AMD Ryzen AI 7 450G vs Intel Xeon 636 Comparison

The AMD Ryzen AI 7 450G and Intel Xeon 636 are two Active-market processors released within weeks of each other in early 2026, yet they land in opposite corners of the market: a 65 W desktop part with integrated graphics on AMD's Socket AM5, and a 170 W Server/Workstation chip on Intel's Socket 4710 with no integrated GPU. The recorded data shows a clean split, with the Xeon taking 9 of 11 head-to-head benchmarks while the Ryzen AI 7 450G holds both single-thread results by a consistent margin. Both sit near the top of the database overall, at the 93rd and 92nd percentiles versus all CPUs.

The Verdict

The data points to a straightforward division. Buyers who need throughput should choose the Intel Xeon 636: its 12 cores and 24 threads deliver a 31.5 percent advantage in PassMark multithread, and it wins every recorded multi-core workload, from data encryption (30.4 percent ahead) to physics simulation (60.5 percent ahead). Its quad-channel DDR5 platform with 204.8 GB/s of memory bandwidth and 80 PCIe Gen 5 lanes makes it the clear fit for memory-hungry and expansion-heavy workstation builds, and its launch MSRP is $639.

Buyers who prioritize responsiveness and efficiency should choose the AMD Ryzen AI 7 450G. It wins single-thread performance by 9.4 percent, 4309 to 3937, which matters for latency-sensitive desktop software, and it does so at a TDP of 65 W versus the Xeon's 170 W. It is also the only one of the two with integrated graphics, the Radeon 860M, meaning it can run without a discrete GPU. Its average benchmark score of 63331 slightly exceeds the Xeon's 61360, though that aggregate masks the direction of individual results.

Notably, neither chip dominates its own neighborhood. The Ryzen AI 7 450G's nearest rivals include the Intel Core Ultra 7 265HX (within 0.2 percent), the Intel Core i7-13790F (within 0.4 percent), and the AMD Ryzen AI Max PRO 390, which is 0.7 percent ahead. The Xeon 636 trails all four of its nearest rivals: the Intel Core i9-13900T, i9-13900K, i9-13900KF, and Core Ultra 7 270HX Plus each edge it by roughly 0.6 to 0.8 percent on average score.

Architecture Differences

These processors take very different routes to adjacent percentile positions. The Ryzen AI 7 450G is built on TSMC's 4 nm process, packs 8 cores and 16 threads in a hybrid Zen 5 / Zen 5c design under the Gorgon Point codename, and belongs to the Ryzen AI 400 generation. Its frequency range is wide: a 2.00 base clock that boosts to 5.10, the highest single number on this page. Cache is modest, with 80 KB of L1 per core, 1 MB of L2 per core, and 8 MB of L3. The die measures 195 mm².

The Xeon 636 is a Granite Rapids design on Intel's own 5 nm node, part of the Xeon 600 generation (Granite Rapids-WS). It runs a tighter 3.50 base and 4.70 boost range across 12 cores and 24 threads. Cache is far more generous: 112 KB of L1 per core, 2 MB of L2 per core, and 48 MB of shared L3, six times the L3 of the AMD part. The die is more than three times larger at 598 mm², a physical reflection of its Server/Workstation positioning.

Platform support diverges sharply. The AMD chip accepts DDR5 or LPDDR5X over a dual-channel bus at 89.6 GB/s, with PCIe Gen 4 at 12 CPU lanes. The Intel chip is DDR5 only, but quad-channel at 204.8 GB/s, more than double the bandwidth, with PCIe Gen 5 at 80 CPU lanes. Both support ECC memory and both ship with unlocked multipliers.

Where Each One Wins

The Xeon 636 owns every parallel workload in the database. Its strengths scale with thread count: the gap widens from 21.6 percent in random string sorting to 40.9 percent in floating point math and 60.5 percent in physics. Find prime numbers, a pure number-crunching test, shows the largest difference at 65.9 percent. Rendering results reinforce this, with Cinebench R23 multi-core at 37757 and R20 multi-core at 15857 recorded for the Xeon. Any workload that scales across cores, or that feeds on memory bandwidth and PCIe lanes, belongs on the Intel side.

The Ryzen AI 7 450G wins where clock speed and per-core throughput dominate. Its 5.10 boost against the Xeon's 4.70 translates directly into the 9.4 percent single-thread lead, 4309 versus 3937. It is also the only option here that can drive a display without a discrete card, thanks to the Radeon 860M. For a compact desktop that must stay cool on a 65 W envelope, the recorded data suggests the AMD part is the practical pick.

An interesting wrinkle: despite losing 9 of 11 tests, the AMD chip posts the higher average benchmark score, 63331 versus 61360, and the higher percentile, 93 versus 92. This implies the single-thread and aggregate metrics weight its strengths differently than the raw head-to-head list does.

FAQ

**Q: Which CPU is faster overall in the head-to-head benchmarks?

A: The Intel Xeon 636 wins 9 of 11 recorded tests, including a 31.5 percent victory in PassMark multithread (44421 versus 30422).

**Q: Which CPU is faster in single-threaded work?

A: The AMD Ryzen AI 7 450G, scoring 4309 against the Xeon 636's 3937 in PassMark single thread, a 9.4 percent lead.

**Q: Which one has integrated graphics?

A: Only the AMD Ryzen AI 7 450G, via its Radeon 860M. The Xeon 636 is listed with no integrated graphics.

**Q: How do their memory platforms differ?

A: The AMD supports DDR5 and LPDDR5X on dual-channel at 89.6 GB/s. The Intel supports DDR5 only, but quad-channel at 204.8 GB/s. Both support ECC.

**Q: What is the difference in power draw?

A: The Ryzen AI 7 450G has a TDP of 65 W, while the Xeon 636 is rated at 170 W.

**Q: What was the Xeon 636's launch MSRP?

A: $639. No launch MSRP is recorded for the Ryzen AI 7 450G.

Head-to-Head Benchmarks

The pattern across the recorded tests is remarkably consistent: the Xeon wins everything that touches multiple threads, and the margin grows with how parallel the workload is.

• Data compression: 550395 for the Xeon against 402831 for the Ryzen, a 26.8 percent gap.
• Data encryption: 27193 versus 18937, Xeon by 30.4 percent.
• Extended instructions: 43282 versus 28223, Xeon by 34.8 percent.
• Find prime numbers: 255 versus 87, the largest recorded gap at 65.9 percent.
• Floating point math: 107826 versus 63683, Xeon by 40.9 percent.
• Integer math: 138281 versus 99732, Xeon by 27.9 percent.
• Multithread: 44421 versus 30422, Xeon by 31.5 percent.
• Physics: 3645 versus 1440, Xeon by 60.5 percent.
• Random string sorting: 54420 versus 42663, the closest multi-core test at 21.6 percent.
• Single thread: 4309 versus 3937, Ryzen by 9.4 percent, the AMD part's lone win.

The physics result invites the most curiosity. A 60.5 percent gap on a 50 percent core-count advantage suggests the Xeon extracts more than proportional scaling in that test, likely aided by its bandwidth headroom. The prime number result, where the Xeon scores nearly three times the AMD figure, points the same direction. Meanwhile the single-thread results hint that clock ceiling, not architecture, decides lightly threaded work, since the 5.10 GHz part beats the 4.70 GHz part by almost exactly what the frequency difference would imply.

For context against the wider database, the Xeon's Cinebench results (R15 multi 3805, R15 single 537, R20 multi 15857, R20 single 2238, R23 multi 37757, R23 single 5330) place it solidly in workstation territory, and its rival set of Core i9-13900 variants all sit within roughly a point of its average.

Specification Differences

• Cores/Threads: 12/24 (Xeon 636) versus 8/16 (Ryzen AI 7 450G)
• Base Clock: 3.50 GHz (Xeon) versus 2.00 GHz (AMD)
• Boost Clock: 5.10 GHz (AMD) versus 4.70 GHz (Xeon)
• TDP: 65 W (AMD) versus 170 W (Intel)
• Socket: AMD Socket AM5 versus Intel Socket 4710
• Process Node: 4 nm TSMC (AMD) versus 5 nm Intel (Intel)
• Die Size: 195 mm² (AMD) versus 598 mm² (Intel)
• L1 Cache: 80 KB per core (AMD) versus 112 KB per core (Intel)
• L2 Cache: 1 MB per core (AMD) versus 2 MB per core (Intel)
• L3 Cache: 8 MB (AMD) versus 48 MB shared (Intel)
• Memory Support: DDR5 and LPDDR5X (AMD) versus DDR5 only (Intel)
• Memory Bus: Dual-channel (AMD) versus Quad-channel (Intel)
• Memory Bandwidth: 89.6 GB/s (AMD) versus 204.8 GB/s (Intel)
• PCIe: Gen 4, 12 CPU lanes (AMD) versus Gen 5, 80 CPU lanes (Intel)
• Integrated Graphics: Radeon 860M (AMD) versus none (Intel)
• Market Segment: Desktop (AMD) versus Server/Workstation (Intel)
• Release Date: 2026-02-28 (AMD) versus 2026-02-01 (Intel)
• Launch MSRP: none recorded (AMD) versus $639 (Intel)
• Part Number: 100-000001917 (AMD) versus SA2DM (Intel)

Both share ECC support, unlocked multipliers, and Active production status, which means the decision ultimately rests on the workload split the benchmarks describe: threads and bandwidth on one side, clocks and efficiency on the other.