Connect with us
Samsung will make video cards much faster The company introduced Samsung will make video cards much faster The company introduced

Components

Samsung will make video cards much faster. The company introduced GDDR6W memory with a bandwidth of HBM2e level

Published

on

Samsung has introduced a new memory for video cards. The new solution is called GDDR6W.

Samsung will make video cards much faster.  The company introduced GDDR6W memory with a bandwidth of HBM2e level

As the name implies, this is not something completely new, but it is not at all the same as GDDR6/GDDR6X.

GDDR6W deserves attention both for its characteristics and for the technologies behind it. In particular, the new packaging technology FOWLP (Fan-Out Wafer-Level Packaging) is used here. It is thanks to her that the new memory is allocated with much more bandwidth and twice the capacity.

Samsung will make video cards much faster.  The company introduced GDDR6W memory with a bandwidth of HBM2e level

In particular, compared to the usual GDDR6 with a frequency of 24 GHz, the bandwidth of GDDR6W is twice as high, as is the capacity. Thanks to this, the volume of one microcircuit was increased to 4 GB, the number of I / O operations also doubled, and the microcircuits themselves became much thinner, since the crystal is mounted directly on a silicon wafer, and not on a printed circuit board. As a result, the height of the chip with doubled capacity is only 0.7 mm versus 1.1 mm for conventional GDDR6.

Speaking of bandwidth, Samsung compares GDDR6W to HBM2e memory, which is 1.4 TB/s for the chip at 22 GHz.

The company completed standardization for the new memory in the second quarter. But it is not yet clear when GDDR6W will hit the market.

Click to comment

Leave a Reply

Your email address will not be published.

Components

The GeForce RTX 4060 Ti can be 20-25% faster than the RTX 3060 Ti. Tests of Nvidia adapter based on GPU AD106 appeared

Published

on

The GeForce RTX 4060 Ti can be 20 25 faster than

The first test results of an Nvidia video card based on the AD106 GPU in its full configuration have appeared on the Web.

The GeForce RTX 4060 Ti can be 20-25% faster than the RTX 3060 Ti.  Tests of Nvidia adapter based on GPU AD106 appeared

What this model will be called is still unclear, but based on early leaks, it could be the GeForce RTX 4060 Ti. True, those very leaks attributed to it a little less CUDA cores, that is, the GPU will be slightly cut.

So, the AD106 configuration allegedly includes 4608 CUDA cores. Also, the specified video card has 8 GB of memory and probably a 128-bit bus. Tests in AIDA64 are not related to reality, but the results in 3DMark are already more useful.

The GeForce RTX 4060 Ti can be 20-25% faster than the RTX 3060 Ti.  Tests of Nvidia adapter based on GPU AD106 appeared

The GeForce RTX 4060 Ti (let’s call it that for convenience) was compared with the RTX 3070 Ti and RTX 2080 Super. You can see that the new product, depending on the test, is either a little faster or a little slower than the RTX 3070 Ti. We can say that there is parity between the cards. At the same time, the TDP of the novelty is 180 W versus 290 W for the RTX 3070 Ti.

The GeForce RTX 4060 Ti can be 20-25% faster than the RTX 3060 Ti.  Tests of Nvidia adapter based on GPU AD106 appeared

According to the leaks, the RTX 4060 Ti will receive not 4608 CUDA cores, but 4352 cores, that is, about 5% less. Accordingly, the performance may be a little less, but here the situation will also depend on the frequencies. Either way, if the information is correct, the RTX 4060 Ti will be about 20-25% faster than the RTX 3060 Ti. However, at the same time, you need to understand that the new product can be more expensive, because so far in the RTX 40 line, this is exactly the situation relative to the RTX 30.

Continue Reading

Components

Nvidia quietly released the most powerful version of the GeForce RTX 3060, but it only has 6 GB of memory

Published

on

Nvidia quietly released the most powerful version of the GeForce

Nvidia seems to have quietly released another version of the GeForce RTX 3060 graphics card. A version with the GA106 GPU in its full configuration has been added to the TechPowerUp database.

Nvidia quietly released the most powerful version of the GeForce RTX 3060, but it only has 6 GB of memory

That is, such an adapter has 3840 CUDA cores versus 3584 cores for the regular RTX 3060. By the way, the mobile RTX 3060 just has the same 3840 CUDA cores, but due to significantly lower frequencies, of course, it cannot compete with the desktop version.

The source says the new RTX 3060 is already in use in some off-the-shelf PCs from Asus, Dell, MSI and others. At the same time, such a card may not appear in retail.

Interestingly, the memory capacity of the new modification is only 6 GB, that is, half as much as the original card. As a reminder, there is also a retail RTX 3060 with 8 GB of memory and a 128-bit bus, which was introduced at the end of last year.

How critical 6 GB of memory will be for the RTX 3060 is difficult to say without tests, but the mobile versions have exactly this amount of memory.

The source suggests that the new version was released in order to get rid of the GA106 GPU stock faster.

Continue Reading

Components

Crisys was launched on a completely Chinese-made video card. Moore Threads MTT S80 also compared to Intel Arc A770

Published

on

Crisys was launched on a completely Chinese made video card Moore

The Moore Threads MTT S80 video card is completely Chinese-made, although it is not able to compete with the flagships or even mid-budget accelerators from AMD and Nvidia, but it can already run the game that once brought to its knees all the 3D cards available at that time on the market.

Crisys was launched on a completely Chinese-made video card.  Moore Threads MTT S80 also compared to Intel Arc A770

Computer enthusiast and collector Löschzwerg managed to run Crysis on a Chinese graphics card. Unfortunately, he did not report anything about performance, and the fps counter, as you can see in the screenshot, clearly does not work.

Crisys was launched on a completely Chinese-made video card.  Moore Threads MTT S80 also compared to Intel Arc A770

However, the question of performance as such is not acute for the S80. Firstly, we have already seen tests where the new product performed approximately at the level of the GeForce GTX 1060, and secondly, there are fresh tests. For the Chinese adapter, optimization is more important, since so far many games on the S80 simply do not start.

In recent tests, older versions of 3DMark compared the adapter to the Intel Arc A770, and the latter was often about three times faster.

Crisys was launched on a completely Chinese-made video card.  Moore Threads MTT S80 also compared to Intel Arc A770

Recall that the Moore Threads MTT S80 is based on a 12-nanometer GPU, has 16 GB of memory and supports PCIe 5.0, which even the newest Nvidia, AMD and Intel video cards cannot boast of.

Continue Reading

Most Popular