SSD can also be overclocked: performance and power consumption soar. The final outcome was unexpected.

When it comes to overclocking, you will think of CPU processors and GPU graphics cards, but do you know? SSD solid state drives can also be overclocked! As long as there is a chip with its own clock frequency, it can theoretically exceed it. YouTube blogger Gabriel Ferraz showed how he overclocked an SSD.

The test subject is unknownRZXPro240GBSATASSD——The new NVMe SSD is not used because they are already running at the highest state, but this SATASSD is different.

Its main control is Huirong SM2259XT2, a single ARC32-bit core.The highest frequency is 550MHz, but here it's only 425MHz.

flash memory isKioxia’s 96-layer BiCS4TLC, single capacity 32GB,The nominal frequency was originally 400MHz, but it actually only ran at 193.75MHz.

Perhaps to control power consumption, perhaps to extend life, the main control and flash memory of this SSD are running at reduced frequencies, which leaves room for overclocking.

Of course, overclocking SSD is not easy. Gabriel used mass production tools, JMS578SATA-USB bridge controller and a series of tools, as well as superb skills and repeated attempts.

final,The main control is restored to full health at 500MHz, with an overclocking range of 17.6%. The flash memory is also at full health at 400MHz, with an overclocking range of 106%.

What is the effect? Take a look at the test:

CDM sequential read and write speed and read latency have almost remained unchanged, while write latency has slightly decreased.The random read and write speeds have increased significantly, with QD4 queue depths of approximately 27% and 10% respectively, and Q1T1 also with 16% and 11% respectively.

3DMark and PCMark10 storage test scores increased by as much as 21% and 11% respectively.

But in actual scenarios, the loading speed of applications and games has not become faster at all.

File copy speed has become faster, especially for a large number of small files, which is consistent with the acceleration of random test performance.

The price is not small either.The pressure test temperature reached 45°C, an increase of 5°C, and the original 40°C is the maximum allowable temperature set by the manufacturer - now I know why the speed is limited.

achievementConsumption also increased across the board, with maximum, average, and standby increases of 73%, 63%, and 32% respectively.

In this way,Actual energy efficiency was reduced by almost 42%, it is true without trying.

This wasn't the end yet. Just when all kinds of tests were completed and we were ready for maintenance, it died. . .