AMD Radeon 860M vs AMD Radeon R9 M290X Comparison

The AMD Radeon 860M and the AMD Radeon R9 M290X represent two very different answers to the same question: how much graphics performance can you squeeze into a mobile platform? One is a modern integrated GPU built on a 4 nm process, sipping power at a 15 W TDP. The other is a dedicated MXM module from an earlier era, carrying its own GDDR5 memory and a 100 W thermal envelope. The recorded data shows the integrated solution winning the only head-to-head benchmark the two share, a result that says a great deal about how far process technology and architecture have advanced between generations.

Where Each One Wins

The benchmark database gives the 860M the edge in overall standing. It sits at the 72nd percentile across all GPUs in the database, while the R9 M290X sits at the 68th percentile. The average benchmark score tells a similar story: 26401 for the 860M against 23276 for the R9 M290X, a gap of roughly 13 percent in favor of the integrated part.

The single shared benchmark, Geekbench OpenCL, goes to the 860M by 3.3 percent, with a score of 22759 against 22028. That is the only direct comparison the database records between these two, and it favors the modern IGP despite the R9 M290X having dedicated video memory and more than double the shading hardware on paper.

Where the R9 M290X still holds advantages is in raw output specifications rather than measured benchmark results. Its pixel rate of 28.80 GPixel/s and texture rate of 72.00 GTexel/s are concrete figures, though the 860M exceeds both with 48.00 GPixel/s and 96.00 GTexel/s respectively. In practice, the R9 M290X wins on memory independence: 4 GB of dedicated GDDR5 on a 256-bit bus delivering 153.6 GB/s of bandwidth, versus system-shared memory on the 860M whose bandwidth depends entirely on the host platform. Workloads sensitive to dedicated VRAM and memory latency may still favor the older discrete part in principle, though no such benchmark appears in the database.

The 860M also wins on feature support. It is a DirectX 12 Ultimate (12_2) part with 8 ray tracing cores and Vulkan 1.4 support, while the R9 M290X tops out at DirectX 12 (11_1) feature level, Vulkan 1.2.170, and has no ray tracing hardware at all.

Architecture Differences

These two GPUs could hardly be more architecturally distinct. The 860M is built on the RDNA 3.5 architecture, part of the Navi III IGP generation associated with Strix Point Mobile, fabricated on TSMC's 4 nm process. Its chip is codenamed Krackan Point. It is an integrated design: no dedicated memory, no power connectors, slot width listed simply as IGP, and a modest 15 W TDP. It connects over PCIe 4.0 x8.

The R9 M290X belongs to the GCN 1.0 architecture, produced on TSMC's 28 nm process with the Neptune chip, and belongs to the Gem System (R9 M200) generation. It is a discrete MXM module with a 100 W TDP, a 2,800 million transistor count, a 212 mm² die, and a transistor density of 13.2M per mm². Its bus interface is PCIe 3.0 x16.

The resource split is lopsided in an interesting way. The R9 M290X has far more of nearly everything: 1280 shading units versus 512, 80 TMUs versus 32, and 32 ROPs versus 16. Yet the 860M compensates with clock speed, boosting to 3000 MHz against a ceiling of 900 MHz on the older part. The result is a higher theoretical FP32 throughput for the 860M, at 3.072 TFLOPS versus 2.304 TFLOPS, and matching FP16 throughput at 3.072 TFLOPS (1:1). The R9 M290X has no FP16 figure recorded.

Feature divergence extends to modern APIs and hardware acceleration. The 860M's 8 ray tracing cores and DirectX 12 Ultimate support make it capable of features the GCN 1.0 design simply cannot expose. Both parts support OpenGL 4.6, which is the one point of API parity.

Production status also separates them sharply: the 860M is Active, released on 2025-02-28, while the R9 M290X is End-of-life, released on 2014-01-08. The R9 M290X's lineage runs from the Solar System predecessor to the Polaris Mobile successor; the 860M follows the Navi II IGP with no recorded successor.

Head-to-Head Benchmarks

The database records one direct benchmark between these two GPUs, and it is a narrow but meaningful win for the integrated design.

Geekbench OpenCL. The 860M scores 22759 against 22028 for the R9 M290X, a 3.3 percent advantage. This is the most striking result on the page: an integrated GPU drawing a fraction of the power edges out a discrete part with more than double the shading units and its own high-bandwidth GDDR5. The result underlines how much per-shader efficiency the RDNA 3.5 architecture and the 4 nm node deliver compared to GCN 1.0 on 28 nm.

Context from each part's rival set reinforces the picture. The 860M's average score of 26401 places it in a cluster that includes the NVIDIA GeForce RTX 5060 (average score 26331, just 0.3 percent behind), the NVIDIA GeForce MX550 (26421, essentially even at 0.1 percent ahead), the AMD Radeon RX 5700 XT 50th Anniversary (26553, 0.6 percent ahead), and the NVIDIA RTX A4000 (26683, 1.1 percent ahead). In other words, the 860M benchmarks within about a percent of several well-known discrete cards.

The R9 M290X's average of 23276 puts it alongside the AMD Radeon RX 6600M (23273, a dead heat at 0 percent delta), the AMD Radeon Pro Vega 16 (23250, 0.1 percent behind), the NVIDIA P106-100 (23249, 0.1 percent behind), and the AMD Radeon AI PRO R9700 (23315, 0.2 percent ahead). Its neighborhood is similarly tightly packed, just at a lower absolute level.

The 860M also posts a Geekbench Vulkan score of 30043, a result the R9 M290X cannot match because no Vulkan run is recorded for it in the database. The R9 M290X carries a Geekbench Metal score of 24524, a test for which no 860M result exists. The overall win count in the database stands at one win for the 860M and zero for the R9 M290X.

Specification Differences

Only the fields where the two GPUs diverge are listed here.

• Chip: Krackan Point (860M) versus Neptune (R9 M290X)
• Architecture: RDNA 3.5 versus GCN 1.0
• Generation: Navi III IGP (Strix Point Mobile) versus Gem System (R9 M200)
• Process node: 4 nm versus 28 nm
• Transistors: unknown for the 860M versus 2,800 million for the R9 M290X
• Die size: unknown versus 212 mm²
• Base clock: 600 MHz versus 850 MHz
• Boost clock: 3000 MHz versus 900 MHz
• Memory: system shared (size, type, bus, and bandwidth all platform dependent) versus 4 GB GDDR5 on a 256-bit bus with 153.6 GB/s of bandwidth; memory clock of 1200 MHz (4.8 Gbps effective)
• Shading units: 512 versus 1280
• TMUs: 32 versus 80
• ROPs: 16 versus 32
• Ray tracing cores: 8 versus none
• FP32: 3.072 TFLOPS versus 2.304 TFLOPS
• FP16: 3.072 TFLOPS (1:1) versus not recorded
• Pixel rate: 48.00 GPixel/s versus 28.80 GPixel/s
• Texture rate: 96.00 GTexel/s versus 72.00 GTexel/s
• TDP: 15 W versus 100 W
• Slot width: IGP versus MXM Module
• Bus interface: PCIe 4.0 x8 versus PCIe 3.0 x16
• DirectX: 12 Ultimate (12_2) versus 12 (11_1)
• Vulkan: 1.4 versus 1.2.170
• Production status: Active versus End-of-life
• Release date: 2025-02-28 versus 2014-01-08
• Predecessor: Navi II IGP versus Solar System
• Successor: none recorded versus Polaris Mobile

Both are AMD parts fabbed at TSMC, both support OpenGL 4.6, both list portable-device-dependent display outputs, and neither requires power connectors. Neither has a recorded launch MSRP.

FAQ

Q: Which GPU is faster in benchmarks?

A: The Radeon 860M. It wins the shared Geekbench OpenCL test 22759 to 22028, a 3.3 percent margin, and holds a higher average benchmark score, 26401 versus 23276, placing it at the 72nd percentile versus the R9 M290X's 68th percentile.

Q: How does the 860M compare to the R9 M290X in power consumption?

A: The 860M is listed at a 15 W TDP as an integrated GPU, while the R9 M290X is a 100 W MXM module. The 860M delivers its benchmark win on a fraction of the power budget.

Q: Does either GPU support ray tracing?

A: Only the 860M. It has 8 ray tracing cores and DirectX 12 Ultimate (12_2) support. The R9 M290X has no ray tracing cores and is limited to DirectX 12 (11_1).

Q: Which GPU has dedicated video memory?

A: The R9 M290X, with 4 GB of GDDR5 on a 256-bit bus and 153.6 GB/s of bandwidth. The 860M uses system-shared memory, so its bandwidth depends on the host system.

Q: How do their theoretical specs compare?

A: The R9 M290X has more hardware units (1280 shading units, 80 TMUs, 32 ROPs), but the 860M's much higher clocks give it higher theoretical output: 3.072 TFLOPS FP32 versus 2.304 TFLOPS, 48.00 GPixel/s versus 28.80 GPixel/s, and 96.00 GTexel/s versus 72.00 GTexel/s.

Q: Are both GPUs still in production?

A: No. The 860M is listed as Active, while the R9 M290X is End-of-life. They were released on 2025-02-28 and 2014-01-08 respectively.