Intel Arc A380E vs Intel Arc B570 Comparison

Intel
GPU

Intel Arc A380E

CORE STATE DG2-128
VRAM 6 GB
CLOCK SPEED 2000 MHz
TDP 75 W
BUS WIDTH 96 bit
ARCHITECTURE Xe-HPG
nm
PROCESS 6 nm
LAUNCH DATE 2024
VS
Intel
GPU

Arc B570

CORE STATE BMG-G21
VRAM 10 GB
CLOCK SPEED 2500 MHz
TDP 150 W
BUS WIDTH 160 bit
ARCHITECTURE Xe2-HPG
nm
PROCESS 5 nm
LAUNCH DATE 2025

PERFORMANCE BENCHMARKS

3dmark_3dmark_steel_nomad_dx12
N/A
2,649
geekbench_opencl
N/A
83,514
geekbench_vulkan
N/A
96,844
passmark_directx_10
N/A
65
passmark_directx_11
N/A
118
passmark_directx_12
N/A
72
passmark_directx_9
N/A
164
passmark_g2d
N/A
661
passmark_g3d
N/A
14,195
passmark_gpu_compute
N/A
7,281

Analysis: Intel Arc A380E vs Intel Arc B570

The Verdict

The Intel Arc A380E and Intel Arc B570 serve completely different segments of the GPU market. The A380E is an embedded, end-of-life Alchemist part designed for low-power, single-slot installations. The B570 is an active Battlemage card aimed at mainstream desktop gaming with substantially higher compute and memory resources.

Benchmark data shows the B570 occupies the 65th percentile among all GPUs, while the A380E sits at the 50th percentile. The B570 delivers an average benchmark score of 20556, placing it within 0.1% of the NVIDIA GeForce RTX 3070 Mobile and effectively tied with the Intel Arc A750. The A380E has no recorded benchmark scores or nearest rivals in the database, meaning its performance profile is defined purely by its specifications.

Choose the B570 if you need a current, active product with 10 GB of memory, 2304 shading units, and 11.52 TFLOPS of FP32 compute. Choose the A380E only for embedded or low-power deployments where the 75 W TDP, slot-powered design, and single-slot footprint are mandatory constraints. The A380E is end-of-life and its successor is Battlemage, which is exactly the architecture the B570 uses.

Architecture Differences

The A380E uses the DG2-128 chip built on the Xe-HPG architecture, belonging to the Alchemist (Arc 3) generation. It is fabricated on TSMC's 6 nm process with 7,200 million transistors on a 157 mm² die, yielding a transistor density of 45.9 million per mm². The B570 uses the BMG-G21 chip on the Xe2-HPG architecture, part of the Battlemage (Arc 5) generation. It is built on TSMC's 5 nm process with 19,600 million transistors on a 272 mm² die, giving a density of 72.1 million per mm².

The B570's newer process node and larger die translate directly into more execution resources. The A380E has 1024 shading units, 64 texture mapping units, 32 raster operation units, and 8 ray tracing cores. The B570 more than doubles those figures with 2304 shading units, 144 TMUs, 80 ROPs, and 18 RT cores. Both cards report no tensor cores in the database.

Memory architecture differs substantially. The A380E uses 6 GB of GDDR6 on a 96-bit bus with 186.0 GB/s of bandwidth. The B570 uses 10 GB of GDDR6 on a 160-bit bus with 380.0 GB/s of bandwidth, more than double the A380E's memory throughput. Clock behavior also diverges: the A380E runs at a flat 2000 MHz base and boost, while the B570 runs at 2500 MHz base and boost. The B570's memory runs at 2375 MHz (19 Gbps effective) versus the A380E's 1937 MHz (15.5 Gbps effective).

Both cards support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, and both use a PCIe 4.0 x8 bus interface. The A380E offers four DisplayPort 2.0 outputs, while the B570 provides one HDMI 2.1a and three DisplayPort 2.1 outputs.

Where Each One Wins

The B570 wins in every compute-heavy scenario. Its FP32 throughput of 11.52 TFLOPS is nearly three times the A380E's 4.096 TFLOPS. Pixel fill rate favors the B570 at 200.0 GPixel/s versus 64.00 GPixel/s, and texture fill rate favors it at 360.0 GTexel/s versus 128.0 GTexel/s. FP16 performance follows the same pattern: 23.04 TFLOPS for the B570 against 8.192 TFLOPS for the A380E, both at a 2:1 ratio.

The A380E wins in physical footprint and power requirements. It is a single-slot card measuring 254 mm long, 127 mm high, and 20 mm wide. It draws 75 W, requires no power connectors, and needs only a 250 W suggested PSU. The B570 is a dual-slot card at 272 mm long and 115 mm high, draws 150 W, requires one 8-pin power connector, and needs a 450 W suggested PSU.

The B570's 10 GB memory capacity and 380.0 GB/s bandwidth give it a clear advantage in texture-heavy workloads and higher-resolution rendering. The A380E's 6 GB capacity and 186.0 GB/s bandwidth limit it to lighter workloads, though its 96-bit bus is adequate for its compute class.

FAQ

Q: Which GPU has better raw compute performance?

A: The B570 delivers 11.52 TFLOPS FP32 and 23.04 TFLOPS FP16, compared to the A380E's 4.096 TFLOPS FP32 and 8.192 TFLOPS FP16. The B570 is roughly 2.8 times faster in both metrics.

Q: How do their memory configurations compare?

A: The B570 has 10 GB of GDDR6 on a 160-bit bus with 380.0 GB/s bandwidth. The A380E has 6 GB of GDDR6 on a 96-bit bus with 186.0 GB/s bandwidth. The B570 also runs its memory at 2375 MHz (19 Gbps effective) versus 1937 MHz (15.5 Gbps effective).

Q: Are these cards suitable for the same power envelope?

A: No. The A380E has a 75 W TDP, requires no power connectors, and suggests a 250 W PSU. The B570 has a 150 W TDP, requires one 8-pin connector, and suggests a 450 W PSU. The A380E is also single-slot, while the B570 is dual-slot.

Q: What is the production status of each card?

A: The A380E is end-of-life and was released on 2024-03-31. Its predecessor is Xe Graphics and its successor is Battlemage. The B570 is active and was released on 2025-01-15; its predecessor is Alchemist, and it has no recorded successor.

Q: How does the B570 compare to its nearest rivals in the database?

A: The B570's average benchmark score of 20556 places it 0.1% above the NVIDIA GeForce RTX 3070 Mobile (20534), 0.1% below the Intel Arc A750 (20582), 0.4% above the NVIDIA Quadro M4000M (20480), and 0.5% below the AMD Radeon R9 M390X (20662).

Q: What display outputs does each card provide?

A: The A380E provides four DisplayPort 2.0 outputs. The B570 provides one HDMI 2.1a and three DisplayPort 2.1 outputs.

Head-to-Head Benchmarks

The database contains no direct head-to-head benchmark entries between the A380E and B570. However, the B570 has a full set of recorded benchmark scores, while the A380E has none. The B570's results include a 3DMark Steel Nomad DX12 score of 2649, a Geekbench OpenCL score of 83514, and a Geekbench Vulkan score of 96844. PassMark results for the B570 show 14195 in G3D, 7281 in GPU Compute, 661 in G2D, 164 in DirectX 9, 118 in DirectX 11, 72 in DirectX 12, and 65 in DirectX 10.

The B570's average benchmark score of 20556 puts it in the 65th percentile of all GPUs. Its nearest rival, the NVIDIA GeForce RTX 3070 Mobile, scores 20534, a delta of 0.1% in the B570's favor. The Intel Arc A750 scores 20582, which is 0.1% higher than the B570. The NVIDIA Quadro M4000M trails at 20480 (0.4% behind), and the AMD Radeon R9 M390X leads slightly at 20662 (0.5% ahead).

Given the A380E's specifications, the performance gap is predictable. The B570 has 2.25 times the shading units, 2.25 times the TMUs, 2.5 times the ROPs, and 2.25 times the RT cores. Memory bandwidth is 2.04 times higher. Texture rate is 2.81 times higher, and pixel rate is 3.13 times higher. These are structural advantages, not clock or driver artifacts.

The B570's boost clock of 2500 MHz is 25% higher than the A380E's 2000 MHz, compounding the execution resource advantage. The combination of more cores and higher clocks explains why the B570's FP32 throughput (11.52 TFLOPS) is 2.81 times the A380E's (4.096 TFLOPS).

Specification Differences

The two cards differ across nearly every measured specification. Process node: 6 nm for the A380E versus 5 nm for the B570. Transistor count: 7,200 million versus 19,600 million. Die size: 157 mm² versus 272 mm². Transistor density: 45.9M per mm² versus 72.1M per mm².

Clock speeds: the A380E runs at 2000 MHz base and boost, while the B570 runs at 2500 MHz base and boost. Memory clocks are 1937 MHz (15.5 Gbps effective) for the A380E and 2375 MHz (19 Gbps effective) for the B570.

Memory: 6 GB GDDR6 on a 96-bit bus with 186.0 GB/s bandwidth versus 10 GB GDDR6 on a 160-bit bus with 380.0 GB/s bandwidth.

Compute resources: 1024 shading units, 64 TMUs, 32 ROPs, 8 RT cores for the A380E versus 2304 shading units, 144 TMUs, 80 ROPs, 18 RT cores for the B570.

Rates: pixel rate of 64.00 GPixel/s versus 200.0 GPixel/s. Texture rate of 128.0 GTexel/s versus 360.0 GTexel/s. FP32 of 4.096 TFLOPS versus 11.52 TFLOPS. FP16 of 8.192 TFLOPS versus 23.04 TFLOPS, both at 2:1 ratio.

Power and physical: 75 W TDP, single-slot, no power connectors, 250 W suggested PSU for the A380E versus 150 W TDP, dual-slot, one 8-pin connector, 450 W suggested PSU for the B570. Dimensions are 254 mm x 127 mm x 20 mm for the A380E and 272 mm x 115 mm for the B570 (width not recorded).

Display outputs: four DisplayPort 2.0 for the A380E versus one HDMI 2.1a and three DisplayPort 2.1 for the B570. The A380E is end-of-life, released 2024-03-31, with a predecessor of Xe Graphics and successor of Battlemage. The B570 is active, released 2025-01-15, with a predecessor of Alchemist and no successor. The B570 has a launch MSRP of 219 USD.

DETAILED SPECIFICATIONS

SPECIFICATION
A380E
B570
Core Specs
Shading Units
1,024
2,304 +125.0%
Shaders
1,024
2,304 +125.0%
TMUs
64
144 +125.0%
ROPs
32
80 +150.0%
Execution Units
128
18 -85.9%
Clocks
Base Clock
2000 MHz
2500 MHz
Boost Clock
2000 MHz
2500 MHz
Memory Clock
1937 MHz 15.5 Gbps effective
2375 MHz 19 Gbps effective
Memory
Memory Size
6 GB
10 GB
VRAM (MB)
6,144
10,240 +66.7%
Memory Type
GDDR6
GDDR6
Memory Bus
96 bit
160 bit
Bandwidth
186.0 GB/s
380.0 GB/s
Cache
L1 Cache
256 KB (per EU)
L2 Cache
4 MB
13.5 MB
Performance
Pixel Rate
64.00 GPixel/s
200.0 GPixel/s
Texture Rate
128.0 GTexel/s
360.0 GTexel/s
FP32 (TFLOPS)
4.096 TFLOPS
11.52 TFLOPS
FP64 (TFLOPS)
1,024.0 GFLOPS (1:4)
720.0 GFLOPS (1:16)
FP16 (TFLOPS)
8.192 TFLOPS (2:1)
23.04 TFLOPS (2:1)
AI/RT
RT Cores
8
18 +125.0%
XMX Cores
128
144 +12.5%
Power
TDP
75 W
150 W
TDP (W)
75
150 +100.0%
Suggested PSU
250 W
450 W
Power Connectors
None
1x 8-pin
Architecture
Architecture
Xe-HPG
Xe2-HPG
GPU Name
DG2-128
BMG-G21
Generation
Alchemist (Arc 3)
Battlemage (Arc 5)
Process Size
6 nm
5 nm
Transistors
7,200 million
19,600 million
Die Size
157 mm²
272 mm²
Foundry
TSMC
TSMC
Density
45.9M / mm²
72.1M / mm²
API Support
DirectX
12 Ultimate (12_2)
12 Ultimate (12_2)
OpenGL
4.6
4.6
Vulkan
1.4
1.4
OpenCL
3.0
3.0
Shader Model
6.6
6.6
Physical
Slot Width
Single-slot
Dual-slot
Length
254 mm 10 inches
272 mm 10.7 inches
Height
127 mm 5 inches
115 mm 4.5 inches
Outputs
4x DisplayPort 2.0
1x HDMI 2.1a3x DisplayPort 2.1
Bus Interface
PCIe 4.0 x8
PCIe 4.0 x8
Other
Launch Price
219 USD
Production
End-of-life
Active
Predecessor
Xe Graphics
Alchemist
Successor
Battlemage
View Arc A380E Details View Arc B570 Details