AMD Radeon RX 7700S vs Intel Arc B390 Comparison

Head-to-Head Benchmarks

The only directly comparable benchmark in the database is 3DMark Steel Nomad DX12. The AMD Radeon RX 7700S records a score of 2185, while the Intel Arc B390 scores 1482. This results in a 47.4% advantage for the AMD part, the sole decisive victory in the head-to-head comparison.

AMD also holds a significant lead in raw compute metrics. The RX 7700S delivers 20.48 TFLOPS of FP32 throughput and 40.96 TFLOPS of FP16, compared to 7.680 TFLOPS and 15.36 TFLOPS respectively for the Arc B390. That is roughly 2.7 times the FP32 output, a gap that aligns with the benchmark delta.

The Arc B390 does not win any recorded test. Its only submitted benchmark, the same Steel Nomad run, places it at the 9th percentile of all GPUs in the database. The RX 7700S, by contrast, sits at the 78th percentile. The average benchmark score tells a similar story: 33849 for AMD versus 1482 for Intel.

The RX 7700S also shows higher pixel and texture throughput. It reaches 160.0 GPixel/s and 320.0 GTexel/s, while the Arc B390 manages 60.00 GPixel/s and 120.0 GTexel/s. These figures matter for fill-rate-bound workloads, and the AMD part leads by more than 2.5 times in both.

Memory bandwidth is another clear separator. The RX 7700S uses 8 GB of GDDR6 on a 128-bit bus, yielding 288.0 GB/s. The Arc B390 relies on system-shared memory, with bandwidth listed as System Dependent. That dependency means the Intel part's real-world throughput varies with the host platform, but the recorded specifications show no dedicated memory pipeline to compete with the AMD solution.

The average benchmark score for the RX 7700S, 33849, places it near a cluster of older discrete GPUs. Its nearest rivals include the AMD Radeon HD 7950 at 33951 (0.3% ahead), the AMD Radeon RX 480 at 33997 (0.4% ahead), the NVIDIA GeForce GTX 1060 5 GB at 33694 (0.5% behind), and the NVIDIA RTX A5000 at 33622 (0.7% behind). These deltas are all within one percentage point, meaning the RX 7700S performs in line with that generation of desktop cards.

The Arc B390's nearest rivals are far less competitive. The NVIDIA GeForce GT 520MX scores 1463 (1.3% behind), the NVIDIA GeForce 800M scores 1460 (1.5% behind), the NVIDIA GeForce GT 625 OEM scores 1446 (2.5% behind), and the NVIDIA GeForce GT 710 scores 1443 (2.7% behind). The Intel part leads this group but only by a few points, a stark contrast to the AMD card's positioning.

Architecture Differences

The two GPUs come from different foundries and process nodes. AMD uses TSMC's 6 nm node for the Navi 33 chip, part of the RDNA 3.0 architecture with the codename Hotpink Bonefish. Intel builds the Arc B390 on its own 3 nm node, using the Xe3-LPG architecture with the Panther Lake chip. The RX 7700S belongs to the Navi Mobile (RX 7000M) generation, while the Arc B390 is part of Arc Graphics-M (Panther Lake).

Transistor counts are only available for one side. The RX 7700S integrates 13,300 million transistors on a 204 mm² die, giving a density of 65.2M per mm². Intel lists the Arc B390's transistor count and die size as unknown. The RX 7700S also has a much larger silicon budget, which helps explain its higher compute throughput.

Shader resources differ substantially. The RX 7700S has 2048 shading units, 128 texture mapping units, and 64 raster operations units. The Arc B390 has 1536 shading units, 48 TMUs, and 24 ROPs. The AMD part carries more than double the TMUs and ROPs, which directly supports its higher texture and pixel rates. Ray tracing hardware also favors AMD: 32 RT cores versus 12.

Clock behavior is interesting because the base clocks are very different. The RX 7700S runs at 1500 MHz base and 2500 MHz boost, with a 2200 MHz game clock. The Arc B390 has a 300 MHz base and 2500 MHz boost. The Intel part's low base clock suggests it relies on boost behavior for performance, likely to keep idle power low. Both reach the same 2500 MHz boost ceiling. Memory clock for the RX 7700S is 2250 MHz, or 18 Gbps effective; the Arc B390 uses system-shared memory with no dedicated clock listed.

Power targets differ as well. The RX 7700S has a TDP of 100 W, while the Arc B390 is rated at 80 W. Both are integrated form factors (IGP), use no power connectors, and are portable-device dependent for display outputs. The AMD part pulls more power but delivers substantially higher performance per the benchmark data.

API support is identical: both list DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The bus interfaces differ, with the RX 7700S using PCIe 4.0 x16 and the Arc B390 using IGP. This reflects the Intel part's tighter integration with Panther Lake.

Where Each One Wins

The RX 7700S wins in every measured category. Its 47.4% lead in Steel Nomad DX12 is the most direct evidence. In synthetic compute, the AMD part's 20.48 TFLOPS FP32 and 40.96 TFLOPS FP16 crush the Arc B390's 7.680 and 15.36 TFLOPS. For pixel-heavy workloads, the 160.0 GPixel/s pixel rate and 320.0 GTexel/s texture rate leave the Intel part's 60.00 GPixel/s and 120.0 GTexel/s far behind. The dedicated 288.0 GB/s memory bandwidth also gives AMD an edge in bandwidth-sensitive scenarios.

The Arc B390's advantages are narrower. It uses less power at 80 W versus 100 W, and it relies on system-shared memory, which can simplify system design. Its 3 nm node suggests a more modern process, though the database records no performance benefit from that node. The Arc B390 also has a lower base clock of 300 MHz, which could help with idle efficiency, but no benchmark data confirms this.

In terms of GPU positioning, the RX 7700S sits near mid-range desktop cards from the Pascal and Polaris era, within one percentage point of the GTX 1060 5 GB, RX 480, and HD 7950. The Arc B390 sits near entry-level mobile GPUs from a decade ago, within a few points of the GT 710 and GT 520MX. The percentile difference, 78 versus 9, captures this gap clearly.

For users looking at portable devices, the RX 7700S offers a complete memory subsystem and much higher throughput. The Arc B390's shared memory means its performance depends on the host system's memory configuration, which the database lists as System Dependent. That introduces variability that the AMD part does not have.

FAQ

Q: Which GPU has the higher 3DMark Steel Nomad DX12 score?

A: The AMD Radeon RX 7700S scores 2185, which is 47.4% higher than the Intel Arc B390's 1482.

Q: How do the two GPUs compare in raw compute power?

A: The RX 7700S delivers 20.48 TFLOPS FP32 and 40.96 TFLOPS FP16. The Arc B390 delivers 7.680 TFLOPS FP32 and 15.36 TFLOPS FP16.

Q: What memory configurations do they use?

A: The RX 7700S has 8 GB of GDDR6 on a 128-bit bus with 288.0 GB/s bandwidth. The Arc B390 uses system-shared memory with bandwidth listed as System Dependent.

Q: Which GPU has better rasterization throughput?

A: The RX 7700S reaches 160.0 GPixel/s and 320.0 GTexel/s. The Arc B390 reaches 60.00 GPixel/s and 120.0 GTexel/s.

Q: How do their power requirements differ?

A: The RX 7700S has a TDP of 100 W. The Arc B390 has a TDP of 80 W.

Q: Where does each GPU rank among all GPUs in the database?

A: The RX 7700S is at the 78th percentile with an average benchmark score of 33849. The Arc B390 is at the 9th percentile with an average score of 1482.

The Verdict

The data gives a clear answer. The AMD Radeon RX 7700S is the faster GPU by a wide margin. It wins the only head-to-head benchmark by 47.4%, offers nearly three times the FP32 throughput, more than double the pixel rate, and a dedicated memory bus with 288.0 GB/s of bandwidth. Its 78th percentile ranking places it alongside capable desktop GPUs like the GTX 1060 5 GB and RX 480, all within one percentage point of the RX 7700S's average score.

The Intel Arc B390 is not competitive in this comparison. Its 9th percentile ranking and average score of 1482 put it in the company of GPUs like the GT 710 and GT 520MX. It does consume 20 W less power and uses a 3 nm process, but the recorded performance data shows no benefit from those traits. The shared memory design adds dependency on the host system, and its nearest rivals are all within 2.7% of its score.

The choice depends on the platform. For a portable device that must handle modern DX12 workloads at reasonable settings, the RX 7700S is the only viable option based on the benchmark evidence. The Arc B390 may suit a system where power efficiency and integration with Panther Lake matter more than graphics performance, but the database shows no scenario where it wins a measured test. Users prioritizing graphics capability should select the RX 7700S. Users prioritizing lower power draw at the cost of performance may consider the Arc B390, but the benchmark records do not support any performance-oriented use case for it.