AMD Instinct MI355X vs Intel Arc B390 Comparison

AMD
RADEON

AMD Instinct MI355X

CORE STATE MI350 256CU
VRAM 288 GB
CLOCK SPEED 2400 MHz
TDP 1400 W
BUS WIDTH 8192 bit
ARCHITECTURE CDNA 4.0
nm
PROCESS 3 nm
LAUNCH DATE 2025
VS
Intel
GPU

Arc B390

CORE STATE Panther Lake
VRAM System Shared
CLOCK SPEED 2500 MHz
TDP 80 W
BUS WIDTH System Shared
ARCHITECTURE Xe3-LPG
nm
PROCESS 3 nm
LAUNCH DATE 2026

PERFORMANCE BENCHMARKS

3dmark_3dmark_steel_nomad_dx12
N/A
1,482

Analysis: AMD Instinct MI355X vs Intel Arc B390

Head-to-Head Benchmarks

The recorded data contains no direct head-to-head benchmark results between the AMD Instinct MI355X and the Intel Arc B390. The AMD Instinct MI355X has no benchmark scores listed, while the Intel Arc B390 has a single recorded score: 1482 in the 3DMark Steel Nomad DX12 test. This places the Arc B390 at the 9th percentile among all GPUs in the database, indicating it sits near the bottom of the performance distribution.

The Intel Arc B390's nearest rivals in the database, based on its average benchmark score, are all NVIDIA parts from older generations. The GeForce GT 520MX scores 1463, which is 1.3% behind the Arc B390. The GeForce 800M scores 1460, a 1.5% deficit. The GeForce GT 625 OEM trails by 2.5% with a score of 1446, and the GeForce GT 710 is 2.7% behind at 1443. These margins are extremely tight, with the Arc B390 leading its closest competitor by just 19 points. The data shows that the Arc B390 is clustered with these entry-level legacy GPUs, all within a narrow band of roughly 40 points.

For the AMD Instinct MI355X, the absence of benchmark scores means no direct performance comparisons can be made from the database. Its percentile ranking of 50 is a positional indicator rather than a measured performance figure, as the average benchmark score is zero. The MI355X is fundamentally a different product category, designed as an OAM module for data center acceleration, while the Arc B390 is an integrated graphics processor for mobile platforms. The head-to-head comparison is therefore limited by the available data, but the specification differences are extensive and well-documented.

FAQ

Q: What is the only benchmark score recorded for the Intel Arc B390?

A: The Intel Arc B390 has one recorded benchmark result: 1482 in the 3DMark Steel Nomad DX12 test, giving it an average benchmark score of 1482 and a percentile rank of 9 among all GPUs.

Q: How does the Intel Arc B390 compare to its nearest rivals in the database?

A: The Arc B390 leads the NVIDIA GeForce GT 520MX by 1.3%, the GeForce 800M by 1.5%, the GeForce GT 625 OEM by 2.5%, and the GeForce GT 710 by 2.7%. All four rivals score within the 1443 to 1463 range.

Q: Does the AMD Instinct MI355X have any recorded benchmark scores?

A: No. The database lists no benchmark entries for the MI355X, and its average benchmark score is zero. Its percentile rank of 50 is a positional value without supporting performance data.

Q: What is the memory configuration of the AMD Instinct MI355X?

A: The MI355X uses 288 GB of HBM3e memory on an 8192-bit bus, providing 8.19 TB/s of memory bandwidth. The memory clock is 2000 MHz with 8 Gbps effective speed.

Q: What is the memory configuration of the Intel Arc B390?

A: The Arc B390 uses System Shared memory, with a System Shared type, bus width, and bandwidth that is System Dependent. Its memory clock is not specified; the database lists the memory section as System Shared throughout.

Q: What API support does each GPU provide?

A: The AMD Instinct MI355X lists N/A for DirectX, OpenGL, and Vulkan. The Intel Arc B390 supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

The Verdict

The data supports a clear separation of roles. The AMD Instinct MI355X is a data center accelerator with no display outputs, no API support for graphics APIs, and a thermal design power of 1400 W. The Intel Arc B390 is an integrated graphics processor with active production status, supporting DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4, and consuming 80 W.

For workloads requiring massive memory capacity and bandwidth, the MI355X delivers 288 GB of HBM3e at 8.19 TB/s, which is a fundamental capability for large-scale compute tasks. The Arc B390, by contrast, relies on system shared memory with system-dependent bandwidth, making it unsuitable for such workloads.

For graphics rendering and API compatibility, the Arc B390 is the only option with any recorded benchmark data. Its score of 1482 in 3DMark Steel Nomad places it in the 9th percentile, and its nearest rivals are all older NVIDIA parts with scores between 1443 and 1463. The MI355X cannot participate in such comparisons because it has no graphics API support and no benchmark entries.

The choice depends entirely on the use case. The MI355X serves compute-heavy data center applications where its 1400 W power envelope is acceptable. The Arc B390 serves portable devices where its 80 W power draw and integrated form factor are priorities. Neither product can substitute for the other based on the recorded data.

Specification Differences

The two GPUs differ across nearly every specification field. The AMD Instinct MI355X uses a 3 nm process from TSMC, while the Intel Arc B390 also uses a 3 nm process but from Intel as the foundry. The MI355X has 185,000 million transistors on a 2380 mm² die, resulting in a transistor density of 77.7M per mm². The Arc B390 lists transistor count, die size, and density as unknown or null.

Clock speeds differ substantially. The MI355X has a base clock of 1000 MHz and a boost clock of 2400 MHz, with memory at 2000 MHz (8 Gbps effective). The Arc B390 has a base clock of 300 MHz and a boost clock of 2500 MHz, with no separate memory clock specified.

Memory configurations are entirely different. The MI355X uses 288 GB of HBM3e with an 8192-bit bus and 8.19 TB/s bandwidth. The Arc B390 uses System Shared memory with a System Shared bus width and System Dependent bandwidth.

Compute resources show a wide gap. The MI355X has 16384 shading units, 1024 texture mapping units, and 0 ROPs. The Arc B390 has 1536 shading units, 48 TMUs, and 24 ROPs. The Arc B390 includes 12 ray tracing cores, while the MI355X lists none. The MI355X produces a pixel rate of 0 MPixel/s and a texture rate of 2,457.6 GTexel/s. The Arc B390 produces 60.00 GPixel/s and 120.0 GTexel/s.

Floating-point performance diverges sharply. The MI355X delivers 78.64 TFLOPS for both FP32 and FP16 (1:1 ratio). The Arc B390 delivers 7.680 TFLOPS FP32 and 15.36 TFLOPS FP16 (2:1 ratio).

Power and form factor also differ. The MI355X has a TDP of 1400 W, uses an OAM Module slot width, has no power connectors, and requires a suggested PSU of 1800 W. The Arc B390 has a TDP of 80 W, uses an IGP slot width, has no power connectors, and lists no suggested PSU. The MI355X measures 102 mm by 165 mm, while the Arc B390 has no recorded dimensions.

The MI355X uses a PCIe 5.0 x16 bus interface and has no display outputs. The Arc B390 uses an IGP bus interface with display outputs described as Portable Device Dependent. The MI355X lists N/A for all graphics APIs, while the Arc B390 supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.

Architecture Differences

The AMD Instinct MI355X is built on the CDNA 4.0 architecture, with the chip designation MI350 256CU. It belongs to the Instinct (MIx) generation. The Intel Arc B390 uses the Xe3-LPG architecture, with the chip designation Panther Lake, and belongs to the Arc Graphics-M (Panther Lake) generation.

The MI355X has no codename listed, nor does the Arc B390. The MI355X lists its predecessor as Radeon Instinct, while the Arc B390 has no predecessor. The MI355X has a release date of 2025-06-11, and the Arc B390 has a release date of 2026-01-26. The MI355X has no production status listed, while the Arc B390 is marked as Active.

The compute architectures reflect their different purposes. CDNA 4.0 is designed for data center compute, with the MI355X's 16384 shading units and 1024 TMUs aimed at massive parallel throughput. The 1:1 FP16 to FP32 ratio on the MI355X indicates that both precision levels are treated equally, a common design for compute accelerators where mixed-precision workloads are standard.

Xe3-LPG is a low-power graphics architecture intended for integrated use. The Arc B390's 2:1 FP16 to FP32 ratio indicates that half-precision operations run at twice the rate of single-precision, a typical arrangement for graphics-oriented GPUs. The presence of 12 ray tracing cores and support for DirectX 12 Ultimate confirms a focus on modern rendering features.

The memory architectures are fundamentally different. HBM3e on the MI355X provides tightly integrated, high-bandwidth storage on a dedicated 8192-bit bus. System Shared memory on the Arc B390 means the GPU uses the system's main memory, with bandwidth dependent on the host platform's memory subsystem. This is a core architectural distinction: dedicated versus shared memory.

The MI355X's lack of ROPs and zero pixel rate indicate it does not perform traditional rasterization. The Arc B390's 24 ROPs and 60.00 GPixel/s pixel rate confirm it does. The MI355X has no display outputs, while the Arc B390's display outputs are portable device dependent, consistent with an integrated graphics solution for laptops and similar devices.

The process nodes are both listed as 3 nm, but the foundries differ: TSMC for AMD and Intel for the Arc B390. The MI355X's transistor count of 185,000 million and die size of 2380 mm² represent a very large chip, while the Arc B390 has no disclosed transistor or die size data. The transistor density of 77.7M per mm² for the MI355X is calculated from those figures.

The thermal design reflects the architectural scale. The MI355X at 1400 W requires an OAM module form factor and a suggested 1800 W power supply. The Arc B390 at 80 W fits an IGP slot width with no separate power supply recommendation, drawing power through the host system. The MI355X uses a PCIe 5.0 x16 connection, while the Arc B390 uses the integrated graphics bus interface.

The release timing also differs by roughly seven months, with the MI355X dated 2025-06-11 and the Arc B390 dated 2026-01-26. The MI355X succeeds the Radeon Instinct line, while the Arc B390 has no predecessor recorded in the database.

DETAILED SPECIFICATIONS

SPECIFICATION
Instinct MI355X
B390
Core Specs
Shading Units
16,384
1,536 -90.6%
Shaders
16,384
1,536 -90.6%
TMUs
1,024
48 -95.3%
ROPs
0
24 +∞%
Compute Units
256
Execution Units
12
Clocks
Base Clock
1000 MHz
300 MHz
Boost Clock
2400 MHz
2500 MHz
Memory Clock
2000 MHz 8 Gbps effective
System Shared
Memory
Memory Size
288 GB
System Shared
VRAM (MB)
294,912
Memory Type
HBM3e
System Shared
Memory Bus
8192 bit
System Shared
Bandwidth
8.19 TB/s
System Dependent
Cache
L1 Cache
32 KB (per CU)
64 KB (per EU)
L2 Cache
32 MB
16 MB
L3 Cache
256 MB
Performance
Pixel Rate
0 MPixel/s
60.00 GPixel/s
Texture Rate
2,457.6 GTexel/s
120.0 GTexel/s
FP32 (TFLOPS)
78.64 TFLOPS
7.680 TFLOPS
FP64 (TFLOPS)
39.32 TFLOPS (1:2)
960.0 GFLOPS (1:8)
FP16 (TFLOPS)
78.64 TFLOPS (1:1)
15.36 TFLOPS (2:1)
AI/RT
RT Cores
12
XMX Cores
96
Matrix Cores
1,024
Power
TDP
1400 W
80 W
TDP (W)
1,400
80 -94.3%
Suggested PSU
1800 W
Power Connectors
None
None
Architecture
Architecture
CDNA 4.0
Xe3-LPG
GPU Name
MI350 256CU
Panther Lake
Generation
Instinct (MIx)
Arc Graphics-M (Panther Lake)
Process Size
3 nm
3 nm
Transistors
185,000 million
unknown
Die Size
2380 mm²
unknown
Foundry
TSMC
Intel
Density
77.7M / mm²
API Support
DirectX
12 Ultimate (12_2)
OpenGL
4.6
Vulkan
1.4
OpenCL
3.0
3.0
Shader Model
6.9
Physical
Slot Width
OAM Module
IGP
Length
102 mm 4 inches
Outputs
No outputs
Portable Device Dependent
Bus Interface
PCIe 5.0 x16
IGP
Other
Production
Active
Predecessor
Radeon Instinct
View Instinct MI355X Details View Arc B390 Details