AMD Radeon R9 370

AMD Radeon R9 370 is a 28 nm graphics card from the Pirate Islands generation, built on the GCN 1.0 architecture with the Trinidad chip. It occupies the 50th percentile among all GPUs in the database, placing it squarely in the middle of the performance spectrum. With a 2,800 million transistor count on a 212 mm² die, the card offers a transistor density of 13.2M per mm². The R9 370 operates at a base clock of 925 MHz and a boost clock of 975 MHz, delivering a peak FP32 throughput of 2.496 TFLOPS.

Benchmark Performance

The R9 370’s average benchmark score is zero in the current dataset, which means no direct synthetic or gaming scores are available for comparison. However, its raw compute specifications allow for meaningful interpretation. The card’s 2.496 TFLOPS of FP32 performance is derived from 1280 shading units running at the boost clock. This places it in a range where it can handle 1080p gaming at medium settings for most titles from its era, though modern games will require reduced resolutions or lower quality presets.

The pixel rate of 31.20 GPixel/s and texture rate of 78.00 GTexel/s indicate that the R9 370 can fill frames at a moderate pace. The pixel rate is calculated from 32 ROPs and the 975 MHz boost clock, while the texture rate comes from 80 TMUs. These figures suggest the card is balanced for its class, with texture throughput being roughly 2.5 times the pixel rate, which is typical for a GCN 1.0 design. The 50th percentile ranking means half of all GPUs in the database perform better, and half perform worse — a true midpoint in the distribution.

How It Compares

The nearestRivals array is empty in the provided data, so no direct percentage deltas can be cited against specific competing models. This absence of comparative scores means the R9 370 must be evaluated on its own merits. Given its 50th percentile standing, it would sit alongside other mid-range cards from the 2015 era, but without explicit rival names and deltaPct values, a precise positional analysis is impossible.

From a generational standpoint, the R9 370 is part of the Pirate Islands series, succeeding the Volcanic Islands generation and preceding Arctic Islands. This lineage places it after the R9 200 series and before the R9 300 series refresh, though the card itself is branded within the R9 300 lineup. The GCN 1.0 architecture is older compared to later GCN revisions, which typically offered better performance-per-clock and improved feature sets.

The 28 nm process node from TSMC was standard for its time, and the 2,800 million transistor count is substantial for a mid-range chip. The die size of 212 mm² is moderate, suggesting a balanced design that prioritizes efficiency over raw size. In the absence of rival scores, the card’s position can be inferred from its percentile ranking — it is neither a budget entry nor a high-end performer.

Ray Tracing and Feature Set

The R9 370 has no dedicated ray tracing cores or tensor cores, as indicated by the null values in the FACT PACK. This means it relies entirely on traditional rasterization for rendering. The card supports DirectX 12 at feature level 11_1, which provides access to basic DX12 features like explicit multi-adapter and asynchronous compute, but not the full DX12 Ultimate feature set that includes ray tracing and mesh shaders.

OpenGL 4.6 and Vulkan 1.2.170 are both supported, offering modern API compatibility for titles that use these interfaces. Vulkan 1.2.170 is a relatively recent version, which is notable for a card from 2015 — it suggests AMD has maintained driver support for this architecture. The GCN 1.0 architecture does not include hardware-accelerated ray tracing, so any ray-traced effects would need to be computed via compute shaders, which would be prohibitively slow on this hardware.

The display outputs include 2x DVI, 1x HDMI 1.4a, and 1x DisplayPort 1.2. The HDMI 1.4a standard limits output to 4K at 30 Hz, while DisplayPort 1.2 can handle 4K at 60 Hz. This makes the card usable for multi-monitor setups, but not ideal for high-refresh-rate 1440p or 4K gaming. The interface is PCIe 3.0 x16, which provides sufficient bandwidth for the card’s 179.2 GB/s memory throughput.

FAQ

Q: Does the AMD Radeon R9 370 support ray tracing?

A: No. The card has no RT cores or tensor cores, and its GCN 1.0 architecture does not include hardware ray tracing acceleration. Any ray-traced effects would be handled by compute shaders, which would be impractically slow.

Q: What is the maximum supported display resolution?

A: The card supports 4K output via DisplayPort 1.2 at 60 Hz, while HDMI 1.4a is limited to 4K at 30 Hz. For higher refresh rates, 1440p or 1080p are more practical resolutions.

Q: Which DirectX version is supported?

A: The R9 370 supports DirectX 12 at feature level 11_1. This provides partial DX12 functionality but lacks features like ray tracing, variable rate shading, and mesh shaders found in later DX12 Ultimate hardware.

Q: How much video memory does the card have?

A: The card comes with 2 GB of GDDR5 memory on a 256-bit bus, delivering 179.2 GB/s of bandwidth. This is sufficient for 1080p gaming but may be limiting at higher resolutions or with high-resolution texture packs.

Q: What is the power consumption of the R9 370?

A: The card has a TDP of 110 W and requires a single 6-pin power connector. A 300 W power supply is recommended, which is modest for a mid-range GPU.

Q: Is the card still in production?

A: No, the production status is listed as end-of-life. The card was released on May 4, 2015, and has been discontinued, though it may still be available on the secondhand market.

Who Should Consider It

The R9 370 is best suited for 1080p gaming at medium settings in older titles or indie games. Its 2.496 TFLOPS of FP32 performance and 31.20 GPixel/s pixel rate are adequate for this resolution, but the 2 GB VRAM capacity becomes a bottleneck with modern games that require more memory. At 1440p, the card would struggle to maintain playable frame rates in demanding titles, and 4K gaming is not realistic beyond very light workloads.

Given its 50th percentile ranking, the card sits in the middle of the performance distribution. Users with a 300 W PSU and a need for a basic GPU that can handle esports titles or older AAA games at 1080p would find it acceptable. However, the lack of ray tracing and limited VRAM make it a poor choice for current-generation gaming. The card’s GCN 1.0 architecture also means it lacks some of the efficiency improvements of later GCN revisions, so performance-per-watt is not exceptional.

For resolution-specific guidance: at 1080p, the card can handle medium settings in most pre-2018 games. At 1440p, expect to drop to low settings or reduce the resolution scale. At 4K, the card is only viable for 2D games or very undemanding titles. The DisplayPort 1.2 output does support 4K at 60 Hz for desktop use, but gaming at that resolution is not practical.

Power and Cooling

The R9 370 has a TDP of 110 W, which is modest for a mid-range GPU. This low power draw means it can be cooled by a dual-slot cooler without excessive noise or heat. The card requires a single 6-pin power connector, and AMD recommends a 300 W power supply. This makes the card compatible with a wide range of existing PSUs, including older units that may lack high-wattage PCIe connectors.

The dual-slot design suggests a substantial heatsink and fan assembly, which should be adequate for the 110 W thermal load. The card’s dimensions are 221 mm in length and 111 mm in height, making it compatible with most mid-tower cases. The power connector placement is standard, and the card does not require additional power beyond the single 6-pin. Given the 28 nm process node, the 110 W TDP is reasonable for the performance on offer, though newer architectures achieve similar performance with lower power.

The 975 MHz boost clock is relatively low, which helps keep power consumption in check. The memory runs at 1400 MHz, or 5.6 Gbps effective, which is standard for GDDR5 of that era. Users with a 300 W PSU should have no issues powering this card alongside a typical CPU and other components, as the total system draw would remain well within the PSU’s capacity.

Memory Subsystem

The R9 370 is equipped with 2 GB of GDDR5 memory on a 256-bit bus, providing a memory bandwidth of 179.2 GB/s. This configuration is well-suited for 1080p gaming at the time of release, as 2 GB was the standard capacity for mid-range cards in 2015. However, the memory bandwidth is a critical factor at higher resolutions, and 179.2 GB/s is modest by modern standards.

At 1080p, the 179.2 GB/s bandwidth is generally sufficient for most textures and effects, but the 2 GB capacity can be exhausted by games with large texture packs or high-resolution assets. At 1440p, the combination of lower bandwidth and limited capacity would cause stuttering and texture pop-in in memory-intensive scenes. The 256-bit bus is wider than the 128-bit buses found on lower-tier cards, which helps maintain throughput, but the 2 GB limit is the primary constraint.

The memory clock of 1400 MHz (5.6 Gbps effective) is moderate, and the 179.2 GB/s bandwidth is roughly 20% lower than what some competing cards offered with similar bus widths. For users who plan to play at 1080p with medium settings, the memory subsystem is adequate. For higher resolutions, the card’s 32 ROPs and 2 GB VRAM would become bottlenecks long before the bandwidth is the limiting factor. The 31.20 GPixel/s pixel rate is also tied to the memory subsystem, and at 4K, the card would be severely limited by both bandwidth and fill rate.