Intel Arc B370 vs NVIDIA GeForce RTX 4070 AD103 Comparison

Head-to-Head Benchmarks

The recorded data contains only one benchmark result for the Intel Arc B370: a 3DMark Steel Nomad DX12 score of 1184. The NVIDIA GeForce RTX 4070 AD103 has no benchmark entries in the database, so a direct head-to-head comparison of measured scores is not possible. The Intel part places in the 5th percentile of all GPUs, while the NVIDIA part sits at the 50th percentile, a difference that reflects their respective positions in the performance distribution.

The Intel Arc B370's nearest rivals in the database provide context for its score. The ATI Mobility Radeon HD 5570 averages 1186, which is 0.2% higher than the B370. The ATI Radeon HD 5770 averages 1190, 0.5% ahead. The AMD Radeon HD 7650M averages 1192, 0.7% ahead. The AMD FirePro M2000 averages 1168, which is 1.4% behind the B370. These deltas are all within roughly one and a half percent, placing the Arc B370 in a tightly clustered performance band around the 1184 mark.

The NVIDIA GeForce RTX 4070 AD103 has no nearest rivals listed and no average benchmark score, so the database does not currently contain comparable measurements for it. Its percentile ranking of 50 indicates it sits at the median of all recorded GPUs, but without a numeric score, the magnitude of any performance gap cannot be stated from the data.

The compute throughput figures in the specification sheets do allow a partial comparison. The Intel Arc B370 delivers 6.144 TFLOPS of FP32 performance, while the NVIDIA GeForce RTX 4070 AD103 delivers 29.15 TFLOPS. The NVIDIA part provides roughly 4.7 times the FP32 throughput. In FP16, the Arc B370 reaches 12.29 TFLOPS using a 2:1 ratio, while the RTX 4070 AD103 reaches 29.15 TFLOPS at a 1:1 ratio, making the NVIDIA part approximately 2.4 times higher in FP16.

Pixel and texture rates follow a similar pattern. The Intel part outputs 48.00 GPixel/s and 96.00 GTexel/s. The NVIDIA part outputs 158.4 GPixel/s and 455.4 GTexel/s. The RTX 4070 AD103 delivers over three times the pixel fill rate and over four and a half times the texture fill rate. Rasterization hardware also differs: the Arc B370 has 1280 shading units, 40 texture mapping units, and 20 render output units, while the RTX 4070 AD103 has 5888 shading units, 184 TMUs, and 64 ROPs.

Ray tracing hardware shows a large gap as well. The Arc B370 includes 10 ray tracing cores, while the RTX 4070 AD103 includes 46. The NVIDIA part also includes 184 tensor cores, a feature class that the Intel part does not list. Both GPUs support DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, so API-level feature support is identical even though hardware capabilities differ substantially.

The Verdict

The data indicates two products aimed at entirely different segments. The Intel Arc B370 is an integrated graphics processor with a 25 W TDP, an IGP slot width, no power connectors, and system-shared memory. Its 5th percentile ranking and 1184 Steel Nomad score place it among much older discrete parts from the early 2010s, such as the ATI Radeon HD 5770 and the AMD Radeon HD 7650M. The NVIDIA GeForce RTX 4070 AD103 is a dual-slot discrete card with a 200 W TDP, a 16-pin power connector, a suggested 550 W power supply, and 12 GB of dedicated GDDR6X memory on a 192-bit bus. Its 50th percentile ranking places it at the median of all recorded GPUs, and its FP32, pixel rate, and texture rate figures are multiple times higher than those of the Arc B370 on every axis.

The recorded data supports only one conclusion: the RTX 4070 AD103 occupies a far higher performance tier. The Arc B370's nearest rivals all score within 1.4% of its 1184 result, and those rivals are legacy ATI and AMD parts. The RTX 4070 AD103 has no rival data to compare, but its specification sheet, particularly the 29.15 TFLOPS FP32 figure and 504.2 GB/s memory bandwidth, indicates a product that operates in a different class. The Arc B370's system-dependent memory bandwidth and system-shared memory further separate it from the dedicated 504.2 GB/s GDDR6X configuration of the NVIDIA part.

For users constrained to a 25 W integrated solution with no add-in card slot, the Arc B370 is the only applicable choice in this pairing. For users with a PCIe 4.0 x16 slot, a dual-slot cooler clearance, and a 550 W power supply, the RTX 4070 AD103 is the higher-performing option according to every comparable specification in the database. The production status also differs: the Arc B370 is listed as active, while the RTX 4070 AD103 is end-of-life, with a GeForce 50 successor already recorded.

FAQ

Q: How does the Intel Arc B370 score in 3DMark Steel Nomad DX12?

A: The Intel Arc B370 records a score of 1184 in 3DMark Steel Nomad DX12, which places it in the 5th percentile of all GPUs in the database.

Q: Does the NVIDIA GeForce RTX 4070 AD103 have a 3DMark Steel Nomad score?

A: No benchmark scores are recorded for the RTX 4070 AD103 in the database. Its average benchmark score is listed as zero, and it has no nearest rival data.

Q: What is the closest rival to the Intel Arc B370 in the database?

A: The ATI Mobility Radeon HD 5570 is the closest, with an average score of 1186, which is 0.2% higher than the Arc B370's 1184.

Q: How much FP32 compute does each GPU deliver?

A: The Intel Arc B370 delivers 6.144 TFLOPS of FP32, while the NVIDIA GeForce RTX 4070 AD103 delivers 29.15 TFLOPS.

Q: What memory configurations do the two GPUs use?

A: The Arc B370 uses system-shared memory with system-dependent bandwidth. The RTX 4070 AD103 uses 12 GB of GDDR6X on a 192-bit bus with 504.2 GB/s of bandwidth.

Q: What is the TDP difference between the two?

A: The Intel Arc B370 has a 25 W TDP. The NVIDIA GeForce RTX 4070 AD103 has a 200 W TDP.

Specification Differences

The two products differ on nearly every measurable specification. The Intel Arc B370 uses a 3 nm process at Intel foundry, while the NVIDIA GeForce RTX 4070 AD103 uses a 5 nm process at TSMC. The NVIDIA part lists 45,900 million transistors on a 379 mm² die with a transistor density of 121.1M per mm². The Intel part lists no transistor count, die size, or density figure.

Clock speeds differ substantially. The Arc B370 runs at a 300 MHz base clock and a 2400 MHz boost clock. The RTX 4070 AD103 runs at a 1920 MHz base clock and a 2475 MHz boost clock. The NVIDIA memory clock is 1313 MHz, or 21 Gbps effective, while the Intel memory clock is listed as system shared.

The memory subsystem is fundamentally different. The Arc B370 has no dedicated memory: size, type, bus width, and bandwidth are all system shared or system dependent. The RTX 4070 AD103 has 12 GB of GDDR6X, a 192-bit bus, and 504.2 GB/s of bandwidth.

Compute resources differ across every category. The Arc B370 has 1280 shading units, 40 TMUs, 20 ROPs, and 10 ray tracing cores. The RTX 4070 AD103 has 5888 shading units, 184 TMUs, 64 ROPs, 46 ray tracing cores, and 184 tensor cores. The Intel part lists no tensor cores.

Power and physical specifications diverge completely. The Arc B370 has a 25 W TDP, an IGP slot width, no power connectors, and no suggested PSU. The RTX 4070 AD103 has a 200 W TDP, a dual-slot width, one 16-pin power connector, and a suggested 550 W PSU. The Intel part uses an IGP bus interface, while the NVIDIA part uses PCIe 4.0 x16. The Arc B370's display outputs are portable device dependent; the RTX 4070 AD103 provides 1x HDMI 2.1 and 3x DisplayPort 1.4a. The RTX 4070 AD103 measures 240 mm by 110 mm by 40 mm; the Arc B370 lists no dimensions.

Release dates differ by roughly two years. The Arc B370 released on 2026-01-26, and the RTX 4070 AD103 released on 2024-02-29. The NVIDIA part is end-of-life with a predecessor in GeForce 30 and a successor in GeForce 50. The Intel part is active with no predecessor or successor listed. The RTX 4070 AD103 has a launch MSRP of 599 USD.

Architecture Differences

The Intel Arc B370 is built on the Xe3-LPG architecture, part of the Arc Graphics-M generation for Panther Lake. The NVIDIA GeForce RTX 4070 AD103 is built on the Ada Lovelace architecture, part of the GeForce 40 series. The manufacturing processes differ: Intel uses a 3 nm node at its own foundry, while NVIDIA uses a 5 nm node at TSMC.

The core configurations reflect different design philosophies. The Arc B370 uses 1280 shading units, 40 TMUs, and 20 ROPs, with 10 ray tracing cores and no tensor cores. The RTX 4070 AD103 uses 5888 shading units, 184 TMUs, and 64 ROPs, with 46 ray tracing cores and 184 tensor cores. The shading unit count on the NVIDIA part is 4.6 times higher, the TMU count is 4.6 times higher, the ROP count is 3.2 times higher, and the ray tracing core count is 4.6 times higher.

The FP16 execution path differs as well. The Arc B370 achieves 12.29 TFLOPS of FP16 through a 2:1 ratio, indicating that its FP16 throughput is derived from the FP32 hardware. The RTX 4070 AD103 achieves 29.15 TFLOPS of FP16 at a 1:1 ratio, matching its FP32 figure, which indicates dedicated or equally provisioned FP16 hardware. The Intel part also lacks tensor cores, while the NVIDIA part includes 184 of them, a difference relevant to any workload that uses tensor operations.

Both GPUs support identical API levels: DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The architectural differences therefore manifest in hardware resources rather than API feature support. The Intel part's 3 nm process and 25 W TDP indicate a design focused on integration and low power draw, while the NVIDIA part's 5 nm process and 200 W TDP indicate a design focused on maximum throughput in a discrete card.

Memory architecture is another major divider. The Arc B370 has no dedicated memory interface; it relies on system-shared memory, which makes its bandwidth system dependent. The RTX 4070 AD103 has a dedicated 192-bit GDDR6X interface with 504.2 GB/s of bandwidth, a figure that does not depend on the host system's memory configuration.

Where Each One Wins

The Intel Arc B370 wins in power efficiency and physical integration. Its 25 W TDP is the lowest recorded figure in this comparison, and its IGP slot width with no power connectors means it requires no additional cooling or power delivery beyond the host device. The 3 nm process node is smaller than the 5 nm node used by the NVIDIA part. The Arc B370 is also active in production status, while the RTX 4070 AD103 is end-of-life. For portable devices where discrete graphics cannot be installed, the Arc B370 is the only viable option in this pairing.

The NVIDIA GeForce RTX 4070 AD103 wins in every raw performance metric recorded. Its FP32 throughput of 29.15 TFLOPS is roughly 4.7 times the Arc B370's 6.144 TFLOPS. Its FP16 throughput of 29.15 TFLOPS is about 2.4 times the Arc B370's 12.29 TFLOPS. Its pixel rate of 158.4 GPixel/s is 3.3 times the Arc B370's 48.00 GPixel/s. Its texture rate of 455.4 GTexel/s is 4.7 times the Arc B370's 96.00 GTexel/s. Its 46 ray tracing cores dwarf the Arc B370's 10, and its 184 tensor cores have no counterpart on the Intel part.

The RTX 4070 AD103 also wins on dedicated memory resources. Its 12 GB of GDDR6X on a 192-bit bus delivers 504.2 GB/s of bandwidth, while the Arc B370's memory bandwidth is system dependent and therefore variable. The NVIDIA part's 5888 shading units, 184 TMUs, and 64 ROPs provide a large resource advantage over the Arc B370's 1280 shading units, 40 TMUs, and 20 ROPs.

The percentile data reinforces the split. The Arc B370 sits at the 5th percentile of all GPUs, with nearest rivals that are over a decade old. The RTX 4070 AD103 sits at the 50th percentile, the median of all recorded GPUs. The benchmark results, where they exist, show the Arc B370 clustered within 1.4% of legacy ATI and AMD parts, while the RTX 4070 AD103 has no recorded score to anchor its position numerically. The specification data, however, consistently indicates that the RTX 4070 AD103 operates at a performance level the Arc B370 cannot approach.

The use-case split is therefore straightforward. The Arc B370 suits systems that require an integrated GPU with a 25 W power envelope, no add-in card slot, and no dedicated memory. The RTX 4070 AD103 suits systems with a PCIe 4.0 x16 slot, a dual-slot cooler, a 550 W power supply, and workloads that benefit from high FP32 and FP16 throughput, dedicated GDDR6X bandwidth, ray tracing cores, and tensor cores. The database contains no recorded benchmark score for the RTX 4070 AD103, but every specification field that can be compared favors the NVIDIA part.