AMD Radeon R9 370X vs AMD Radeon RX 9070 XT Comparison

Head-to-Head Benchmarks

The recorded data contains two directly comparable benchmark results between the AMD Radeon R9 370X and the AMD Radeon RX 9070 XT. The two cards split these head-to-head tests evenly, with one decisive victory for each. In Geekbench OpenCL, the R9 370X scores 25,893 against the RX 9070 XT's 17,428. That is a 48.6% advantage for the older card, a surprisingly large margin that reflects the very different driver and compute paths these two architectures expose to OpenCL workloads. The RX 9070 XT is not merely behind here; it is substantially behind, and this result dominates the average benchmark comparison in an unusual way.

In Geekbench Vulkan, the situation reverses completely. The RX 9070 XT scores 65,879, while the R9 370X manages only 9,018. The delta is reported as -86.3% from the perspective of the RX 9070 XT, meaning the R9 370X trails by that enormous percentage. In absolute terms, the newer card is roughly 7.3 times faster in Vulkan. That is a generational chasm in API-level performance, and it aligns with the architectural divergence between GCN 1.0 and RDNA 4.0.

The overall average benchmark scores in the database tell a more nuanced story. The R9 370X carries an average score of 15,862, while the RX 9070 XT averages 13,543. The older card actually leads the newer one by about 17% in this aggregate metric, but that is heavily skewed by the OpenCL outlier. The RX 9070 XT has ten recorded benchmark entries, including several PassMark tests that pull its average down, while the R9 370X has only three entries. The average is therefore not a clean apples-to-apples comparison; the head-to-head tests are the more reliable signal.

Looking at the nearest rivals for each card provides additional context. The R9 370X sits at the 58th percentile of all GPUs, with rivals like the AMD Radeon RX 7700 (average 15,852, delta 0.1%), the AMD Radeon RX 9060 (average 16,014, delta -1%), the AMD Radeon Pro W5500 (average 15,679, delta 1.2%), and the NVIDIA GeForce RTX 3060 Ti (average 16,129, delta -1.7%). The R9 370X is essentially neck-and-neck with these cards in average score, despite being from an older generation. The RX 9070 XT, at the 55th percentile, has nearest rivals that are far less powerful: the AMD Radeon HD 7770M (average 13,536, delta 0.1%), the NVIDIA GeForce GTX 570 (average 13,515, delta 0.2%), the NVIDIA P106-090 (average 13,470, delta 0.5%), and the AMD Radeon Pro 555 (average 13,407, delta 1%). This is a striking result: the RX 9070 XT's average score places it alongside decade-old midrange parts, purely because of the OpenCL and PassMark results that dominate its benchmark profile.

Where Each One Wins

The R9 370X wins decisively in OpenCL compute workloads. Its 25,893 Geekbench OpenCL score is 48.6% higher than the RX 9070 XT's 17,428. For any application that relies heavily on OpenCL for general-purpose GPU compute, the older card is the clear performer according to the recorded data. This could include legacy scientific computing, certain video encoding pipelines, or older rendering packages that use OpenCL kernels. The R9 370X also holds its own in the aggregate average score, leading the RX 9070 XT by roughly 17% overall, though that gap is inflated by the newer card's poor OpenCL showing.

The RX 9070 XT wins in Vulkan, a modern cross-platform graphics and compute API. Its 65,879 Geekbench Vulkan score is 86.3% higher in relative terms than the R9 370X's 9,018. That is a massive margin, and it reflects the RDNA 4.0 architecture's native support for Vulkan 1.4, versus the GCN 1.0 card's Vulkan 1.2.170 support. For modern games, VR applications, or any Vulkan-based compute workload, the RX 9070 XT is overwhelmingly faster. The newer card also has more benchmark entries in the database, including PassMark DirectX 9 through 12 tests, which the R9 370X lacks entirely. The RX 9070 XT's PassMark G3D score of 26,795 and PassMark GPU Compute score of 15,864 are not directly comparable to the R9 370X's Geekbench results, but they indicate a broad capability across multiple test suites.

The use-case split is clean. Legacy OpenCL compute favors the R9 370X. Modern Vulkan graphics and compute heavily favor the RX 9070 XT. The R9 370X also appears to be a more consistent performer in the average metric, likely because its three benchmark results are all from the Geekbench suite and thus share a similar driver path. The RX 9070 XT's average is dragged down by lower scores in PassMark DirectX 10 (154) and DirectX 12 (78), which are not representative of its Vulkan performance but still weigh on the aggregate.

FAQ

Q: Which card has a higher Geekbench OpenCL score?

A: The AMD Radeon R9 370X scores 25,893 in Geekbench OpenCL, which is 48.6% higher than the AMD Radeon RX 9070 XT's 17,428.

Q: How much faster is the RX 9070 XT in Vulkan?

A: The RX 9070 XT scores 65,879 in Geekbench Vulkan, while the R9 370X scores 9,018. The newer card leads by 86.3% in relative terms, making it roughly 7.3 times faster in that test.

Q: What is the average benchmark score for each card?

A: The R9 370X has an average benchmark score of 15,862, while the RX 9070 XT averages 13,543. The older card leads by about 17% in this aggregate metric.

Q: How do the two cards compare in terms of percentile rank among all GPUs?

A: The R9 370X sits at the 58th percentile of all GPUs, while the RX 9070 XT is at the 55th percentile. Despite the newer architecture, the RX 9070 XT's percentile is lower due to its benchmark profile.

Q: Which card has more recorded benchmark entries?

A: The RX 9070 XT has ten benchmark entries, including multiple PassMark tests and a 3DMark Steel Nomad result. The R9 370X has only three entries, all from Geekbench.

Q: What are the closest rivals for each card based on average score?

A: The R9 370X's nearest rival is the AMD Radeon RX 7700 with a delta of 0.1%, followed by the RX 9060 (-1%), the Pro W5500 (1.2%), and the RTX 3060 Ti (-1.7%). The RX 9070 XT's nearest rival is the AMD Radeon HD 7770M (0.1%), then the GTX 570 (0.2%), the P106-090 (0.5%), and the Pro 555 (1%).

Specification Differences

The two cards differ in nearly every major specification category. The R9 370X uses 2 GB of GDDR5 memory on a 256-bit bus, yielding 179.2 GB/s of bandwidth. The RX 9070 XT uses 16 GB of GDDR6 memory on the same 256-bit bus, but bandwidth jumps to 644.6 GB/s, a 3.6x increase. The memory clock is also different: the R9 370X runs at 1400 MHz with 5.6 Gbps effective, while the RX 9070 XT runs at 2518 MHz with 20.1 Gbps effective.

The GPU clocks diverge sharply. The R9 370X has a base clock of 980 MHz and a boost of 1030 MHz. The RX 9070 XT has a base of 1660 MHz, a boost of 2970 MHz, and a game clock of 2400 MHz. The compute resources are also vastly different. The R9 370X has 1,280 shading units, 80 TMUs, and 32 ROPs. The RX 9070 XT has 4,096 shading units, 256 TMUs, and 128 ROPs. The RX 9070 XT also includes 64 ray tracing cores, which the R9 370X lacks entirely.

The process node and physical dimensions differ as well. The R9 370X is built on TSMC's 28 nm node with a die size of 212 mm² and 2,800 million transistors. The RX 9070 XT uses a 4 nm TSMC node with a die size of 357 mm² and 53,900 million transistors. Transistor density is 13.2 million per mm² for the older card and 151.0 million per mm² for the newer one. Thermal and power specifications also change: the R9 370X has a TDP of 180 W with dual 6-pin power connectors and a suggested 450 W PSU, while the RX 9070 XT has a TDP of 304 W with dual 8-pin connectors and a suggested 700 W PSU.

The bus interface advances from PCIe 3.0 x16 to PCIe 5.0 x16. Display outputs differ: the R9 370X offers 2x DVI, 1x HDMI 1.4a, and 1x DisplayPort 1.2, while the RX 9070 XT offers 1x HDMI 2.1b and 3x DisplayPort 2.1a. The R9 370X is 221 mm long and 111 mm tall, while the RX 9070 XT has no recorded dimensions. The R9 370X is end-of-life with a launch date of 2015-08-26, while the RX 9070 XT is active with a launch date of 2025-03-05. The R9 370X has a launch MSRP of 199 USD, while the RX 9070 XT has a launch MSRP of 599 USD.

Architecture Differences

The R9 370X is built on GCN 1.0, AMD's Graphics Core Next architecture from the Pirate Islands generation (R9 300 series). It uses the Trinidad chip. The RX 9070 XT is built on RDNA 4.0, part of the Navi IV generation (RX 9000 series), using the Navi 48 chip. This is a fundamental architectural shift from a unified shader design with a focus on compute to a gaming-oriented design with dedicated ray tracing hardware.

The R9 370X supports DirectX 12 (11_1), OpenGL 4.6, and Vulkan 1.2.170. The RX 9070 XT supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The newer card's Vulkan 1.4 support is a major differentiator, as it enables features and optimizations that the older card cannot access. The DirectX 12 Ultimate designation also implies full support for hardware ray tracing and mesh shaders, which are absent on the GCN 1.0 part.

The FP32 compute throughput is vastly different: the R9 370X delivers 2.637 TFLOPS, while the RX 9070 XT delivers 48.66 TFLOPS. The RX 9070 XT also supports FP16 at a 1:1 ratio, matching its FP32 throughput, while the R9 370X has no recorded FP16 capability. Pixel rate and texture rate follow the same pattern: the R9 370X has 32.96 GPixel/s and 82.40 GTexel/s, while the RX 9070 XT has 380.2 GPixel/s and 760.3 GTexel/s. These are 11.5x and 9.2x increases, respectively.

The transistor density difference is notable: 13.2 million transistors per mm² on 28 nm versus 151.0 million per mm² on 4 nm. This is a 11.4x density increase, enabled by the much smaller process node. The R9 370X's predecessor is Volcanic Islands and its successor is Arctic Islands, indicating its place in the GCN lineage. The RX 9070 XT's predecessor is Navi III, and it has no successor recorded, sitting at the top of the Navi IV generation.

The memory architecture also differs in type and speed, though both use a 256-bit bus. The R9 370X's GDDR5 at 5.6 Gbps effective provides 179.2 GB/s, while the RX 9070 XT's GDDR6 at 20.1 Gbps effective provides 644.6 GB/s. The 16 GB capacity on the newer card is eight times the 2 GB on the older one, which has implications for texture-heavy workloads and high-resolution rendering. The ray tracing cores on the RX 9070 XT are a distinct hardware addition that GCN 1.0 simply cannot emulate in software, and this is reflected in the Vulkan benchmark gap.