Does SNB's new platform cost power?

As we all know, Intel's Sandy Bridge series of processors using 32nm process technology, in theory, will bring lower power consumption and heat. How much has it improved compared to the previous 45nm products (Bloomfield, Lynnfield core)? I believe that everyone is very interested in this issue, just in the hands of us now have two Intel Sandy Bridge processors Core i7-2600 and Core i5-2300, since this is a hands-on platform to explore together.

Core i5-2300 uses LGA 1155 interface and 4-core 4-thread design, 32nm process, clocked at 2.8GHz, configured with 6MB L3 cache, integrated dual-channel DDR3 memory controller and Intel HD Graphics 2000 graphics card, support Turbo Boost 2.0 Technology, TDP is 95W; Core i7-2600 uses a higher frequency (3.4GHz) and larger L3 cache (8MB), and supports Hyper-Threading Technology (4 core 8 threads), and the rest specifications are with Core The i5-2300 is the same.

Compared with them, the core i5-750 processor of the Lynnfield core, which is the closest generation product, is located closely. It adopts the LGA 1156 interface, 45nm process, 4-core and 4-thread design, configures 8MB of L3 cache, and integrates dual-channel DDR3 memory control. , support Turbo Boost technology, TDP with 95W. The reason why Clarkdale Core i5/i3 does not use a 32nm CPU+45nm GPU is compared, mainly because Clarkdale is only a dual-core processor, and the Core i7-2600/i5-2300 has a large gap.

It should be noted that, in the Core i5-750 and Core i5-2300 processors with similar specifications, the former does not have an integrated GPU, so there is theoretical advantage in terms of heat generation and power consumption.

This test is mainly to explore the heat and power consumption of Sandy Bridge series processors. Compared with the Core i5-750, the Core i7-2600 and Core i5-2300 have higher clock speeds. The Core i7-2600 also supports Hyper-Threading Technology. The remaining specifications are roughly the same.

In the temperature test, the CPU load is provided by ORTHOS, and the average temperature of each core under standby and full load conditions is recorded using AIDA64 software. The power consumption test is performed using Seasonic's PowerAngel power meter, and both are packed and subjected to the sealing test.

During the testing process, the processors of the two platforms will use the default frequency, and EIST energy-saving technology and Turbo Boost Turbo Boost technology will be enabled. In terms of heat sinks, in addition to the use of super-cooled super-cooled Super Mega, we have also added Intel's original heat sink test.

In the power consumption test power consumption test, since the motherboards of the two sets of platforms are not the same, the difference in the final results does not mean the power consumption difference of the processors but the power consumption difference of the entire platform.

Standby power consumption was recorded 10 minutes after the system was turned on and allowed to stand, while the full load power was recorded by the ORTHOS SP2004 after 10 minutes of load. The instrument was a Seasonic PowerAngel.

Under stand-by conditions, the P55 platform equipped with a Core i5-750 leads the P67 platform of the Core i7-2600 and Core i5-2300 in a 9-10W range, but the main reason is that the standby frequency is lower; after full load, With Turbo Boost technology enabled, the actual frequency of the reference processor has been improved over the default frequency. At this time, the Core i5-2300 platform beats the Core i5-750 platform with a 9W advantage, while the Core i7-2600 platform has a power consumption of 148W, because the processor frequency is significantly ahead of the other two.

Considering that Sandy Bridge integrates GPUs, in this case, the Core i5-2300 platform has a higher frequency but consumes less power. It can be seen that the Sandy Bridge platform is superior in power consumption control, and the advantages of the 32 nm process are obvious. Embodied.

Temperature test results The standby temperature is recorded after the system is turned on and left standing for 10 minutes. The full load temperature is recorded after the ORTHOS SP2004 provides the load and waits for a stable temperature profile. The software for detecting temperature is AIDA64 Ver 1.5 software, which records the temperature of each processor core and calculates the average value as the final score.

First of all, it was the ProlimaTech Super Mega radiator that made its debut on the thermal test. It is an excellent air-cooled radiator and it is more than enough to deal with these three processors without overclocking. In the standby case, the temperatures of the three CPUs are basically the same. At full load, the Core i5-2300, which consumes less power, performs equally well and continues to lead the Core i5-750 at 2.6 degrees Celsius. Of course, the highest power consumption Core i7-2600 is also the highest heat, the temperature reached 52 degrees.

Compared to the higher-priced ProlimaTech Super Mega, many users who do not require much heat dissipation are still using the original radiator, so we have also added the original radiator test. Since the LGA 1155 is compatible with the LGA 1156 platform heatsink, the same original heatsink was used in the test, which is derived from the Core i7-2600.

The heat sink of the original radiator can certainly not be compared with Super Mega. In the standby mode, they are all around 31 degrees Celsius, and the difference is only 0.5 degrees Celsius, which can be regarded as due to the detection error. In the case of full load, it is still Core i7-2600 The highest temperature has reached 79 degrees Celsius, followed by the Core i5-750 at 74.1 degrees Celsius, and the lowest is still the Core i5-2300 at only 68.5 degrees Celsius. When using the original radiator, the Core i5-2300 was able to reflect the temperature advantage of the Core i5-750, which also proved the excellent temperature control of the Sandy Bridge platform.

The CPU temperature under the original radiator is still very high at full load. For players who love overclocking, we recommend using a third-party high-end heat sink with better performance.

summary:

From the test results of power consumption and temperature, we can see that the product with the TDP of 95W is the same. The Sandy Bridge processor with 32nm technology is obviously better at platform power consumption and temperature control. The Core i5-2300 has a higher clock speed than the Core i5-750 and incorporates a GPU. However, the CPU temperature and platform power consumption are even smaller. This means that the Sandy Bridge platform has lower requirements for the user's cooling system and power supply, and it is even more Meet the current "energy saving and environmental protection" world theme.

Although the Core i7-2600 has the highest CPU temperature and power consumption in this test, it is intended to replace the Core i7-800 series. Its own default clock speed is as high as 3.4 GHz, which is higher than the 2.66 GHz Core i5-750. 800MHz, but also supports Hyper-Threading technology, but its platform power consumption is only about 10W higher than Core i5-750, also can be regarded as very good.