AMD Radeon 840M vs AMD Steam Machine GPU Comparison

Head-to-Head Benchmarks

The database records no head-to-head benchmark results for the AMD Radeon 840M versus the AMD Steam Machine GPU. Both entries carry an average benchmark score of 0 and sit at the 50th percentile among all GPUs tracked. With zero recorded wins for either part, the comparative performance picture must be drawn from the architectural and specification data on file rather than from direct frame-rate or synthetic score comparisons.

The Steam Machine GPU delivers a peak FP32 throughput of 17.56 TFLOPS, which is roughly 11.8 times the 1,484.8 GFLOPS recorded for the Radeon 840M. Pixel throughput tells a similar story: the Steam Machine GPU reaches 156.8 GPixel/s against 23.20 GPixel/s for the integrated part, a margin of approximately 6.8x. Texture rate stands at 274.4 GTexel/s versus 46.40 GTexel/s, a ratio near 5.9x in favor of the discrete console GPU. These are the largest wins in the recorded data for the Steam Machine GPU, and they reflect the fundamental gap between a 110 W dedicated part and a 15 W integrated graphics processor.

The Radeon 840M counters with a boost clock of 2900 MHz, which exceeds the Steam Machine GPU's 2450 MHz boost by 450 MHz. Its base clock of 400 MHz, however, sits far below the Steam Machine GPU's 1720 MHz base. The integrated part also holds a process-node advantage, built on TSMC 4 nm versus the Steam Machine GPU's TSMC 6 nm. The Radeon 840M further posts a higher Vulkan version in its API support list, 1.4, while the Steam Machine GPU matches that figure. In DirectX and OpenGL, both parts list 12 Ultimate (12_2) and 4.6 respectively, so no separation exists there.

The lack of any recorded head-to-head benchmarks means the database cannot show a single workload where one chip outperforms the other. The specification sheet, however, indicates that the Steam Machine GPU should dominate in raw compute, rasterization, and memory bandwidth, while the Radeon 840M retains advantages in clock speed, manufacturing process, and power envelope. Without measured results, these remain projections from the recorded hardware parameters.

Architecture Differences

The two GPUs diverge sharply in their underlying architecture. The Radeon 840M uses RDNA 3.5, built on the Krackan Point chip, and belongs to the Navi III IGP generation for Strix Point Mobile. The Steam Machine GPU uses RDNA 3.0, built on the Navi 33 chip with the codename Hotpink Bonefish, and is classified as a Console GPU for Valve. The process nodes differ as well: TSMC 4 nm for the integrated part, TSMC 6 nm for the console part.

Shading resources show a clear hierarchy. The Steam Machine GPU carries 1792 shading units, 112 texture mapping units, 64 render output units, and 28 ray tracing cores. The Radeon 840M carries 256 shading units, 16 TMUs, 8 ROPs, and 4 RT cores. These figures translate directly into the pixel, texture, and FP32 rates already cited. The Steam Machine GPU also reports a transistor count of 13,300 million on a 204 mm² die, with a transistor density of 65.2M per mm². The Radeon 840M lists transistor count and die size as unknown in the database, so no density comparison is possible for that part.

Memory architecture separates the two completely. The Radeon 840M uses system shared memory, with a system dependent bandwidth figure. The Steam Machine GPU uses 8 GB of GDDR6 on a 128-bit bus, delivering 288.0 GB/s of bandwidth. Its memory clock is listed at 2250 MHz with 18 Gbps effective data rate. The integrated part's memory clock is likewise listed as system shared, reinforcing that it depends entirely on the host platform's DRAM configuration.

Clock behavior also differs. The Radeon 840M boosts to 2900 MHz from a 400 MHz base. The Steam Machine GPU has a 1720 MHz base, a 2250 MHz game clock, and a 2450 MHz boost. The presence of a dedicated game clock on the Steam Machine GPU indicates a sustained performance target for gaming workloads, while the integrated part's wide base-to-boost range suggests aggressive power management.

Feature support is largely identical at the API level. Both parts list DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The Steam Machine GPU adds specific display outputs: 1x HDMI 2.1a and 1x DisplayPort 2.1. The Radeon 840M lists its display outputs as portable device dependent, consistent with an IGP that ships inside a mobile platform.

Power draw separates the two by a wide margin. The Radeon 840M is rated at 15 W TDP with no power connectors, while the Steam Machine GPU is rated at 110 W TDP with no power connectors listed. The integrated part occupies an IGP slot width, whereas the Steam Machine GPU measures 156 mm by 152 mm by 162 mm. The bus interface for the Radeon 840M is PCIe 4.0 x8; the Steam Machine GPU has no bus interface recorded, consistent with a console-integrated part.

The Verdict

The recorded data points to two different products built for two different roles. The Radeon 840M is a 15 W integrated GPU on TSMC 4 nm with a 2900 MHz boost and system shared memory. The Steam Machine GPU is a 110 W discrete console part on TSMC 6 nm with 8 GB GDDR6, 288.0 GB/s bandwidth, and roughly 11.8x the FP32 throughput.

For anyone choosing based on the data, the Steam Machine GPU is the clear compute leader. Its 17.56 TFLOPS, 156.8 GPixel/s, and 274.4 GTexel/s dwarf the integrated part's corresponding figures. Its 28 RT cores against 4 also points to substantially stronger ray tracing capability. The 8 GB dedicated GDDR6 pool with 288.0 GB/s bandwidth removes any dependency on host memory, which the Radeon 840M retains through its system shared architecture.

The Radeon 840M holds advantages in process node, boost clock, and power consumption. The 4 nm node versus 6 nm, the 2900 MHz boost versus 2450 MHz, and the 15 W TDP versus 110 W all favor the integrated part for efficiency-sensitive mobile designs. Its API support matches the Steam Machine GPU across DirectX, OpenGL, and Vulkan, so software compatibility should not differ on that front.

Strictly from the data, the Steam Machine GPU is the higher-performing part in every measured compute and memory metric. The Radeon 840M is the only option for platforms that need an IGP with a minimal power footprint. The choice is not between two equivalent competitors; it is between a low-power integrated solution and a high-power console GPU with a 95 W TDP gap between them.

Specification Differences

The following fields differ between the two entries in the database:

• Chip: Krackan Point (Radeon 840M) versus Navi 33 (Steam Machine GPU)
• Architecture: RDNA 3.5 versus RDNA 3.0
• Codename: None listed versus Hotpink Bonefish
• Generation: Navi III IGP (Strix Point Mobile) versus Console GPU (Valve)
• Process Node: 4 nm versus 6 nm, both TSMC
• Transistors: Unknown versus 13,300 million
• Die Size: Unknown versus 204 mm²
• Transistor Density: Not listed versus 65.2M / mm²
• Base Clock: 400 MHz versus 1720 MHz
• Boost Clock: 2900 MHz versus 2450 MHz
• Game Clock: Not listed versus 2250 MHz
• Memory Clock: System Shared versus 2250 MHz 18 Gbps effective
• Memory Size: System Shared versus 8 GB
• Memory Type: System Shared versus GDDR6
• Bus Width: System Shared versus 128 bit
• Bandwidth: System Dependent versus 288.0 GB/s
• Shading Units: 256 versus 1792
• TMUs: 16 versus 112
• ROPs: 8 versus 64
• RT Cores: 4 versus 28
• Pixel Rate: 23.20 GPixel/s versus 156.8 GPixel/s
• Texture Rate: 46.40 GTexel/s versus 274.4 GTexel/s
• FP32: 1,484.8 GFLOPS versus 17.56 TFLOPS
• FP16: 1,484.8 GFLOPS (1:1) versus 17.56 TFLOPS (1:1)
• TDP: 15 W versus 110 W
• Slot Width: IGP versus not listed
• Bus Interface: PCIe 4.0 x8 versus not listed
• Display Outputs: Portable Device Dependent versus 1x HDMI 2.1a, 1x DisplayPort 2.1
• Dimensions: Not listed versus 156 mm by 152 mm by 162 mm
• Release Date: 2025-02-28 versus 2026-06-28
• Predecessor: Navi II IGP versus not listed

Fields that match include manufacturer (AMD), DirectX 12 Ultimate (12_2), OpenGL 4.6, Vulkan 1.4, production status (Active), and the absence of a launch MSRP in both records.

FAQ

Q: Which GPU has the higher boost clock?

A: The AMD Radeon 840M boosts to 2900 MHz, while the AMD Steam Machine GPU boosts to 2450 MHz.

Q: How much memory bandwidth does the Steam Machine GPU provide?

A: The Steam Machine GPU delivers 288.0 GB/s over an 8 GB GDDR6 memory pool on a 128-bit bus.

Q: What is the FP32 performance difference between the two?

A: The Steam Machine GPU reaches 17.56 TFLOPS, compared to 1,484.8 GFLOPS for the Radeon 840M, a margin of roughly 11.8x.

Q: Are both GPUs on the same manufacturing process?

A: No. The Radeon 840M uses TSMC 4 nm, while the Steam Machine GPU uses TSMC 6 nm.

Q: What ray tracing resources does each GPU carry?

A: The Radeon 840M has 4 RT cores, while the Steam Machine GPU has 28 RT cores.

Q: Do the two GPUs support the same graphics APIs?

A: Yes. Both list DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4.