The AMD Ryzen 7 4800U is a processor for thin and light laptops based on the Renoir architecture. The 4800U integrates all eight cores based on the Zen 2 microarchitecture. They are clocked at 1.8 (guaranteed base clock) to 4.2 GHz (Turbo) and support SMT / Hyperthreading (16 threads). The chip is manufactured in the modern 7 nm process at TSMC and partly thanks to it AMD advertises a 2x improved performance per Watt for the Renoir chips. Compared to the similar named Ryzen 7 4800H, the 4800U offers a lower TDP (15 versus 45 Watt) and base clock (1.8 versus 2.9 GHz).
According to AMD, the Ryzen 7 4800U is 4% faster in Cinebench R20 single thread test and 90 % faster in the multithreaded test compared to the fast Intel Core i7-1065G7 (Ice Lake, 4 cores, 3.9 GHz). This means the CPU should be one of the fastest processors in the 15 Watt TDP range and well suited for demanding tasks. The sustained performance however ist strongly depending on the cooling system and the TDP settings of the laptop.
In addition to the eight CPU cores, the APU also integrates a Radeon RX Vega 8 integrated graphics card with 8 CUs and up to 1750 MHz. The dual channel memory controller supports DDR4-3200 and energy efficient LPDDR4-4266 RAM. Furthermore, 8 MB level 3 cache can be found on the chip. See our hub page on the Renoir Processors for more information.
The TDP of the APU is specified at 15 Watt (default) and can be configured from 10 to 25 Watt by the laptop vendor. That means the chip is intended for thin and light laptops (but with fans).
The Intel Celeron N4020 is a slow dual-core processor (SoC) of the Gemini Lake refresh product family; as such, it is designed for use in laptops and mini-PCs of the most affordable flavor. The Celeron was launched in Q4 2019. Its CPU cores run at 1.1 GHz to 2.8 GHz, the latter being a moderate 200 MHz upgrade over the preceding N4000 chip. Other key specs include a DDR4/LPDDR4 memory controller (up to 2,400 MHz and up to 8 GB, with independent reports claiming as much as 32 GB will work just fine) and an integrated UHD 600 graphics adapter.
Architecture & Features
Just like Apollo Lake family products, the ever-popular N3350 included, the N4020 is manufactured on a really old, as of late 2023, 14 nm process. What makes the newer Celerons different are the slightly improved processor cores with double the L2 cache and also, somewhat counterintuitively, a reduction in physical size.
The Goldmont Plus microarchitecture is not much different from what was used in Gemini Lake processors like the N4000. A relatively large 4 MB L2 cache features prominently on the rather short list of N4020's strong sides, allowing for a marginal increase in performance-per-MHz figures compared to processors of previous generations. Still, Gemini Lake refresh processors are a clear step down from Core i3/i5/i7/i9 series processors, both in performance and in features.
The N4020 has six PCI-Express 2.0 lanes at its disposal. While very few N4020-based SBCs, nettops and laptops feature an NVMe M.2 slot, you can use an NVMe SSD as a boot drive with this processor (read/write rates will be limited to 2 GB/s, though). Furthermore, partial Wi-Fi 5 support is built into the CPU. The Celeron also supports up to eight USB 3.0 ports and two SATA III storage devices.
Please note this is not a user-replaceable CPU. They solder it straight to the motherboard for good (BGA1090 socket interface).
Performance
While slightly faster than the outgoing Celeron N4000, the average N4020 in our extensive database only just manages to match the N6211, as far as multi-thread performance is concerned. These three chips deliver multi-thread CB R15, CB R20 and CB R23 scores that are so low, they lag behind a single-thread score of any half-decent CPU such as an i5-1135G7. In other words, these Celeron N chips are good enough for basic tasks only such as word processing and Web browsing with two or three tabs open at a time.
The Celeron N4120, a quad-core chip with a similar name, has little trouble leaving the N4020 behind in most workloads - which is not to say it is a fast CPU.
Jacking the long-term power limit value up to something like 9 W will help improve system responsiveness noticeably.
Graphics
The UHD Graphics 600 is based on Intel's Generation 9 architecture, much like the HD Graphics 520 or the UHD Graphics 615 or other widespread Intel iGPUs found in Core i3/i5/i7/i9 processors of generations six to ten.
Just like the HD Graphics 500, the UHD Graphics 600 is DX12 compatible. The iGPU's 12 EUs can run at up to 650 MHz. The Iris Plus G7 iGPU that certain 10th Gen Ice Lake processors have packs 64 EUs, for reference. As a low-end solution, UHD Graphics 600 will let you play some seriously old titles, but that's about as far as its talents go.
Perhaps more importantly, this graphics solution can drive up to 3 monitors with resolutions as high as 4096x2160@60. Furthermore, it will have no trouble HW-decoding AVC, HEVC and VP9-encoded videos. The newer AV1 codec will be decoded via software, with the limited CPU horsepower imposing a limit on video resolutions that can be played back without stuttering. 1080p60 videos are out of reach while 720p25 videos run fine, to give you an example.
Power consumption
The low 6 W TDP (also known as the long-term Power Limit) makes it easy for laptop makers to ditch the fan. Performance sustainability will be poor unless the long-term Power Limit is set to a value higher than the default 6 W and a fan is available to aid in heat dissipation.
The Celeron N4020 is built with one of the old 14 nm Intel processes for poor, as of mid 2023, energy efficiency.
- Range of benchmark values for this graphics card - Average benchmark values for this graphics card * Smaller numbers mean a higher performance 1 This benchmark is not used for the average calculation
v1.28
log 30. 09:12:06
#0 checking url part for id 11681 +0s ... 0s
#1 checking url part for id 11535 +0s ... 0s
#2 redirected to Ajax server, took 1719731526 +0s ... 0s
#3 did not recreate cache, as it is less than 5 days old! Created at Fri, 28 Jun 2024 05:15:21 +0200 +0s ... 0s
#4 composed specs +0.007s ... 0.007s
#5 did output specs +0s ... 0.007s
#6 getting avg benchmarks for device 11681 +0.004s ... 0.011s
#7 got single benchmarks 11681 +0.008s ... 0.019s
#8 getting avg benchmarks for device 11535 +0.004s ... 0.023s
#9 got single benchmarks 11535 +0.015s ... 0.037s
#10 got avg benchmarks for devices +0s ... 0.037s
#11 min, max, avg, median took s +0.041s ... 0.078s