Preface:
Regarding this year’s 618, I wonder if any of the DIY players have picked up the graphics card of their choice, or are they waiting for the launch of the new RTX40 series cards with the mentality of: If you don’t buy it, I won’t buy it. The graphics card will be reduced by 500 tomorrow.
According to Ah Lu’s observation, I believe there are many people with the latter mentality. After all, miners have ruined the graphics card market for too long. Players who want to buy graphics cards at the original price have experienced more than 2 years of waiting, and their enthusiasm for installing them is overwhelming. After a lot of consumption, the price of the graphics card has basically returned to its original price today, and players have long since lost their love for it. Therefore, first install a CPU with core display and other RTX40 series, firstly to enjoy the latest game performance, and secondly to eliminate the possibility of buying mining card . Well, if you think so, today's test should have some reference significance.
Yes, what I am testing today is the sensitivity of core graphics card to memory frequency . The test platform uses i5 12500, which integrates UHD770 core display. GPU has 32 execution units (EU) and a basic frequency. is 300MHz, and the maximum dynamic frequency is 1.45GHz. The GPU is based on the 10nm process, based on the Alder Lake GT1 graphics processor, and supports DirectX 12. It basically belongs to the top core display specifications of Intel's 12th generation Core. It is the same as 32EU. Compared with the i9 12900K, the frequency is only 100MHz lower.
product unboxing:
memory uses Kingston FURY Beast DDR5 5600MHz RGB 16G*2, which is an RGB light strip.
As a light bar produced by a world-renowned storage brand, it supports the RGB lighting effect control of major mainstream motherboard manufacturers. It is only a basic operation. If your motherboard is not from the above four companies and you want to play with light pollution, you only need the official FURY CTRL software of Kingston . Yes, and the cool effects are not inferior to them at all.
toothpaste factory
comes with FURY faith stickers with the memory. The red and black color scheme has an e-sports style.
The memory vest is made of metal, with brushing added for a delicate touch. The raised irregular stripes make the vest look less monotonous.
RGB light strip has a very smooth edge transition and a relatively rounded surface shape, which looks a bit like jade. The sticker on the back of
contains various certifications, product serial numbers, voltages, memory packaging location in Bay and other information. It does not mark its timing like other brands.
Let’s take a look at the lighting performance of Kingston FURY Beast DDR5 5600MHz 16G*2 first.
By the way, after installing FURY CTRL, the memory lighting effect can be played to perfection. In addition, the software can also view memory information and real-time operating temperature of the memory. How about
? Do you feel that the lighting effect of FURY CTRL is not inferior to the lighting effect that comes with the board factory software?
performance test:
PS: The following test results are DDR5 4800MHz in the upper picture and DDR5 5600MHz in the lower picture.
memory reading DDR5 5600MHz leads DDR5 4800MHz by about 13%, and memory writing DDR5 5600MHz leads DDR5 4800MHz by about 14%.
Let’s first look at the CPU-Z running scores. The CPU single-core score of 5600MHz leads the 4800MHz group by 1.2%, and the multi-core score of 5600MHz leads the 4800MHz group by about 0.5%.
has a maximum thread increase of 0.3% and a single thread increase of 0.1%.
3DMARK NR running integrated graphics card running score increased by 2.4%.
LOL actual measurement, 1080P is about 4-5 frames higher, the difference is not big.
2K resolution, the gap between frame rate and has widened to 10 frames, but both can still achieve a smooth five-kill level.
Summary:
In general, this test is more about fun. Although the gap between DDR5 4800MHz and DDR5 5600MHz in memory reading and writing performance is as high as 10%+, it is not obvious in current practical applications. Maybe Is it because the memory frequency gap is not big enough? Next time I have the opportunity to try the performance gap between DDR4 and DDR5 high-frequency memory on AMD GPU, it should be more interesting. Thanks for reading the full article. Welcome to give Alu a thumbs up.
