AMD Ryzen 5 PRO 8640HS vs Intel Arc G3 EXTREME Comparison

The Verdict

The recorded benchmark data presents a clear split between these two mobile processors. The Intel Arc G3 EXTREME wins 15 of the 17 head-to-head comparisons, while the AMD Ryzen 5 PRO 8640HS takes only 2 wins. The average benchmark score confirms the gap: the Intel part scores 36,699 versus 28,478 for the AMD part, a difference that places Intel at the 85th percentile of all CPUs while AMD sits at the 80th percentile.

The Intel Arc G3 EXTREME delivers its strongest advantages in compute-heavy workloads. PassMark floating point math shows a 50.5% lead, physics tests show a 58.5% lead, and data encryption shows a 37.3% lead. The AMD Ryzen 5 PRO 8640HS fights back in Cinebench R23, where it leads by 23.5% in multi-core and 37.2% in single-core. These are not small margins; they represent fundamentally different performance profiles.

The data indicates that the Intel part is the stronger choice for general throughput, mathematical workloads, and compression tasks. The AMD part, however, holds a decisive edge in the newer Cinebench R23 workload, which suggests its architecture handles that specific rendering test more efficiently. The choice between the two depends entirely on which workload family matters more.

Architecture Differences

The two processors come from different manufacturing and design philosophies. The AMD Ryzen 5 PRO 8640HS uses Zen 4 architecture under the Hawk Point codename, built on a 4 nm process at TSMC with 25,000 million transistors on a 178 mm² die. The Intel Arc G3 EXTREME uses the Panther Lake codename with a 3 nm process at Intel's own foundry, though the database records no transistor count or die size for it.

Core configurations differ substantially. The AMD part has 6 cores and 12 threads, while the Intel part has 14 cores and 14 threads. The Intel chip has no hyperthreading, which explains the equal core and thread counts. The AMD chip relies on simultaneous multithreading to reach 12 threads from 6 cores.

Cache hierarchies diverge at every level. The AMD processor allocates 64 KB of L1 per core, 1 MB of L2 per core, and 16 MB of shared L3. The Intel processor allocates 192 KB of L1 per core, 2.5 MB of L2 per core, and 18 MB of shared L3. The larger per-core L1 and L2 caches on the Intel side likely contribute to its strengths in certain workloads.

Memory support also differs. AMD uses DDR5 with a dual-channel bus and 89.6 GB/s of bandwidth. Intel uses LPDDR5X with a dual-channel bus and 136.5 GB/s of bandwidth, a 52.3% higher memory bandwidth figure. ECC memory support exists on the AMD part but not on the Intel part. PCIe connectivity favors AMD with Gen 4 and 20 lanes, while Intel offers Gen 5 but only 4 lanes.

The integrated graphics differ as well. AMD pairs its CPU with the Radeon 760M, while Intel uses the Arc B390. The Intel part carries a 3 nm process advantage, which typically enables higher density and efficiency, though the recorded data does not include power efficiency measurements.

Head-to-Head Benchmarks

Cinebench R23 produces the most striking divergence. The AMD Ryzen 5 PRO 8640HS scores 18,104 in multi-core and 2,555 in single-core, while the Intel Arc G3 EXTREME scores 14,655 and 1,862 respectively. The AMD leads by 23.5% in multi-core and 37.2% in single-core. These are the only two wins for AMD in the entire comparison.

The earlier Cinebench versions tell the opposite story. In R15 multi-core, Intel wins with 2,192 versus 1,824, a 16.8% lead. In R15 single-core, Intel wins with 267 versus 257, a 3.7% lead. In R20, Intel extends its advantage to 30.5% in multi-core (10,945 versus 7,603) and 30.6% in single-core (1,545 versus 1,073). The shift between R20 and R23 is dramatic: Intel dominates the older tests while AMD dominates the newest one.

PassMark results consistently favor Intel. Floating point math shows the largest margin at 86,929 versus 43,019, a 50.5% lead. Physics scores 2,515 versus 1,043, a 58.5% lead. Find prime numbers shows 248 versus 71, a 71.4% lead. Data encryption reaches 23,881 versus 14,962, a 37.3% lead. Extended instructions score 24,017 versus 18,507, a 22.9% lead. Data compression sits at 307,094 versus 246,446, a 19.7% lead.

Smaller Intel wins appear in integer math (71,164 versus 69,839, only 1.9% ahead), random string sorting (37,331 versus 30,235, 19% ahead), and multi-thread PassMark (30,659 versus 21,465, 30% ahead). The single-thread PassMark result gives Intel 4,293 versus 3,561, a 17.1% lead. The database shows two identical single-thread entries for each processor, confirming consistency in those measurements.

Specification Differences

The recorded specifications show several fields where the two processors differ. Clock speeds favor AMD on paper: 3.50 GHz base and 4.90 GHz boost versus 1.90 GHz base and 4.70 GHz boost for Intel. Despite the lower clocks, Intel wins most benchmarks, which points to architectural efficiency and the higher core count as decisive factors.

Thermal design power favors Intel at 25 watts versus 28 watts for AMD. Both are mobile parts with active production status. Power draw does not explain the performance differences, since Intel delivers higher scores at a lower TDP.

The socket and platform differ completely. AMD uses Socket FP7, while Intel uses BGA 2540. Neither processor has an unlocked multiplier. The AMD part numbers include 100-000001354 (FP7r2) and 100-000001382 (FP7), while the Intel part number is SA4R3.

Release dates show a significant gap. AMD launched on 2024-04-15, while Intel launched on 2026-05-27. The database records no launch MSRP for either part. The Intel processor belongs to the Panther Lake generation, while AMD belongs to the 8000 series with Zen 4 (Hawk Point). The market segment for both is mobile, and both are currently in active production.

FAQ

Q: Which processor has more cores?

A: The Intel Arc G3 EXTREME has 14 cores and 14 threads, while the AMD Ryzen 5 PRO 8640HS has 6 cores and 12 threads.

Q: Does the Intel processor support ECC memory?

A: No. The database shows ECC memory support as false for the Intel Arc G3 EXTREME. The AMD Ryzen 5 PRO 8640HS supports ECC memory.

Q: Which processor wins in Cinebench R23?

A: The AMD Ryzen 5 PRO 8640HS wins both Cinebench R23 tests. It scores 18,104 in multi-core versus 14,655 for Intel (23.5% lead) and 2,555 in single-core versus 1,862 (37.2% lead).

Q: What is the memory bandwidth difference?

A: The Intel Arc G3 EXTREME provides 136.5 GB/s with LPDDR5X memory, while the AMD Ryzen 5 PRO 8640HS provides 89.6 GB/s with DDR5 memory.

Q: Which processor has a higher average benchmark score?

A: The Intel Arc G3 EXTREME has an average benchmark score of 36,699, compared to 28,478 for the AMD Ryzen 5 PRO 8640HS. Intel also ranks at the 85th percentile versus AMD's 80th percentile.

Q: What is the process node for each processor?

A: The AMD Ryzen 5 PRO 8640HS uses a 4 nm process at TSMC. The Intel Arc G3 EXTREME uses a 3 nm process at Intel.

Where Each One Wins

The Intel Arc G3 EXTREME dominates across the PassMark suite and the older Cinebench versions. Its 14 cores provide a clear multi-threading advantage in data compression, where it scores 307,094, and in multi-thread PassMark, where it reaches 30,659. The physics test result of 2,515 versus 1,043 shows a 58.5% lead that suggests strong compute throughput for simulation-style workloads. Floating point math at 86,929 versus 43,019 indicates the Intel part handles numerical processing much more effectively. The 71.4% lead in find prime numbers (248 versus 71) reinforces this pattern.

The AMD Ryzen 5 PRO 8640HS wins specifically in Cinebench R23. The 37.2% single-core lead (2,555 versus 1,862) and 23.5% multi-core lead (18,104 versus 14,655) show that this architecture performs well on that rendering workload. Users running Cinebench R23 would see substantially better scores on the AMD part. The higher boost clock of 4.90 GHz versus 4.70 GHz may contribute to this advantage, though the data does not isolate that variable.

For general productivity, the Intel part appears stronger based on the 30% multi-thread PassMark lead and the 19.7% data compression lead. For encryption workloads, Intel holds a 37.3% advantage. For single-threaded PassMark tasks, Intel leads by 17.1%. The only notable AMD wins outside Cinebench R23 do not appear in the recorded data, which means the Intel part holds the advantage in every other measured workload.

The database reveals an interesting nuance in the Cinebench progression. Intel wins R15 and R20 by wide margins, then loses R23 by equally wide margins. This could indicate workload-specific optimization or changes in how the newer test stresses the processors. The AMD chip's 12 threads on 6 cores may interact differently with the R23 workload than the Intel chip's 14 threads on 14 cores.

The percentile data places both processors in the upper range of all CPUs, with Intel at the 85th percentile and AMD at the 80th. The nearest rivals for each confirm their competitive positions. Intel's closest competitor, the Intel Core Ultra 5 225, scores 36,938 with a 0.6% delta. AMD's closest rival, the Intel Xeon E-2436, scores 28,530 with a 0.2% delta. These figures place each processor in a distinct performance tier.