AMD Radeon RX 7900 GRE vs NVIDIA B300 SXM6 AC Comparison

AMD
RADEON

AMD Radeon RX 7900 GRE

CORE STATE Navi 31
VRAM 16 GB
CLOCK SPEED 2245 MHz
TDP 260 W
BUS WIDTH 256 bit
ARCHITECTURE RDNA 3.0
nm
PROCESS 5 nm
LAUNCH DATE 2023
VS
NVIDIA
GEFORCE

B300 SXM6 AC

CORE STATE GB110
VRAM 288 GB
CLOCK SPEED 2032 MHz
TDP 1100 W
BUS WIDTH 8192 bit
ARCHITECTURE Blackwell Ultra
nm
PROCESS 5 nm
LAUNCH DATE 2025

PERFORMANCE BENCHMARKS

3dmark_3dmark_steel_nomad_dx12
4,814
N/A
geekbench_opencl
175,758
369,831
geekbench_vulkan
99,850
N/A
passmark_directx_10
139
N/A
passmark_directx_11
300
N/A
passmark_directx_12
107
N/A
passmark_directx_9
310
N/A
passmark_g2d
1,180
N/A
passmark_g3d
27,089
N/A
passmark_gpu_compute
15,016
N/A

Analysis: AMD Radeon RX 7900 GRE vs NVIDIA B300 SXM6 AC

Head-to-Head Benchmarks

The recorded database contains exactly one shared benchmark between these two accelerators: Geekbench OpenCL. The NVIDIA B300 SXM6 AC posts a score of 369,831, while the AMD Radeon RX 7900 GRE scores 175,758. That represents a delta of 52.5 percent in favor of the NVIDIA part. In other words, the B300 SXM6 AC delivers more than double the OpenCL performance of the RX 7900 GRE in this workload.

The magnitude of this gap is striking. The RX 7900 GRE sits at the 77th percentile among all GPUs in the database, with an average benchmark score of 32,456 across all recorded tests. The B300 SXM6 AC, by contrast, sits at the 100th percentile, meaning it outperforms every other GPU in the database on average. Its average benchmark score of 369,831 is more than eleven times higher than the RX 7900 GRE's average, though this comparison is skewed by the fact that the two parts have very different benchmark portfolios recorded.

Looking at the nearest rivals for each card helps contextualize the OpenCL result. The RX 7900 GRE's closest competitors are all older or lower-tier AMD parts: the AMD FirePro S10000 (average score 32,388, just 0.2 percent below), the AMD FirePro S9300 X2 (32,540, 0.3 percent above), the AMD Radeon RX 590 GME (32,601, 0.4 percent above), and the AMD Radeon Pro 570X (32,176, 0.9 percent below). These are extremely tight margins, indicating the RX 7900 GRE's average score is in a densely packed performance band. The B300 SXM6 AC's nearest rivals, in contrast, are all enterprise-class accelerators: the NVIDIA B200 (345,482, 7 percent below), the NVIDIA H200 NVL (334,891, 10.4 percent below), the AMD Instinct MI300X (317,994, 16.3 percent below), and the NVIDIA L40S (295,763, 25 percent below). The B300 SXM6 AC leads each of these by substantial margins.

The single head-to-head result shows a decisive victory for the NVIDIA part. The RX 7900 GRE has zero wins in shared benchmarks, while the B300 SXM6 AC has one. However, the RX 7900 GRE has ten recorded benchmarks spanning DirectX 9, 10, 11, 12, 3DMark Steel Nomad, Geekbench Vulkan, Passmark G2D, G3D, and GPU compute. The B300 SXM6 AC has only the single OpenCL result. This asymmetry means the comparison is limited in scope but still reveals a clear performance hierarchy in the one workload where both are measured.

Architecture Differences

The two products represent fundamentally different design philosophies. The AMD Radeon RX 7900 GRE uses the Navi 31 chip built on RDNA 3.0 architecture, codenamed Plum Bonito, and belongs to the Navi III (RX 7000) generation. The NVIDIA B300 SXM6 AC uses the GB110 chip built on Blackwell Ultra architecture and belongs to the Server Blackwell (Bxx) generation. Both are fabricated by TSMC on a 5 nm process, but the similarities largely end there.

The transistor counts differ enormously. The RX 7900 GRE packs 57,700 million transistors on a 529 mm² die, yielding a density of 109.1 million transistors per square millimeter. The B300 SXM6 AC contains 208,000 million transistors on a 1628 mm² die, with a density of 127.8 million per square millimeter. The B300 SXM6 AC has roughly 3.6 times the transistor count and a die more than three times larger, while also achieving a higher transistor density.

Memory configurations diverge sharply as well. The RX 7900 GRE uses 16 GB of GDDR6 on a 256-bit bus, delivering 576.0 GB/s of bandwidth. The B300 SXM6 AC uses 288 GB of HBM3e on an 8192-bit bus, delivering 8.19 TB/s. That is over fourteen times the memory capacity and over fourteen times the bandwidth. The memory clock specifications also differ: the RX 7900 GRE runs memory at 2250 MHz with 18 Gbps effective, while the B300 SXM6 AC runs at 2000 MHz with 8 Gbps effective, though the massive bus width compensates entirely.

Compute resources are scaled dramatically. The RX 7900 GRE has 5120 shading units, 320 texture mapping units, 160 ROPs, and 80 RT cores. The B300 SXM6 AC has 18,944 shading units, 592 TMUs, 24 ROPs, and 592 tensor cores. Notably, the B300 SXM6 AC has no recorded RT core count, while the RX 7900 GRE has no recorded tensor core count. The ROP disparity is unusual: the B300 SXM6 AC has far fewer ROPs (24 versus 160), which directly impacts pixel rate. The RX 7900 GRE achieves 359.2 GPixel/s, while the B300 SXM6 AC manages only 48.77 GPixel/s. Texture rate favors the NVIDIA part: 718.4 GTexel/s versus 1,202.9 GTexel/s.

Floating point performance shows a nuanced picture. The RX 7900 GRE delivers 45.98 TFLOPS FP32 and 91.96 TFLOPS FP16 with a 2:1 ratio. The B300 SXM6 AC delivers 76.99 TFLOPS FP32 and 76.99 TFLOPS FP16 with a 1:1 ratio. The NVIDIA part is 67 percent faster in FP32 but actually slower in FP16, since it does not double-rate FP16 the way the AMD part does. This suggests the B300 SXM6 AC is optimized for FP32-heavy workloads, while the RX 7900 GRE's FP16 advantage points toward different application profiles.

Clock speeds differ as well. The RX 7900 GRE has a base clock of 1287 MHz, a boost clock of 2245 MHz, and a game clock of 1880 MHz. The B300 SXM6 AC has a base clock of 1665 MHz and a boost clock of 2032 MHz, with no recorded game clock. The AMD part boosts higher, but the NVIDIA part has a higher base clock. Power draw is in different leagues entirely: the RX 7900 GRE has a TDP of 260 W with a suggested PSU of 600 W, while the B300 SXM6 AC has a TDP of 1100 W and a suggested PSU of 1500 W.

Interface and output capabilities also reflect their intended markets. The RX 7900 GRE uses PCIe 4.0 x16 and provides 1x HDMI 2.1a, 2x DisplayPort 2.1, and 1x USB Type-C. The B300 SXM6 AC uses PCIe 6.0 x16 and has no display outputs. The RX 7900 GRE is a dual-slot card with 2x 8-pin power connectors, measuring 276 mm in length, 110 mm in height, and 51 mm in width. The B300 SXM6 AC is an SXM module with no recorded dimensions and no power connector details. API support differs completely: the RX 7900 GRE supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, while the B300 SXM6 AC lists N/A for DirectX, OpenGL, and Vulkan.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA B300 SXM6 AC has an average benchmark score of 369,831, placing it at the 100th percentile of all GPUs. The AMD Radeon RX 7900 GRE has an average score of 32,456, placing it at the 77th percentile.

Q: How much faster is the B300 SXM6 AC in the shared Geekbench OpenCL test?

A: The B300 SXM6 AC scores 369,831 versus 175,758 for the RX 7900 GRE, a delta of 52.5 percent in favor of the NVIDIA part.

Q: What memory configurations do the two GPUs use?

A: The RX 7900 GRE uses 16 GB of GDDR6 on a 256-bit bus with 576.0 GB/s bandwidth. The B300 SXM6 AC uses 288 GB of HBM3e on an 8192-bit bus with 8.19 TB/s bandwidth.

Q: Which GPU has more shading units?

A: The B300 SXM6 AC has 18,944 shading units, while the RX 7900 GRE has 5,120. The NVIDIA part also has 592 TMUs versus 320, and 592 tensor cores versus none recorded for the AMD part. The RX 7900 GRE has 160 ROPs versus 24 for the B300 SXM6 AC.

Q: What are the power requirements for each?

A: The RX 7900 GRE has a TDP of 260 W and a suggested PSU of 600 W. The B300 SXM6 AC has a TDP of 1100 W and a suggested PSU of 1500 W.

Q: What display outputs does each GPU provide?

A: The RX 7900 GRE provides 1x HDMI 2.1a, 2x DisplayPort 2.1, and 1x USB Type-C. The B300 SXM6 AC provides no display outputs.

The Verdict

The data indicates two products designed for entirely different workloads. The NVIDIA B300 SXM6 AC is the clear performance leader in the only shared benchmark, doubling the RX 7900 GRE in OpenCL. Its 100th percentile ranking, 288 GB of HBM3e memory, 8.19 TB/s bandwidth, and 76.99 TFLOPS FP32 performance place it firmly in the server accelerator class. Its nearest rivals are the B200, H200 NVL, MI300X, and L40S, all enterprise parts, and it beats each by 7 to 25 percent.

The AMD Radeon RX 7900 GRE belongs to the consumer desktop segment. It has display outputs, a dual-slot form factor, PCIe 4.0 x16, DirectX 12 Ultimate support, and a 260 W TDP. Its nearest rivals are the FirePro S10000, FirePro S9300 X2, RX 590 GME, and Radeon Pro 570X, all with average scores within 0.9 percent of each other. The RX 7900 GRE does lead in pixel rate (359.2 GPixel/s versus 48.77 GPixel/s) and FP16 throughput (91.96 TFLOPS versus 76.99 TFLOPS), which suggests strengths in rasterization-heavy and FP16-oriented tasks.

The B300 SXM6 AC cannot be used for conventional graphics output, as it has no display connectors and no DirectX, OpenGL, or Vulkan support recorded. The RX 7900 GRE cannot match the B300 SXM6 AC in memory capacity, bandwidth, or raw FP32 compute. The choice between them is not a matter of preference but of application domain: the data shows the RX 7900 GRE for client-side graphics and the B300 SXM6 AC for server-side compute.

Specification Differences

| Field | AMD Radeon RX 7900 GRE | NVIDIA B300 SXM6 AC |

|--------|------------------------|---------------------|

| Architecture | RDNA 3.0 | Blackwell Ultra |

| Chip | Navi 31 | GB110 |

| Transistors | 57,700 million | 208,000 million |

| Die Size | 529 mm² | 1628 mm² |

| Transistor Density | 109.1M / mm² | 127.8M / mm² |

| Base Clock | 1287 MHz | 1665 MHz |

| Boost Clock | 2245 MHz | 2032 MHz |

| Memory Size | 16 GB | 288 GB |

| Memory Type | GDDR6 | HBM3e |

| Memory Bus Width | 256 bit | 8192 bit |

| Memory Bandwidth | 576.0 GB/s | 8.19 TB/s |

| Shading Units | 5120 | 18944 |

| TMUs | 320 | 592 |

| ROPs | 160 | 24 |

| RT Cores | 80 | N/A |

| Tensor Cores | N/A | 592 |

| Pixel Rate | 359.2 GPixel/s | 48.77 GPixel/s |

| Texture Rate | 718.4 GTexel/s | 1,202.9 GTexel/s |

| FP32 | 45.98 TFLOPS | 76.99 TFLOPS |

| FP16 | 91.96 TFLOPS (2:1) | 76.99 TFLOPS (1:1) |

| TDP | 260 W | 1100 W |

| Suggested PSU | 600 W | 1500 W |

| Slot Width | Dual-slot | SXM Module |

| Bus Interface | PCIe 4.0 x16 | PCIe 6.0 x16 |

| Display Outputs | 1x HDMI 2.1a, 2x DisplayPort 2.1, 1x USB Type-C | No outputs |

| DirectX | 12 Ultimate (12_2) | N/A |

| OpenGL | 4.6 | N/A |

| Vulkan | 1.4 | N/A |

| Release Date | 2023-07-26 | 2025-09-10 |

Where Each One Wins

The RX 7900 GRE wins in pixel throughput, delivering 359.2 GPixel/s against 48.77 GPixel/s for the B300 SXM6 AC. This indicates a strong advantage for traditional rasterization and framebuffer operations. It also wins in FP16 compute with 91.96 TFLOPS versus 76.99 TFLOPS, and it has a higher boost clock at 2245 MHz versus 2032 MHz. The RX 7900 GRE is the only one of the two with display outputs, making it the sole candidate for direct video output. It also supports consumer graphics APIs including DirectX 12 Ultimate, OpenGL 4.6, and Vulkan 1.4, none of which are available on the B300 SXM6 AC.

The B300 SXM6 AC wins in nearly every compute and memory metric. It has more than triple the transistor count (208,000 million versus 57,700 million), nearly fourteen times the memory capacity (288 GB versus 16 GB), and over fourteen times the memory bandwidth (8.19 TB/s versus 576.0 GB/s). Its FP32 throughput of 76.99 TFLOPS is 67 percent higher than the RX 7900 GRE's 45.98 TFLOPS. It has 18,944 shading units versus 5,120, 592 TMUs versus 320, and 592 tensor cores where the RX 7900 GRE records none. Its texture rate of 1,202.9 GTexel/s exceeds the RX 7900 GRE's 718.4 GTexel/s. The B300 SXM6 AC also uses PCIe 6.0 x16, while the RX 7900 GRE uses PCIe 4.0 x16.

The use-case split follows directly from these figures. The RX 7900 GRE suits client systems requiring graphics output, high pixel fill, and FP16 acceleration. The B300 SXM6 AC suits server deployments requiring massive memory capacity, extreme bandwidth, tensor core compute, and the highest FP32 throughput in the database. The 100th percentile ranking of the B300 SXM6 AC, combined with its 7 to 25 percent lead over its nearest rivals, confirms its position at the top of the recorded performance hierarchy. The 77th percentile ranking of the RX 7900 GRE, surrounded by rivals within 0.9 percent, places it in a competitive but crowded mid-to-high segment of the consumer GPU market.

DETAILED SPECIFICATIONS

SPECIFICATION
RX 7900 GRE
B300 SXM6 AC
Core Specs
Shading Units
5,120
18,944 +270.0%
Shaders
5,120
18,944 +270.0%
TMUs
320
592 +85.0%
ROPs
160
24 -85.0%
Compute Units
80
—
SM Count
—
148
Clocks
Base Clock
1287 MHz
1665 MHz
Boost Clock
2245 MHz
2032 MHz
Game Clock
1880 MHz
—
Shader Clock
1880 MHz
—
Memory Clock
2250 MHz 18 Gbps effective
2000 MHz 8 Gbps effective
Memory
Memory Size
16 GB
288 GB
VRAM (MB)
16,384
294,912 +1700.0%
Memory Type
GDDR6
HBM3e
Memory Bus
256 bit
8192 bit
Bandwidth
576.0 GB/s
8.19 TB/s
Cache
L1 Cache
256 KB per Array
256 KB (per SM)
L2 Cache
6 MB
126 MB
L3 Cache
64 MB
—
L0 Cache
64 KB per WGP
—
Performance
Pixel Rate
359.2 GPixel/s
48.77 GPixel/s
Texture Rate
718.4 GTexel/s
1,202.9 GTexel/s
FP32 (TFLOPS)
45.98 TFLOPS
76.99 TFLOPS
FP64 (TFLOPS)
1,436.8 GFLOPS (1:32)
1,202.9 GFLOPS (1:64)
FP16 (TFLOPS)
91.96 TFLOPS (2:1)
76.99 TFLOPS (1:1)
AI/RT
RT Cores
80
—
Tensor Cores
—
592
Matrix Cores
160
—
Power
TDP
260 W
1100 W
TDP (W)
260
1,100 +323.1%
Suggested PSU
600 W
1500 W
Power Connectors
2x 8-pin
—
Architecture
Architecture
RDNA 3.0
Blackwell Ultra
GPU Name
Navi 31
GB110
Codename
Plum Bonito
—
Generation
Navi III (RX 7000)
Server Blackwell (Bxx)
Process Size
5 nm
5 nm
Transistors
57,700 million
208,000 million
Die Size
529 mm²
1628 mm²
Foundry
TSMC
TSMC
Density
109.1M / mm²
127.8M / mm²
AMD MCM
GCD Transistors
45,400 million
—
GCD Die Size
304.35 mm²
—
MCD Transistors
2,050 million x6
—
API Support
DirectX
12 Ultimate (12_2)
—
OpenGL
4.6
—
Vulkan
1.4
—
OpenCL
2.2
3.0
CUDA
—
10.3
Shader Model
6.8
—
Physical
Slot Width
Dual-slot
SXM Module
Length
276 mm 10.9 inches
—
Height
110 mm 4.3 inches
—
Outputs
1x HDMI 2.1a2x DisplayPort 2.11x USB Type-C
No outputs
Bus Interface
PCIe 4.0 x16
PCIe 6.0 x16
Other
Launch Price
549 USD
—
Production
Active
Active
Predecessor
Navi II
Server Hopper
Successor
Navi IV
Server Rubin
View Radeon RX 7900 GRE Details View B300 SXM6 AC Details