AMD Radeon RX 7900 GRE vs NVIDIA GeForce RTX 2080 Ti Comparison

The AMD Radeon RX 7900 GRE and the NVIDIA GeForce RTX 2080 Ti make for an intriguing comparison precisely because the recorded data refuses to follow the usual narrative. One is a 2023 RDNA 3.0 card, the other a 2018 Turing flagship that has since reached end-of-life, yet the database places them just two percentile points apart overall. Digging into the individual tests, however, reveals a contest that is far more lopsided than the aggregate scores suggest: the 7900 GRE wins 8 of 10 head-to-head benchmarks, several by margins that completely reorder the relationship between these two designs.

The Verdict

For anyone choosing strictly on the recorded measurements, the Radeon RX 7900 GRE is the stronger and more current option. It wins 8 of 10 benchmarks, posts an average benchmark score of 32,456 against 29,783, and sits in the 77th percentile versus the 75th percentile for the RTX 2080 Ti. Its advantage is not marginal in the tests it wins: 57.9% in PassMark DirectX 11, 49.3% in GPU compute, 37.1% in Geekbench OpenCL, and 36.1% in 3DMark Steel Nomad. It also carries 16 GB of GDDR6 against 11 GB, holds an active production status while the 2080 Ti is end-of-life, and offers a newer platform: PCIe 4.0 x16 instead of PCIe 3.0 x16, plus DisplayPort 2.1 and HDMI 2.1a outputs.

The case for the RTX 2080 Ti is narrower but real. It wins Geekbench Vulkan by 24.9% and PassMark DirectX 10 by 9.2%, and it holds two architectural aces the GRE lacks: 544 tensor cores (the 7900 GRE lists none) and higher memory bandwidth, 616.0 GB/s versus 576.0 GB/s thanks to a 352-bit bus against a 256-bit one. If Vulkan API work or tensor-based workloads dominate the use case, the data indicates the older card still has a lane where it leads.

Architecture Differences

These are two different philosophies of GPU design separated by several process generations. The RX 7900 GRE is built on the Navi 31 chip using TSMC's 5 nm process, packing 57,700 million transistors into a 529 mm² die, for a density of 109.1M per mm². The RTX 2080 Ti uses the TU102 chip on a 12 nm TSMC process, with 18,600 million transistors spread across a much larger 754 mm² die and a density of only 24.7M per mm². That is a density gap of well over four times, and it explains much of what follows: the GRE simply does more with far less silicon.

The architectural split shows up in the execution resources. The GRE fields 5120 shading units, 320 TMUs, 160 ROPs and 80 RT cores, against 4352 shading units, 272 TMUs, 88 ROPs and 68 RT cores on the 2080 Ti. Those ratios compound at the throughput level: 359.2 GPixel/s versus 136.0 GPixel/s pixel rate, 718.4 versus 420.2 GTexel/s texture rate, and 45.98 versus 13.45 TFLOPS FP32, with FP16 doubling on both (91.96 versus 26.90 TFLOPS).

Clocking strategies differ too. The GRE runs a 1287 MHz base and 2245 MHz boost with a 1880 MHz game clock, and its memory runs at 2250 MHz, 18 Gbps effective. The 2080 Ti clocks 1350 MHz base and 1545 MHz boost with memory at 1750 MHz, 14 Gbps effective. Yet the older card counters with a wider 352-bit bus, giving it 616.0 GB/s of bandwidth against 576.0 GB/s, despite slower memory chips. The NVIDIA card also carries 544 tensor cores where the GRE lists none, the one hardware block that tilts decisively toward Turing. Both support DirectX 12 Ultimate (12_2), OpenGL 4.6 and Vulkan 1.4, both are dual-slot cards with 2x 8-pin connectors and a 600 W suggested PSU, and TDPs are close: 260 W for the GRE, 250 W for the 2080 Ti.

FAQ

Q: Which GPU is faster overall according to the database?

A: The RX 7900 GRE, with an average benchmark score of 32,456 versus 29,783, an 8-2 win record in head-to-head tests, and a 77th versus 75th percentile placement among all GPUs.

Q: Does the RTX 2080 Ti win anything?

A: Yes, two tests: Geekbench Vulkan (132,930 vs 99,850, a 24.9% lead) and PassMark DirectX 10 (153 vs 139, a 9.2% lead).

Q: How big is the gap in the GPU compute test?

A: Large. The 7900 GRE scores 15,016 in PassMark GPU compute against 10,056, a 49.3% advantage.

Q: Which card has more memory?

A: The 7900 GRE has 16 GB of GDDR6 on a 256-bit bus; the 2080 Ti has 11 GB on a 352-bit bus. The wider bus gives the 2080 Ti more bandwidth, 616.0 versus 576.0 GB/s.

Q: Are the power requirements similar?

A: Very similar: 260 W TDP for the GRE and 250 W for the 2080 Ti, both dual-slot, both using 2x 8-pin connectors, and both with a 600 W suggested PSU.

Q: Is either card still in production?

A: The 7900 GRE is listed as active; the RTX 2080 Ti is end-of-life.

Specification Differences

The meaningful differences cluster in five areas: silicon, throughput, memory, accelerators, and platform.

• Process and die: 5 nm, 57,700M transistors, 529 mm², 109.1M/mm² (GRE) vs 12 nm, 18,600M transistors, 754 mm², 24.7M/mm² (2080 Ti).
• Shading and output hardware: 5120 shading units, 320 TMUs, 160 ROPs, 80 RT cores vs 4352, 272, 88 and 68 respectively.
• Theoretical rates: 359.2 vs 136.0 GPixel/s; 718.4 vs 420.2 GTexel/s; 45.98 vs 13.45 TFLOPS FP32.
• Memory: 16 GB on a 256-bit bus at 576.0 GB/s vs 11 GB on a 352-bit bus at 616.0 GB/s.
• Accelerators: the 2080 Ti has 544 tensor cores; the GRE lists none.
• Platform: PCIe 4.0 x16, HDMI 2.1a, DisplayPort 2.1, USB Type-C vs PCIe 3.0 x16, HDMI 2.0, DisplayPort 1.4a, USB Type-C.
• Physical and power: the GRE is 276 mm long, 110 mm high and 51 mm wide at 260 W TDP; the 2080 Ti is 267 mm long, 116 mm high and 35 mm wide at 250 W TDP.
• Status and launch MSRP: the GRE launched at 549 USD and remains active; the 2080 Ti launched at 999 USD and is end-of-life.

Head-to-Head Benchmarks

The single most dramatic result is PassMark DirectX 11, where the 7900 GRE scores 300 against 190, a 57.9% win. Legacy DirectX 9 tells a similar story, 310 versus 235, a 31.9% margin, which is notable given how old that API is relative to both cards. Modern DirectX 12 also goes to the GRE, 107 to 84, by 27.4%.

Compute tests split by API in an interesting way. Geekbench OpenCL heavily favors the GRE, 175,758 to 128,171 (37.1%), and PassMark GPU compute agrees at 49.3% (15,016 vs 10,056). But Geekbench Vulkan flips the table: the 2080 Ti scores 132,930 versus 99,850, a 24.9% lead that stands out as the largest anomaly in the entire dataset. Why would a card losing badly in OpenCL and DirectX dominate Vulkan? The database cannot answer directly, but the 2080 Ti's superior memory bandwidth and its 544 tensor cores are the two hardware traits that distinguish it here.

3DMark Steel Nomad, a modern DirectX 12 stress test, favors the GRE 4814 to 3537, 36.1%, consistent with its rasterization superiority. PassMark 3D goes to the GRE as well, 27,089 versus 21,548 (25.7%), as does 2D graphics, 1180 versus 927 (27.3%). The 2080 Ti's only other win is PassMark DirectX 10, 153 to 139, a modest 9.2% edge in a legacy API.

Where Each One Wins

RX 7900 GRE territory: the data points here are unambiguous. Modern rasterized gaming (Steel Nomad, PassMark 3D, DirectX 12), legacy DirectX 9 and 11 gaming, general GPU compute under OpenCL, 2D workloads, and any scenario that benefits from 16 GB of memory or a PCIe 4.0 x16 host connection. Its 8-2 record, higher percentile placement, and active production status make it the default pick for gaming and compute outside Vulkan.

RTX 2080 Ti territory: Vulkan API work is its stronghold, with a 24.9% Geekbench Vulkan lead, and legacy DirectX 10 performance also leans its way. Workloads that lean on tensor cores, 544 of them versus none listed on the GRE, fall outside what the benchmark suite directly measures but represent the one hardware capability the older card holds exclusively. Its 616.0 GB/s of bandwidth also suits bandwidth-hungry tasks, even though the GRE's raw throughput wins most recorded tests.

The overall picture is one of generational progress: a newer process node, dramatically higher theoretical throughput, and a commanding win record for the 7900 GRE, with the 2080 Ti retaining two specific niches that its architecture still defends well.