AMD Radeon Instinct MI308X vs Lisuan Tech LX PRO Comparison
AMD Radeon Instinct MI308X
Lisuan Tech LX PRO
Analysis: AMD Radeon Instinct MI308X vs Lisuan Tech LX PRO
FAQ
Q: What are the memory capacities of the AMD Radeon Instinct MI308X and the Lisuan Tech LX PRO?
A: The AMD Radeon Instinct MI308X features 192 GB of HBM3 memory on a 8192-bit bus, delivering 10.3 TB/s of bandwidth. The Lisuan Tech LX PRO uses 24 GB of GDDR6 memory on a 192-bit bus, providing 432.0 GB/s.
Q: Which product uses a larger process node?
A: The Lisuan Tech LX PRO is built on a 6 nm process at TSMC, while the AMD Radeon Instinct MI308X uses a smaller 5 nm process, also at TSMC.
Q: How do the FP32 compute figures compare between the two?
A: The AMD Radeon Instinct MI308X delivers 81.72 TFLOPS of FP32 compute, which is 3.3 times higher than the 24.58 TFLOPS of the Lisuan Tech LX PRO.
Q: What is the thermal design power (TDP) of each product?
A: The AMD Radeon Instinct MI308X has a TDP of 750 W and requires a suggested power supply of 1150 W. The Lisuan Tech LX PRO has a TDP of 225 W with a suggested power supply of 550 W.
Q: Do both products support the same PCIe interface?
A: No. The AMD Radeon Instinct MI308X uses PCIe 5.0 x16, while the Lisuan Tech LX PRO uses PCIe 4.0 x16.
Q: Which product includes display outputs?
A: The Lisuan Tech LX PRO includes 4x DisplayPort 1.4a outputs. The AMD Radeon Instinct MI308X has no display outputs, consistent with its OAM Module form factor.
Architecture Differences
The two accelerators share a common foundry, TSMC, but diverge immediately after that. The AMD Radeon Instinct MI308X uses the Aqua Vanjaram chip based on the CDNA 3.0 architecture, manufactured on a 5 nm process. The Lisuan Tech LX PRO uses the 7G105 chip with the TrueGPU architecture, manufactured on a 6 nm process. The AMD part integrates a substantial 153,000 million transistors on a 1017 mm² die, giving a transistor density of 150.4M per mm². Transistor count and die size for the Lisuan Tech part are listed as unknown in the database.
The memory subsystems are fundamentally different. The MI308X uses HBM3 across an 8192-bit bus, reaching a bandwidth of 10.3 TB/s. The LX PRO uses GDDR6 across a 192-bit bus, reaching 432.0 GB/s. The AMD accelerator has 19456 shading units and 1216 texture mapping units, while the Lisuan Tech part has 6144 shading units, 192 texture mapping units, and 96 render output units. The MI308X has no render output units, and its pixel rate is recorded as 0 MPixel/s. The LX PRO delivers a pixel rate of 192.0 GPixel/s.
Clock behavior also separates the two. The AMD product has a base clock of 1000 MHz and a boost clock of 2100 MHz, with memory clocked at 2525 MHz (10.1 Gbps effective). The Lisuan Tech product has no listed base or boost clock, but its memory runs at 2250 MHz (18 Gbps effective). The AMD part supports an 8:1 FP16 ratio, delivering 653.7 TFLOPS, while the LX PRO supports a 2:1 FP16 ratio, delivering 49.15 TFLOPS. The AMD texture rate is 2,553.6 GTexel/s versus 384.0 GTexel/s for the LX PRO.
Form factor and interface details also differ. The MI308X is an OAM Module with no power connectors and no display outputs. The LX PRO is a Dual-slot card with a 1x 16-pin power connector and 4x DisplayPort 1.4a outputs. The AMD product uses PCIe 5.0 x16, while the LX PRO uses PCIe 4.0 x16. The LX PRO also carries full API support with DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.3, whereas the AMD part has no listed API support in the database. The Lisuan Tech LX PRO is marked as Active in production status and has physical dimensions of 248 mm length, 118 mm height, and 48 mm width.
The Verdict
The data separates these two products into different usage domains. The AMD Radeon Instinct MI308X is a data center accelerator with a 192 GB HBM3 pool, 10.3 TB/s of memory bandwidth, and 81.72 TFLOPS of FP32 compute. Its 750 W TDP and OAM Module form factor align with server deployments. The Lisuan Tech LX PRO is a 225 W dual-slot card with 24 GB GDDR6, display outputs, and full consumer API support, which aligns with workstation or edge use cases.
The MI308X is the clear choice for memory-capacity-bound workloads. Its 192 GB of HBM3 is eight times the 24 GB of GDDR6 on the LX PRO, and its bandwidth advantage is 23.8 times. For compute-heavy tasks, the AMD part offers 3.3 times the FP32 throughput and 13.3 times the FP16 throughput (at the respective ratios listed). The LX PRO counters with a 192.0 GPixel/s pixel rate, which the MI308X cannot match at 0 MPixel/s, plus the presence of 96 ROPs and display connectivity.
The MI308X targets the highest tier of parallel compute and large memory residency. The LX PRO targets a lower-power, display-capable segment where its 225 W TDP and 550 W suggested power supply are practical. The LX PRO also provides a ready API stack for graphics-oriented software, while the MI308X omits display and graphics API support entirely. The recorded data shows two products with little overlap: one for large-scale accelerator tasks, the other for compact, versatile deployment.
Specification Differences
| Specification | AMD Radeon Instinct MI308X | Lisuan Tech LX PRO |
|---|---|---|
| Chip | Aqua Vanjaram | 7G105 |
| Architecture | CDNA 3.0 | TrueGPU |
| Generation | Radeon Instinct (MIx) | 7G100 |
| Process Node | 5 nm | 6 nm |
| Transistors | 153,000 million | unknown |
| Die Size | 1017 mm² | unknown |
| Transistor Density | 150.4M / mm² | null |
| Base Clock | 1000 MHz | null |
| Boost Clock | 2100 MHz | null |
| Memory Clock | 2525 MHz, 10.1 Gbps effective | 2250 MHz, 18 Gbps effective |
| Memory Size | 192 GB | 24 GB |
| Memory Type | HBM3 | GDDR6 |
| Memory Bus Width | 8192 bit | 192 bit |
| Memory Bandwidth | 10.3 TB/s | 432.0 GB/s |
| Shading Units | 19456 | 6144 |
| TMUs | 1216 | 192 |
| ROPs | 0 | 96 |
| Pixel Rate | 0 MPixel/s | 192.0 GPixel/s |
| Texture Rate | 2,553.6 GTexel/s | 384.0 GTexel/s |
| FP32 | 81.72 TFLOPS | 24.58 TFLOPS |
| FP16 | 653.7 TFLOPS (8:1) | 49.15 TFLOPS (2:1) |
| TDP | 750 W | 225 W |
| Slot Width | OAM Module | Dual-slot |
| Power Connectors | None | 1x 16-pin |
| Suggested PSU | 1150 W | 550 W |
| Bus Interface | PCIe 5.0 x16 | PCIe 4.0 x16 |
| Display Outputs | No outputs | 4x DisplayPort 1.4a |
| DirectX | null | 12 Ultimate (12_2) |
| OpenGL | null | 4.6 |
| Vulkan | null | 1.3 |
| Dimensions | null | 248 mm x 118 mm x 48 mm |
| Production Status | null | Active |
| Release Date | 2023-12-05 | 2026-03-16 |
Head-to-Head Benchmarks
The recorded data shows a dominant pattern in raw compute and memory metrics, with the AMD Radeon Instinct MI308X leading in nearly every throughput category. FP32 performance sits at 81.72 TFLOPS versus 24.58 TFLOPS, a 3.3 times margin. FP16 output differs even more: 653.7 TFLOPS (8:1) against 49.15 TFLOPS (2:1), giving the AMD part a 13.3 times advantage. Texture rate follows the same direction, at 2,553.6 GTexel/s for the MI308X versus 384.0 GTexel/s for the LX PRO, a 6.7 times gap. Memory bandwidth is the largest proportional difference, with 10.3 TB/s versus 432.0 GB/s, a 23.8 times spread.
The Lisuan Tech LX PRO records its wins in areas the AMD part does not engage. Pixel rate is 192.0 GPixel/s on the LX PRO, while the MI308X is listed at 0 MPixel/s. The LX PRO also has 96 ROPs versus 0 on the AMD part. Display connectivity is exclusive to the LX PRO, which provides 4x DisplayPort 1.4a outputs; the MI308X has no outputs. The LX PRO supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.3, while the MI308X has no listed API entries.
Efficiency metrics also favor the Lisuan Tech part. The LX PRO operates at 225 W with a 550 W suggested power supply, while the MI308X operates at 750 W with a 1150 W suggested power supply. The LX PRO uses a 6 nm process and a 192-bit GDDR6 interface, which aligns with its lower power envelope. The MI308X uses a 5 nm process and an 8192-bit HBM3 interface, consistent with its far higher bandwidth and power draw.
The MI308X also leads in shading unit count, with 19456 units against 6144, and in texture mapping units, 1216 against 192. The LX PRO counters with a higher memory clock in effective Gbps terms, 18 Gbps versus 10.1 Gbps, though its 192-bit bus limits aggregate bandwidth. The MI308X has a larger transistor count and die area, while the LX PRO does not list those figures. The MI308X uses PCIe 5.0 x16; the LX PRO uses PCIe 4.0 x16. The two products also differ in release timing: the MI308X launched in 2023-12-05, while the LX PRO is dated 2026-03-16 with an Active production status.