AMD Radeon Pro Vega II vs Intel Arc Pro A60 Comparison

The AMD Radeon Pro Vega II and the Intel Arc Pro A60 occupy very different positions in the professional GPU landscape, and the recorded data makes the split unusually clean. The Vega II is an end-of-life workstation flagship with an enormous memory subsystem and roughly double the benchmark throughput of the A60 across every shared compute test. The Arc Pro A60 is an active, current-generation card built on a newer node, with ray tracing hardware, modern display connectivity, and a far smaller power and physical footprint. Neither card competes with the other on its own terms, and the numbers below show exactly where each one still makes sense.

The Verdict

The data points to the Vega II for anyone whose workload is raw GPU compute or memory-bandwidth-bound. Its average benchmark score of 109,617 sits in the 94th percentile of all GPUs in the database, and it beats the Arc Pro A60 by 56% in OpenCL and 74.3% in Vulkan. Its 32 GB of HBM2 delivering 825.3 GB/s of bandwidth dwarfs the A60's 12 GB of GDDR6 at 384.0 GB/s, so large datasets and heavy texture or buffer workloads clearly favor the Vega II. The trade-offs are structural: it is an end-of-life product with a launch MSRP of 2,199 USD, a 475 W TDP, a quad-slot footprint, and an Apple MPX bus interface, meaning it is effectively tied to the Mac Pro ecosystem it was designed for.

The Arc Pro A60 is the pick where efficiency, availability, and modern feature support matter more than peak throughput. It is in active production, uses a single-slot design with a 130 W TDP and a suggested 300 W power supply versus the Vega II's 850 W recommendation, and connects over standard PCIe 4.0 x16. It supports DirectX 12 Ultimate (12_2) and Vulkan 1.4, both newer API targets than the Vega II's DirectX 12 (12_1) and Vulkan 1.3, and it carries 16 ray tracing cores the Vega II lacks entirely. Its average score of 60,326 places it in the 88th percentile, which is respectable, but it simply loses every head-to-head benchmark here.

Architecture Differences

These two cards come from opposite ends of the design spectrum. The Vega II uses AMD's Vega 20 chip on the GCN 5.1 architecture, fabricated on TSMC's 7 nm process, with 13,230 million transistors on a 331 mm² die for a density of 40.0M per mm². The A60 uses Intel's DG2-256 chip on the Xe-HPG (Alchemist) architecture, built on TSMC's 6 nm process, packing 11,500 million transistors into a smaller 269 mm² die at 42.8M per mm². The newer node gives the A60 a density edge, but the Vega II's die is substantially larger and carries far more hardware.

The resource counts reflect that. The Vega II has 4096 shading units, 256 texture mapping units, and 64 render output units, exactly double the A60's 2048 shading units and 128 TMUs, though the ROP counts match at 64 each. Where the A60 pulls ahead in fixed-function hardware is ray tracing: it has 16 RT cores while the Vega II has none.

Clock strategies differ sharply. The Vega II runs a narrow 1574 MHz base to 1720 MHz boost range, typical of a big-die pro part. The A60 starts at a 900 MHz base and boosts to 2050 MHz, a much wider range on a smaller chip. The memory story is the real divider: HBM2 on a 4096-bit bus at 825.3 GB/s for the Vega II versus GDDR6 on a 192-bit bus at 384.0 GB/s for the A60, running at 1612 Mbps and 16 Gbps effective respectively.

Where Each One Wins

The Vega II wins every benchmark the two share. Its 14.09 TFLOPS of FP32 and 28.18 TFLOPS of FP16 compute, combined with the massive memory bandwidth, make it the stronger card for GPU compute, OpenCL workloads, Vulkan-based rendering, and any task that streams large volumes of data through memory. Its texture rate of 440.3 GTexel/s is well ahead of the A60's 262.4 GTexel/s, favoring heavy texturing workloads.

The A60 wins on pixel throughput, interestingly, with 131.2 GPixel/s against the Vega II's 110.1 GPixel/s, a function of its higher boost clock paired with the same 64 ROPs. It also wins on everything the benchmark suite cannot score: power efficiency, physical size, current availability, ray tracing support, and modern API and display standards. Four DisplayPort 2.0 outputs and Vulkan 1.4 support make it the more future-proof card for driver and API longevity, while the Vega II is already end-of-life with a 1x HDMI 2.0b and 4x Thunderbolt output configuration specific to its Apple ecosystem.

FAQ

Q: Which GPU is faster in benchmarks? A: The Radeon Pro Vega II. It wins both shared tests, scoring 99,048 versus 63,485 in Geekbench OpenCL (56% ahead) and 99,621 versus 57,166 in Geekbench Vulkan (74.3% ahead).

Q: Which card has more memory bandwidth? A: The Vega II, with 825.3 GB/s from 32 GB of HBM2 on a 4096-bit bus. The A60 delivers 384.0 GB/s from 12 GB of GDDR6 on a 192-bit bus.

Q: Do either of these GPUs support ray tracing? A: Only the Arc Pro A60, which includes 16 ray tracing cores. The Vega II has no RT cores.

Q: How do they compare in power consumption and size? A: The Vega II has a 475 W TDP, a quad-slot footprint, and a suggested 850 W power supply. The A60 has a 130 W TDP, a single-slot design, and a suggested 300 W power supply.

Q: Are both cards still in production? A: No. The Vega II is listed as end-of-life, while the Arc Pro A60 is an active product.

Q: How do they rank against other GPUs? A: The Vega II averages 109,617 and sits in the 94th percentile; the A60 averages 60,326 and sits in the 88th percentile.

Head-to-Head Benchmarks

Only two tests overlap in the database, and the Vega II takes both decisively.

In Geekbench OpenCL, the Vega II scores 99,048 against the A60's 63,485, a 56% margin. This aligns with the raw compute gap: 14.09 versus 8.397 TFLOPS FP32, and 28.18 versus 16.79 TFLOPS FP16. OpenCL workloads that saturate shading hardware scale with those figures, and the result tracks almost exactly.

In Geekbench Vulkan, the gap widens. The Vega II scores 99,621 while the A60 manages 57,166, a 74.3% difference. The Vulkan result also illustrates how much the memory subsystem matters here: with 825.3 GB/s of HBM2 bandwidth versus 384.0 GB/s, the Vega II can keep its larger shader array fed in ways the A60's narrower bus cannot match, and the benchmark spread grows beyond what the compute ratio alone would predict.

For context against the wider field, the Vega II's average score of 109,617 lands it within striking distance of the AMD Radeon PRO W7900 at 110,725 (1% ahead of the Vega II) and ahead of cards like the NVIDIA RTX A5500 Mobile at 113,944 (3.8% ahead of the Vega II) and the Radeon Pro Vega II Duo at 106,750 (2.7% behind it). The A60's average of 60,326 places it essentially level with the NVIDIA GeForce RTX 4090's average of 60,347 in this database (a 0% delta), marginally behind the Radeon Pro W6600M at 61,896 (2.5% ahead of the A60), and ahead of the Radeon Pro Vega 48 at 60,140 and the Radeon PRO V710 at 58,657.

Specification Differences

• Process node: 7 nm (Vega II) vs 6 nm (A60), both on TSMC
• Transistors / die size: 13,230 million on 331 mm² vs 11,500 million on 269 mm²
• Shading units: 4096 vs 2048
• TMUs: 256 vs 128
• RT cores: none vs 16
• Base / boost clock: 1574 / 1720 MHz vs 900 / 2050 MHz
• Memory: 32 GB HBM2, 4096-bit, 825.3 GB/s vs 12 GB GDDR6, 192-bit, 384.0 GB/s
• FP32 / FP16 compute: 14.09 / 28.18 TFLOPS vs 8.397 / 16.79 TFLOPS
• Pixel rate: 110.1 vs 131.2 GPixel/s; texture rate: 440.3 vs 262.4 GTexel/s
• TDP / suggested PSU: 475 W / 850 W vs 130 W / 300 W
• Slot width: quad-slot vs single-slot
• Bus interface: Apple MPX vs PCIe 4.0 x16
• Display outputs: 1x HDMI 2.0b plus 4x Thunderbolt vs 4x DisplayPort 2.0
• API support: DirectX 12 (12_1), Vulkan 1.3 vs DirectX 12 Ultimate (12_2), Vulkan 1.4
• Release date: June 2019 for the Vega II vs June 2023 for the A60
• Production status: end-of-life vs active
• Launch MSRP: 2,199 USD for the Vega II; no launch MSRP recorded for the A60