOV[14] or Direct3D 9 natively, instead falling back on the D3D9On12 wrapper which translates Direct3D 9 calls to their Direct3D 12 equivalents.[15][16]
Arc support OpenCL 3.0[note 1] for example, this GPU can work in the grid World Community Grid.[17]
Display connections: DisplayPort 2.0 (40 Gbit/s bandwidth) and HDMI 2.1
Desktop
[edit]| Branding and Model[18] | Launch | MSRP (USD) |
Code name | Process | Transistors (billion) | Die size (mm2) |
Core config [b] | L2 cache | Clock rate (MHz)[c] |
Fillrate | Memory | Processing power (TFLOPS) | TDP | Bus interface | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Pixel (GP/s) |
Texture (GT/s) |
Type | Size (GB) | Bandwidth (GB/s) |
Bus width | Clock (MT/s) |
Half precision (base) |
Single precision (base) |
Double precision (base) | |||||||||||||
| Arc 3 | A310 | Sep 28, 2022 | $110 | ACM-G11 (DG2-128) |
TSMC N6 |
7.2 | 157 | 6 Xe cores 768:32:16:6 (192:96:2) |
4 MB | 2000 2000 |
32 | 64 | GDDR6 | 4 GB | 124 | 64-bit | 15500 | 6.144 | 3.072 | n/a | 75 W | PCIe 4.0 x8 |
| A380 | Jun 14, 2022 | $139 | 8 Xe cores 1024:64:32:8 (256:128:2) |
2000 2050 |
64 65.6 |
128 131.2 |
6 GB | 186 | 96-bit | 8.192 8.3968 |
4.096 4.1984 |
n/a n/a | ||||||||||
| Arc 5 | A580 | Oct 10, 2023 | $179 | ACM-G10 (DG2-512) |
21.7 | 406 | 24 Xe cores 3072:192:96:24 (768:384:6) |
8 MB | 1700 1700 |
163.2 | 326.4 | 8 GB | 512 | 256-bit | 16000 | 20.890 | 10.445 | n/a | 175 W | PCIe 4.0 x16 | ||
| Arc 7 | A750 | Oct 14, 2022 | $289 | 28 Xe cores 3584:224:112:28 (896:448:7) |
16 MB | 2050 2400 |
229.6 268.8 |
393.6 460.8 |
29.3888 34.4064 |
14.6944 17.2032 |
n/a n/a |
225 W | ||||||||||
| A770 8GB | $329 | 32 Xe cores 4096:256:128:32 (1024:512:8) |
2100 2400 |
268.8 307.2 |
537.6 614.4 |
34.4064 39.3216 |
17.2032 19.6608 |
n/a n/a | ||||||||||||||
| A770 16GB | $349 | 16 GB | 560 | 17500 | ||||||||||||||||||
- ^ In OpenCL 3.0, OpenCL 1.2 functionality has become a mandatory baseline, while all OpenCL 2.x and OpenCL 3.0 features were made optional.
- ^ Shading cores (ALU): texture mapping units (TMU): render output units (ROP): ray tracing units
(tensor cores (XMX): execution units: render slices) - ^ Boost values (if available) are stated below the base value in italic.
Mobile
[edit]| Branding and Model[19] | Launch | Code name | Process | Transistors (billion) | Die size (mm2) |
Core config[a][b] | L2 cache |
Core clock (MHz)[c] |
Fillrate[d] | Memory | Processing power (TFLOPS) | TDP | Bus interface | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Pixel (GP/s) |
Texture (GT/s) |
Type | Size | Bandwidth (GB/s) |
Bus width | Clock (MT/s) |
Half precision |
Single precision |
Double precision | ||||||||||||
| Arc 3 | A350M | Mar 30, 2022 | ACM-G11 (DG2-128) |
TSMC N6 |
7.2 | 157 | 6 Xe cores 768:48:24:6 (96:96:2) |
4 MB | 1150 2200 |
27.6 52.8 |
55.2 105.6 |
GDDR6 | 4 GB | 112 | 64-bit | 14000 | 3.5328 6.7584 |
1.7664 3.3792 |
0.4416 0.8448 |
25–35 W | PCIe 4.0 ×8 |
| A370M | 8 Xe cores 1024:64:32:8 (128:128:2) |
1550 2050 |
49.6 65.6 |
99.2 131.2 |
6.3488 8.3968 |
3.1744 4.1984 |
0.7936 1.0496 |
35–50 W | |||||||||||||
| Arc 5 | A530M | Q3 2023 | ACM-G12 (DG2-256) |
12 Xe cores 1536:96:48:12 (192:192:3) |
8 MB | 1300 | 4 GB 8 GB |
224 | 128-bit | 65–95 W | |||||||||||
| A550M | Q2 2022 | ACM-G10 (DG2-512) |
21.7 | 406 | 16 Xe cores 2048:128:64:16 (256:256:4) |
900 1700 |
57.6 108.8 |
115.2 217.6 |
8 GB | 7.3728 13.9264 |
3.6864 6.9632 |
0.9216 1.7408 |
60–80 W | ||||||||
| A570M | Q3 2023 | ACM-G12 (DG2-256) |
1300 | 75–95 W | |||||||||||||||||
| Arc 7 | A730M | Q2 2022 | ACM-G10 (DG2-512) |
21.7 | 406 | 24 Xe cores 3072:192:96:24 (384:384:6) |
12 MB | 1100 2050 |
105.6 196.8 |
211.2 393.6 |
12 GB | 336 | 192-bit | 13.5168 25.1904 |
6.7584 12.5952 |
1.6896 3.1488 |
80–120 W | PCIe 4.0 ×16 | |||
| A770M | 32 Xe cores 4096:256:128:32 (512:512:8) |
16 MB | 1650 2050 |
211.2 262.4 |
422.4 524.8 |
16 GB | 512 | 256-bit | 16000 | 27.0336 33.5872 |
13.5168 16.7936 |
3.3792 4.1984 |
120–150 W | ||||||||
- ^ Shading cores (ALU): texture mapping units (TMU): render output units (ROP): ray tracing units
(tensor cores (XMX): execution units: render slices) - ^ Texture fillrate is calculated as the number of texture mapping units (TMUs) multiplied by the base (or boost) core clock speed.
- ^ Boost values (if available) are stated below the base value in italic.
- ^ Pixel fillrate is calculated as the lowest of three numbers: number of ROPs multiplied by the base core clock speed, number of rasterizers multiplied by the number of fragments they can generate per rasterizer multiplied by the base core clock speed, and the number of streaming multiprocessors multiplied by the number of fragments per clock that they can output multiplied by the base clock rate.
Workstation
[edit]| Branding and Model[20] | Launch | Code name | Process | Transistors (billion) | Die size (mm2) |
Core config[a] | L2 cache |
Core clock (MHz)[b] |
Fillrate[c][d] | Memory | Processing power (TFLOPS) | TDP | Bus interface | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Pixel (GP/s) |
Texture (GT/s) |
Type | Size | Bandwidth (GB/s) |
Bus width | Clock (MT/s) |
Half precision |
Single precision |
Double precision | ||||||||||||
| Arc Pro | A30M | Aug 8, 2022 | ACM-G11 (DG2-128) |
TSMC N6 |
7.2 | 157 | 8 Xe cores 1024:64:32:8 (128:128:2) |
4 MB | 1550 | GDDR6 | 4 GB | 112 | 64-bit | 14000 | 4.20[20] |
50 W | PCIe 4.0 x8 | ||||
| A40 | 6 GB | 192 | 96-bit | 16000 | 5.02[20] |
||||||||||||||||
| A50 | 2050 | 75 W | |||||||||||||||||||
| A60M | June 6, 2023 | ACM-G12 (DG2-256) |
16 Xe cores 2048:128:64:16 (256:256:4) |
1300 | 8 GB | 256 | 128-bit | 9.42[20] |
95 W | PCIe 4.0 x16 | |||||||||||
| A60 | 2000 | 12 GB | 384 | 192-bit | 10.04[20] |
130 W | |||||||||||||||
- ^ Shading cores (ALU): texture mapping units (TMU): render output units (ROP): ray tracing units
(tensor cores (XMX): execution Units: render slices) - ^ Boost values (if available) are stated below the base value in italic.
- ^ Pixel fillrate is calculated as the lowest of three numbers: number of ROPs multiplied by the base core clock speed, number of rasterizers multiplied by the number of fragments they can generate per rasterizer multiplied by the base core clock speed, and the number of streaming multiprocessors multiplied by the number of fragments per clock that they can output multiplied by the base clock rate.
- ^ Texture fillrate is calculated as the number of texture mapping units (TMUs) multiplied by the base (or boost) core clock speed.
Battlemage
[edit]
Battlemage (Xe2) is the second-generation Xe architecture that debuted with its low power variant in Lunar Lake mobile processors that released in September 2024.[21] On December 3, 2024, Intel announced two Arc B-Series desktop graphics cards based on the Xe2-HPG graphics architecture.[22]
Desktop
[edit]| Branding and Model[23] | Launch | MSRP (USD) |
Code name | Process | Transistors (billion) | Die size (mm2) |
Core | Cache | Memory | Fillrate[a][b] | Processing power (TFLOPS) | TDP | Bus interface | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Core Config[c] | Clock (MHz)[d] |
L1 | L2 | Type | Size | Bandwidth (GB/s) |
Bus width | Clock
(MT/s) |
Pixel
(GP/s) |
Texture
(GT/s) |
Half precision | Single precision | Double precision | ||||||||||
| Arc 5 | B570 | Jan 16, 2025 | $219 | BMG-G21 | TSMC | 19.6 | 272 | 18 Xe Cores (144) 2304:144:72:18:144
(128:128:5) |
1700 2500 |
4.5 MB | 10 MB | GDDR6 | 10 GB | 380 | 160-bit | 19000 | 122.4 200.0 | 244.8 360.0 |
23.04 |
11.52 |
1.44 |
150 W | PCIe 4.0 x8 |
| B580 | Dec 13, 2024 | $249 | 20 Xe Cores (160) 2560:160:80:20:160
(160:160:5) |
1700 2670 |
5 MB | 12 MB | 12 GB | 456 | 192-bit | 136.0 213.6 |
272.0 427.2 |
27.34 |
13.67 |
1.709 |
190 W | ||||||||
- ^ Pixel fillrate is calculated as the number of render output units (ROPs) multiplied by the base (or boost) core clock speed.
- ^ Texture fillrate is calculated as the number of texture mapping units (TMUs) multiplied by the base (or boost) core clock speed.
- ^ Xe2-HPG Cores (Xe Vector Engines)
Unified Shaders : Texture Mapping Units : Render Output Units : Ray Tracing Cores : XMX Cores - ^ Boost values (if available) are stated below the base value in italics.
Workstation
[edit]| Branding and Model[24] | Launch | MSRP (USD) | Code name | Process | Transistors (billion) | Die size (mm2) |
Core | L2 cache |
Fillrate[a][b] | Memory | Processing power (TFLOPS) | TDP | Bus interface | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Config[c] | Clock (MHz)[d] |
Pixel (GP/s) |
Texture (GT/s) |
Type | Size | Bandwidth (GB/s) |
Bus width | Clock (MT/s) |
Half precision | Single precision | Double precision | XMX Half Precision | |||||||||||
| Arc Pro | B50[25][26] | Sept 3, 2025 | $349 | BMG-G21 | TSMC | 19.6 | 272 | 16 Xe2-cores 2048:128:64:16:128 (128:128:4) |
1700
2600 |
4 MB | 87
133 |
218
332.8 |
GDDR6 | 16 GB | 224 | 128-bit | 14000 | 21.3 | 10.65 | 1.33 | 170 | 70 W | PCIe 5.0 x8 |
| B60 | Q3 2025 | 20 Xe2-cores 2560:160:80:20:160 (160:160:5) |
2400 | 16 MB | 192 | 384 | 24 GB | 456 | 192-bit | 19000 | 24.5 | 12.8 | 1.54 | 197 | 120-200 W | ||||||||
- ^ Pixel fillrate is calculated as the lowest of three numbers: number of ROPs multiplied by the base core clock speed, number of rasterizers multiplied by the number of fragments they can generate per rasterizer multiplied by the base core clock speed, and the number of streaming multiprocessors multiplied by the number of fragments per clock that they can output multiplied by the base clock rate.
- ^ Texture fillrate is calculated as the number of texture mapping units (TMUs) multiplied by the base (or boost) core clock speed.
- ^ Shading cores (ALU): texture mapping units (TMU): render output units (ROP): ray tracing units
(tensor cores (XMX): execution Units: render slices) - ^ Boost values (if available) are stated below the base value in italic.
Future generations
[edit]Intel has revealed future generations of Intel Arc GPUs under development: Celestial (Xe3P), and Druid (Xe4).[27][28] Additionally, Panther Lake series iGPUs will be based on the Xe3 architecture.
Intel XeSS
[edit]Intel XeSS is a real-time deep learning image upsampling technology developed primarily for use in video games as a competitor to Nvidia's DLSS and AMD's FSR technologies. Additionally, XeSS is not restricted to Arc graphics cards, similar to FSR. It utilizes XMX instructions exclusive to Arc graphics cards, but will fall back to utilizing DP4a instructions on competing GPUs that have support for DP4a instructions. XeSS is trained with 64 samples per pixel as opposed to Nvidia DLSS's 16 samples per pixel (16K reference images).[29][30] XeSS 3 supports Multi-Frame Generation (MFG), similar to DLSS 4. This feature is exclusive to Intel Arc graphics cards, and will be supported on both Arc Battlemage and Alchemist series cards. Games that support XeSS 2 will support XeSS 3 at launch of XeSS 3.
| Quality preset[a] | Scale factor[b] | Render scale[c] | Scale factor[b]
(1.0-1.2) |
Render scale[c]
(1.0-1.2) |
|---|---|---|---|---|
| Native Anti-Aliasing(since 1.3) | 1.00x(since 1.3) | 100% | N/A | |
| Ultra Quality Plus(since 1.3) | 1.30x(since 1.3) | 77.0% | ||
| Ultra Quality | 1.50×(since 1.3) | 66.7% | 1.30x | 77.0% |
| Quality | 1.70×(since 1.3) | 58.8% | 1.50x | 66.7% |
| Balanced | 2.00×(since 1.3) | 50.0% | 1.70x | 58.8% |
| Performance | 2.30×(since 1.3) | 43.5% | 2.00x | 50.0% |
| Ultra Performance(since 1.3) | 3.00×(since 1.3) | 33.3% | N/A | |
- ^ The algorithm does not necessarily need to be implemented using these presets; it is possible for the implementer to define custom input and output resolutions.
- ^ a b The linear scale factor used for upsampling the input resolution to the output resolution. For example, a scene rendered at 540p with a 2.00× scale factor would have an output resolution of 1080p.
- ^ a b The linear render scale, compared to the output resolution, that the technology uses to render scenes internally before upsampling. For example, a 1080p scene with a 50% render scale would have an internal resolution of 540p.
Issues
[edit]Drivers
[edit]Performance on Intel Arc GPUs has suffered from poor driver support, particularly at launch. An investigation by Gamers Nexus discovered 43 known driver issues with Arc GPUs, prompting a response and acknowledgement of the issues from Intel.[32][33][34] Intel CEO Pat Gelsinger also blamed driver problems as a reason for Arc's delayed launch.[35] A beta driver from October 2022 accidentally reduced the memory clock by 9% on the Arc A770 from 2187 MHz to 2000 MHz, resulting in a 17% reduction in memory bandwidth.[36] This particular issue was later fixed.[37] Intel provides an open source driver for Linux.[38]
DirectX 9 compatibility
[edit]As of the Alchemist generation, Arc only includes direct hardware support for the DirectX 11 & 12 and Vulkan graphics APIs, with the older DirectX 9 & 10 and OpenGL APIs being supported via a real-time compatibility layer built into Intel's graphics driver.[39] As a result, Alchemist GPUs perform noticeably worse than competing Nvidia and AMD GPUs in software that can only use these older APIs, including multiple DirectX 9-based esports games such as Counter-Strike: Global Offensive, League of Legends and StarCraft II: Wings of Liberty.[40] There is also a performance gap between DirectX 11 and DirectX 12.
A December 2022 driver update improved Arc compatibility and performance with DirectX 9-based games.[41] According to Intel, the driver update made Arc GPUs up to 1.8x faster in DirectX 9 games.[42] A February 2023 driver update further improved Arc's performance on DirectX 9-based games.[43]
Legacy BIOS compatibility
[edit]Intel Arc requires a UEFI BIOS with resizable BAR support for optimal performance.[44] UEFI Class 1 and Class 2 BIOS are not supported by Intel Arc.
Footnotes
[edit]- ^ In OpenCL 3.0, OpenCL 1.2 functionality has become a mandatory baseline, while all OpenCL 2.x and OpenCL 3.0 features were made optional.
References
[edit]- ^ a b c hachman, Mark (March 14, 2022). "Intel will launch its Arc GPUs on March 30". PCWorld. Retrieved December 25, 2023.
- ^ Warren, Tom (August 16, 2021). "Intel enters the PC gaming GPU battle with Arc". The Verge. Retrieved August 29, 2021.
- ^ "Intel® Arc™ A750 Graphics – Product Specifications". Intel. Retrieved July 24, 2023.
- ^ Gartenberg, Chaim (March 30, 2022). "Intel's first Arc GPUs are now available for laptops". The Verge. Retrieved April 2, 2022.
- ^ "Intel Unveils Arc Pro GPU Products". Intel. August 8, 2022. Retrieved September 15, 2022.
- ^ "Intel Launches Arc B-Series Graphics Cards". Intel. Retrieved December 12, 2024.
- ^ Ung, Gordon (August 19, 2021). "Intel's Arc gaming GPU: Price, specs and availability". PCWorld. Archived from the original on August 29, 2021. Retrieved August 29, 2021.
- ^ "Intel Introduces New High-Performance Graphics Brand: Intel Arc". Intel Newsroom. August 16, 2021. Retrieved October 5, 2022.
- ^ Rutherford, Sam (August 16, 2021). "Intel Names New Brand of GPUs That Will Hit Shelves Next Year". Gizmodo. Archived from the original on January 3, 2022. Retrieved August 29, 2021.
- ^ Cunningham, Andrew (August 20, 2021). "Intel provides more details on its Arc GPUs, which will be made by TSMC". Ars Technica. Archived from the original on August 29, 2021. Retrieved August 29, 2021.
- ^ "Intel® Arc™ A-series Graphics Gaming API Guide". Intel. Archived from the original on July 20, 2022. Retrieved September 15, 2022.
- ^ Szewczyk, Chris (April 4, 2022). "Intel Arc GPUs will support AV1 encode and decode". PC Gamer. Archived from the original on July 6, 2022. Retrieved September 15, 2022.
- ^ "intel: ASTC support was removed on Gfx12.5 (!13206) · Merge requests · Mesa / mesa · GitLab". GitLab. October 5, 2021. Archived from the original on October 7, 2021. Retrieved October 7, 2021.
- ^ "Graphics Virtualization Technologies Support for Each Intel Graphics Family". Intel. Archived from the original on August 9, 2022. Retrieved September 15, 2022.
- ^ "Which Intel® Graphics Products Support DirectX 9* (DX9)?". Intel. Archived from the original on August 27, 2022. Retrieved September 15, 2022.
- ^ Killian, Zak (August 15, 2022). "Intel Xe And Arc Graphics Lack DX9 Support Forcing DX12 Emulation". HotHardware. Retrieved October 10, 2022.
- ^ "Help". World Community Grid. Retrieved September 15, 2022.
- ^ "Intel Arc Graphics". Intel. Retrieved December 23, 2022.
- ^ "Intel Arc Graphics". Intel.
- ^ a b c d e "Intel® Arc™ Pro A-Series Graphics". Intel.
- ^ Lam, Chester (October 8, 2024). "Lunar Lake's iGPU: Debut of Intel's Xe2 Architecture". Chips and Cheese. Retrieved December 3, 2024.
- ^ Hollister, Sean (December 3, 2024). "Intel announces $249 Arc B580 and $219 Arc B570 'Battlemage' graphics cards". The Verge. Retrieved December 3, 2024.
- ^ "Intel Arc Graphics". Intel. Retrieved December 23, 2022.
- ^ "Intel® Arc™ Pro B-Series Graphics". Intel.
- ^ Discuss, AleksandarK (September 3, 2025). "Intel Arc Pro B50 GPU Arrives at $349 for Small-Form-Factor Workstations". TechPowerUp. Retrieved September 11, 2025.
- ^ Lee, John (September 3, 2025). "Intel Arc Pro B50 Review A 16GB SFF Mini GPU". ServeTheHome. Retrieved September 11, 2025.
- ^ Smith, Ryan (August 16, 2021). "Intel Video Cards Get a Brand Name: Arc, Starting with 'Alchemist' in Q1 2022". AnandTech. Archived from the original on January 8, 2022. Retrieved August 29, 2021.
- ^ "Intel Introduces New High-Performance Graphics Brand: Intel Arc". Intel. Archived from the original on March 16, 2022. Retrieved August 29, 2021.
- ^ Solca, Bogdan (August 8, 2022). "More details on Intel's AI-based Xe SuperSampling tech launching with the ARC GPUs revealed by principal engineer". NotebookCheck. Archived from the original on July 3, 2022. Retrieved July 3, 2022.
- ^ "Intel® Iris® Xe MAX Graphics Open Source Programmer's Reference Manual For the 2020 Discrete GPU formerly named "DG1" Volume 11: Media Engines February 2021, Revision 1.0" (PDF). Intel. Archived (PDF) from the original on July 3, 2022. Retrieved September 15, 2022.
- ^ Mujtaba, Hassan (March 24, 2022). "Intel Showcases Arc Alchemist GPU Performance With XeSS & Raytracing Enabled, Calls XeSS Better Than Temporal Upscaling". Wccftech. Archived from the original on March 24, 2022. Retrieved March 24, 2022.
- ^ Gamers Nexus (August 1, 2022). "Worst We've Tested: Broken Intel Arc GPU Drivers". YouTube. Retrieved December 8, 2022.
- ^ McLoughlin, Aleksha (October 8, 2022). "Intel Arc driver issues – Are they fixed?". PC Guide. Retrieved December 8, 2022.
- ^ Pearce, Lisa (August 19, 2022). "Engineering Arc – 8/19/2022". Intel. Retrieved December 8, 2022.
- ^ Cunningham, Andrew (August 9, 2022). "Rumors, delays, and early testing suggest Intel's Arc GPUs are on shaky ground". Ars Technica. Retrieved December 8, 2022.
- ^ "Some Intel Arc A770 Limited Edition GPUs may show up with lower memory clock". VideoCardz. October 24, 2022. Retrieved December 8, 2022.
- ^ Klotz, Aaron (October 28, 2022). "Intel Arc A770 GPU Memory Clock Bug Fixed With Driver Update". Tom's Hardware. Retrieved December 8, 2022.
- ^ Larabel, Michael (August 25, 2022). "Intel Arc Graphics Running On Fully Open-Source Linux Driver". Phoronix. Retrieved January 13, 2023.
- ^ Roach, Jacob (August 22, 2022). "Bad news: Intel's Arc GPU issues run much deeper than performance". Digital Trends. Retrieved December 8, 2022.
- ^ Kan, Michael (December 7, 2022). "With New Driver, Intel Arc GPUs Run Older DirectX 9 Games Up to 79% Faster". PCMag. Retrieved December 8, 2022.
- ^ Shrout, Ryan (December 6, 2022). "Upward Trajectory: Improvements to DirectX 9 Games on Intel® Arc™ Graphics". Intel. Archived from the original on June 16, 2023. Retrieved December 8, 2022.
- ^ "Intel Arc GPUs get performance boost for DirectX 9 games, CS:GO now up to 1.8x faster". VideoCardz. December 7, 2022. Retrieved December 8, 2022.
- ^ "Intel Arc Graphics Updates, New Bundle, and Pricing". Intel. Archived from the original on October 1, 2023. Retrieved February 1, 2023.
- ^ "Intel® Arc™ A-Series Graphics – Desktop Quick Start..." Intel. Retrieved July 4, 2024.
External links
[edit]Intel Arc
View on GrokipediaHistory and Development
Brand Announcement
Intel's efforts in discrete graphics began with the i740 in 1998, marking its initial foray into dedicated GPUs, but the company subsequently shifted focus to integrated graphics solutions throughout the 2000s and 2010s. A notable attempt to re-enter the discrete market came with the Larrabee project, announced in 2008 as a high-performance GPU architecture, which was canceled in December 2009 due to development challenges and performance concerns.[9] This left Intel primarily emphasizing integrated graphics like the Iris Xe series until the resurgence of ambitions in discrete GPUs under the Xe architecture umbrella. On August 16, 2021, during Intel's annual Architecture Day event, the company officially unveiled the Intel Arc brand, signaling its renewed commitment to competing in the discrete graphics market against established players like NVIDIA and AMD.[10][11] Arc was positioned as a comprehensive brand encompassing high-performance graphics hardware, software, and services tailored for gaming and content creation workloads, with plans to span multiple generations of products.[12] The announcement included teasers of the underlying Xe-HPG (High Performance Graphics) architecture, which promised advancements in ray tracing and AI-accelerated features. Intel revealed that the first Arc products, codenamed Alchemist, would launch in the first quarter of 2022, with subsequent generations named Battlemage, Celestial, and Druid.[11] To differentiate in the competitive landscape, Intel highlighted strategic elements such as manufacturing on TSMC's 6nm process node for improved efficiency, hardware support for AV1 video encoding to enable superior content creation capabilities, and the introduction of XeSS (Xe Super Sampling) as an AI-based upscaling technology akin to NVIDIA's DLSS.[13]Etymology and Naming
The name "Arc" for Intel's high-performance graphics brand was developed by the branding agency Catchword in collaboration with Intel, emphasizing a short, evocative term that conveys progression and innovation in the gaming and creative ecosystems. Drawing from narrative structures, "Arc" references the concept of story arcs in video games and storytelling, symbolizing the evolving journey of users, creators, and the technology itself—"Every game, gamer, and creator has a story, and every story has an Arc." This choice also evokes geometric and dynamic imagery, such as a curved trajectory or segment of a circle, representing comprehensive graphics solutions that guide users toward future advancements. The brand was officially unveiled on August 16, 2021, as part of Intel's entry into the discrete GPU market.[14] Positioned distinctly within Intel's portfolio, "Intel Arc" serves as the dedicated brand for discrete graphics products, encompassing hardware, software, and services tailored for gamers and creators, while "Intel Graphics" remains the umbrella for integrated graphics solutions in CPUs. This separation allows Arc to focus on high-end, standalone performance, differentiating it from Intel's broader Xe architecture branding, which applies to all graphics technologies. The Arc moniker thus establishes a dedicated identity for consumer-facing discrete GPUs, aligning with Intel's strategy to compete directly with established players like NVIDIA and AMD in the dedicated graphics segment. The generational codenames under the Arc brand adopt a thematic convention inspired by character classes from fantasy role-playing games (RPGs), evoking magical and adventurous archetypes to underscore the innovative "superpowers" of each iteration. The first generation, Alchemist, embodies transformation and foundational change, reflecting the alchemy of turning raw compute power into visual experiences. Subsequent generations build on this narrative: Battlemage for the second, emphasizing combat-oriented magic and enhanced performance in demanding scenarios; Celestial for the third, drawing from heavenly and divine motifs to signify elevated capabilities; and Druid for the fourth, rooted in nature-based harmony and sustainability themes. These names were revealed alongside the brand launch, highlighting Intel's intent to infuse its GPU lineup with a storytelling progression akin to RPG character development.[15] Internally, early development used alphanumeric codenames such as DG2 for the Alchemist generation, which transitioned to the public-facing Arc branding to better communicate the product's consumer appeal and ecosystem integration. This evolution from technical project labels to thematic, marketable names supports Intel's long-term vision for Arc as a cohesive family of graphics solutions spanning multiple generations.Initial Launches and Milestones
Intel Arc's initial product launches began with the Alchemist generation in 2022, marking Intel's entry into the discrete graphics market for both mobile and desktop segments. The mobile variants debuted on March 30, 2022, integrated into laptops from manufacturers including Acer, ASUS, Dell, HP, and Lenovo, featuring models such as the Arc A350M and A370M targeted at 1080p gaming and content creation.[16] Desktop models followed later that year, with the entry-level Arc A380 releasing in June 2022 at approximately $150, built on the Xe-HPG architecture with 8 Xe-cores.[17] Higher-end desktop options, including the Arc A750 and A770, launched in October 2022, priced at $289 and $449 respectively, expanding availability to retail channels.[18] The second-generation Battlemage GPUs advanced the lineup in late 2024, with the Arc B580 Limited Edition becoming available on December 13, 2024, starting at $249 and featuring the Xe2 architecture for improved performance in gaming and AI workloads.[2] The mid-range Arc B570 followed on January 16, 2025, at $219, broadening the discrete graphics offerings for mainstream users.[19] Higher-end Battlemage models, such as the anticipated Arc B770 based on the BMG-G31 die with 32 Xe2 cores and 16GB VRAM, were reported in development for a potential late 2025 release, signaling Intel's continued push toward more powerful discrete solutions.[20] Looking ahead, the third-generation Celestial GPUs, codenamed Xe3, entered pre-silicon validation in May 2025, a critical milestone where the architecture undergoes testing for frequency, power, and voltage optimization prior to tapeout.[21] In October 2025, Intel announced details of the Xe3 architecture as part of its Panther Lake mobile processors, building on Xe2 with enhancements for AI and graphics, and confirmed that Xe3P variants would power the next-generation C-series GPUs.[22][23] This phase positions Celestial for an anticipated market entry in late 2025 or early 2026, focusing on enhanced AI and ray tracing capabilities to compete in high-performance computing.[24] Key milestones in the early years included hardware support for AV1 encoding and decoding, introduced with Alchemist in 2022 as the first discrete GPUs to enable royalty-free AV1 hardware acceleration in applications like Handbrake and Adobe tools.[25] That same year, Intel integrated Xe Super Sampling (XeSS), an AI-based upscaling technology, into the Arc ecosystem, with initial support in games via SDKs for Unreal Engine and Unity, achieving performance parity with competitors like DLSS 2.0 and FSR 2.0.[26] In September 2025, Intel intensified its recruitment for high-end GPU development, posting job openings for SoC engineers to optimize future Arc desktop graphics, underscoring ongoing commitment to the discrete market despite industry challenges.[27] Partnerships played a pivotal role in Arc's rollout, with Intel collaborating with TSMC for fabrication of Alchemist and subsequent generations on processes like 6nm, leveraging external capacity to accelerate production.[28] Additionally, Intel worked with game developers to optimize titles for Arc hardware, including early XeSS integrations in over 200 games by 2025 through plugins for major engines, enhancing upscaling and frame generation adoption.[29][30]Technical Architecture
Xe Microarchitecture Overview
The Intel Xe microarchitecture represents a unified GPU architecture introduced by Intel in 2020, designed to span a wide range of applications from low-power integrated graphics in processors to high-performance discrete GPUs under the Arc brand.[31] This architecture, encompassing variants such as Xe-LP for efficient integrated solutions and Xe-HPG for discrete high-performance graphics processing, emphasizes scalability across power envelopes and form factors while maintaining a common instruction set and design principles.[32] Core to its design is modularity, allowing flexible configuration of compute resources to address diverse workloads like gaming, AI inference, and media processing, marking a departure from the more rigid structures in prior generations. At the heart of the Xe microarchitecture are Xe-cores, which serve as the primary execution units integrating vector processing, ray tracing hardware, and media engines for video decode/encode. Each Xe-core in the Xe-HPG variant includes 16 vector engines for general-purpose graphics and compute tasks, alongside 16 matrix engines dedicated to AI acceleration via XMX units that support efficient INT8 and BF16 operations.[33] Complementing these are DP4a instructions, which enable compact matrix mathematics by performing four 8-bit integer multiplications and additions in a single 32-bit operation, facilitating AI and imaging workloads without specialized hardware in fallback scenarios.[34] The architecture supports a unified programming model for memory in oneAPI, allowing developers to write code that shares data between CPU and GPU, with discrete implementations using dedicated VRAM and PCIe-based system memory access via Resizable BAR for hybrid workloads.[32] Xe implementations leverage advanced process nodes for improved density and efficiency, with the first-generation Alchemist discrete GPUs fabricated on TSMC's 6 nm (N6) node, second-generation Battlemage on TSMC's 5 nm (N5) process, and third-generation Celestial planned for Intel's 18A node to enable further scaling.[35] It fully supports modern APIs including DirectX 12 Ultimate for features like variable rate shading and mesh shaders, Vulkan for cross-platform graphics, and OpenCL for parallel computing.[3] Compared to earlier Intel graphics generations like Gen9 (Broadwell/Skylake) and Gen11 (Ice Lake), which relied on fixed execution unit arrays with limited scalability, Xe introduces greater modularity through slice-based tiling of Xe-cores and enhanced interconnects, allowing seamless expansion from 16 to over 500 execution units while doubling performance-per-watt in graphics tasks.[36][37] This evolution prioritizes a data-parallel execution model optimized for throughput in both rasterization and ray-traced rendering.Core Technologies and Innovations
Intel Arc GPUs incorporate hardware support for AV1 video encoding and decoding, a royalty-free codec that delivers superior compression efficiency compared to predecessors like H.264 and HEVC. This capability enables high-quality 8K video streaming and playback with reduced bandwidth requirements, starting from the Alchemist generation and continuing across subsequent architectures.[38] The integrated media engine handles both 8-bit and 10-bit AV1 formats at resolutions up to 8K, supporting applications in content creation and streaming services.[38] Ray tracing acceleration is provided through dedicated ray-tracing units (RTUs) attached to each Xe-core, facilitating hardware-accelerated ray-triangle intersection and bounding volume hierarchy traversal for real-time rendering. These units enable hybrid rendering pipelines that combine rasterization with ray-traced effects, improving visual fidelity in games and simulations while maintaining performance.[39] Complementing this, Intel Arc supports DirectX 12 Ultimate features such as variable rate shading (VRS), which allows developers to apply different shading rates across the screen for optimized performance, and mesh shading, which streamlines geometry processing to reduce overhead in complex scenes.[39] AI acceleration is driven by XMX engines, specialized matrix multiply units integrated into each Xe-core that perform dense matrix operations for deep learning tasks. These engines support upscaling technologies like XeSS and noise reduction in imaging workflows, delivering 128 FP16 operations per cycle per engine to enhance inference and training efficiency.[40] Power efficiency innovations include adaptive boosting mechanisms that dynamically adjust clock speeds based on workload demands and deep learning-based power management, which uses AI models to predict and optimize energy allocation across the GPU.[41] In the evolution of these technologies, the Battlemage (Xe2) generation introduces enhanced RT units with doubled bounding volume hierarchy buffer sizes and improved dispatch efficiency, yielding up to 50% better ray tracing performance over Alchemist.[42] The Celestial (Xe3) architecture further advances AI capabilities with evolved tensor cores derived from XMX, providing greater matrix throughput for advanced inference and supporting higher core counts in power-constrained environments.[43]Graphics Processor Generations
Alchemist (First Generation)
The Alchemist generation marked Intel's entry into the discrete graphics processing unit (GPU) market, debuting as the first family under the Arc brand in 2022. Built on the Xe-HPG microarchitecture, these GPUs were fabricated using TSMC's 6 nm process node, enabling a balance of performance and efficiency through features like hardware-accelerated ray tracing and mesh shading.[3][44] The architecture supports up to 32 Xe-cores, each containing 16 execution units for a total of up to 4,096 shaders, and integrates up to 16 GB of GDDR6 memory across a 256-bit interface for high-bandwidth tasks.[6] This design choice emphasized scalability across consumer and professional applications, with innovations like dedicated matrix engines for AI workloads distinguishing it from prior Intel integrated graphics.[3] For desktop systems, Alchemist offered variants targeting entry-level to high-end gaming and content creation. The Arc A380 served as the entry-level option with 8 Xe-cores, 6 GB of GDDR6 memory on a 96-bit bus delivering 186 GB/s bandwidth, and a 75 W thermal design power (TDP), making it suitable for budget builds and light 1080p gaming.[45] The mid-range Arc A750 featured 28 Xe-cores, 8 GB GDDR6 on a 256-bit bus with 512 GB/s bandwidth, and a 225 W TDP, positioning it for 1440p performance in rasterization-heavy workloads.[46] At the top end, the Arc A770 provided 32 Xe-cores with options for 8 GB or 16 GB GDDR6 (512 GB/s or 560 GB/s bandwidth, respectively) and a 225 W TDP, aimed at enthusiasts seeking higher frame rates and future-proofing through expanded VRAM.[6][47]| Variant | Xe-Cores | Memory | Bandwidth | TDP |
|---|---|---|---|---|
| Arc A380 | 8 | 6 GB GDDR6 | 186 GB/s | 75 W |
| Arc A750 | 28 | 8 GB GDDR6 | 512 GB/s | 225 W |
| Arc A770 | 32 | 8/16 GB GDDR6 | 512/560 GB/s | 225 W |