We all know, graphics cards and motherboards have "memory", but the kind of motherboards are called memory sticks, and graphics cards on the known as video memory. So far, video memory and system memory use the exact same technology. But high-end graphics cards require faster memory than system memory, so more and more graphics card makers are turning to DDR2 and DDR3 technology.
GDDR2 memory
The DDR2 and DDR3 used in graphics cards are different from the DDR2 and DDR3 used on motherboards, not least of which is the difference in voltage. Therefore, they are referred to as GDDR2 and GDDR3 to differentiate them from each other (the "G" here is an abbreviation for Graphics, the English word for graphics card).
GDDR3 memory
Besides the difference in principle, one of the main differences between DDR and DDR2 is the voltage: DDR operates at 2.5 volts, while DDR2 operates at 1.8 volts. Because of the lower operating voltage, DDR2 consumes less power and generates less heat than DDR.
The Dualsim 5700 Ultra using GDDR3 - Little Demon G TURBO 5718GXIII
Because GDDR2 operates at a much higher frequency than the DDR2 of the system memory, it uses an operating voltage of 2.5 volts instead of 1.8 volts, which is a bit more heat-generating. This is the reason why very few graphics cards use GDDR2 memory - only the GeForce FX 5700 Ultra and FX 5800 Ultra. Shortly after the GeForce FX 5700 Ultra came out, many manufacturers shifted to the GDDR3 5700 Ultra, presumably for the sake of GDDR3. Ultra, presumably for its low power consumption and low heat generation.
The 5800 Ultra with GDDR2 memory
GDDR3 can operate at 2.0 volts (on chips such as Samsung's) or 1.8 volts (on chips other than Samsung's), which solves the heat problem. This is the reason why high-end graphics cards are using this type of memory.
There are no DDR3 system memory sticks yet, but expect them to be rated at a likely 1.5 volts, not unlike GDDR3.
Physically, both GDDR2 and GDDR3 come in BGA packages, as does DDR2 memory. Go to Google or the homepage of the memory chip manufacturer and look for the number on the chip, and we can superficially identify whether the card uses GDDR2 or GDDR3 memory.
At present, graphics cards are divided into two main categories: the familiar home graphics cards that are used for gaming, and workstation graphics cards that are used for professional engineering, film and animation. Compared to the former, most consumers are unfamiliar with professional cards, and there are a lot of misconceptions when it comes to purchasing them, so we will briefly analyze the differences between the two.
First, let's take a look at some of the questions consumers often have when purchasing a gaming card, as these basic terms are ****common to professional cards and gaming cards, and therefore useful when purchasing a professional card.
A:Currently, PCI-E graphics cards are very hot, are they necessarily better than AGP ones? What is the difference between the two?
Before exploring this question, let's take a brief look at the history of these two types of interface buses. AGP is a bus interface introduced by Intel in 1996, from the earliest AGP1×, AGP2×, AGP4× to the current mainstream AGP8×, the operating frequency has also been increased from 66 MHz to 533 MHz, and the peak working The bandwidth was 533MB/s and jumped to 2.1GB/s. However, compared to other computer accessories, this development is undoubtedly slow, especially for professional cards, which are subject to constraints. Therefore, in 2001, Intel launched the uplink and downlink transmission rate of up to 4GB/s for the PCI-E bus specification, which has flourished for 10 years, and the AGP bus has gradually withdrawn from the history stage!
From the introduction above, it is easy to find that the biggest advantage of PCI-E over AGP is the data transfer rate, which is the root cause of the rapid development of PCI-E graphics cards, and many manufacturers, including Elsa, already have a complete lineup of PCI-E products, so as long as the conditions allow for PCI-E graphics cards should be the consumer's choice. However, AGP is still the preferred choice for many users due to its long development time and mature technology, which gives it a clear advantage in the mid-range and low-end product lines
B:Why can't my graphics card reach its maximum nominal resolution?
This is a point of confusion for many ordinary users, as the resolution that a graphics card can provide is often greater than the monitor's resolution. In general, the maximum resolution of a home monitor is 1280x1024, so even if the graphics card can provide a resolution greater than this value, the monitor can not support it, so there is no practical significance of overemphasizing the resolution when purchasing.
C:What is video memory? Is it the same as RAM, the bigger the better?
Graphics memory, known as display memory, is basically the same as the memory in our computers. Its main function is to store the data processed by the display chip, and the larger its capacity, the higher the mapping accuracy will be, so in a sense the performance of the memory increase will be significantly improved. However, for the average user, the pursuit of high-capacity memory will only increase expenses, 128M memory capacity of the card is sufficient to cope with most of the current game and work software needs, so consumers should buy products according to their own actual situation.
Graphics memory is no less important to a graphics card than the display core, and involves parameters such as bandwidth, speed, and package type in addition to capacity, so here's a quick explanation of some of the parameters involved.
D: Why is there a difference between DDR and DDR3 memory?
The so-called DDR and DDR3 are both types of graphics memory. DDR, which is similar to the memory we all know, has been adopted not only because of its mature technology and lower cost, but also because it is easier to increase the frequency during production. But just as DDR2 is sweeping the memory market, many mid-range and high-end products in the current market, including the Elsa Shadow Thunderbolt P940ultra, Shadow Thunderbolt A660GT, and the Phantom Thunderbolt X70PRO, are using the faster, more frequent DDR3 memory.
E:I often see package form factor in product descriptions, what does this mean?
Let's start by clarifying the concept of memory packaging. The so-called graphics package is commonly referred to as a protective jacket for the graphics chip to avoid unnecessary damage to the graphics chip, which is the same as the CPU. At present, the most widely used graphics card is TSOP (thin small package), and BGA (Ball Grid Array Package) is also due to its good overclocking and heat dissipation is being favored by many manufacturers, including Elsa, such as the Shadow Thunder 620TC is the use of BGA packaging. However, due to a number of technical and pricing issues, a very small number of manufacturers have adopted these packages for their individual products.
F:I've heard that memory bandwidth is important, is it?
There's no doubt about it. Memory bandwidth does have a significant impact on the performance of a graphics card. It is a measure of the data transfer rate between the graphics chip and the memory, and generally speaking, the larger the bandwidth, the faster the data transfer rate, in bytes per second. The higher the bandwidth, the higher the performance of the graphics card at the same frequency. This is similar to how a six-lane highway has more traffic than a four-lane highway at the same time.
The next step is to introduce workstation graphics cards, which are often referred to as professional cards. Compared to the gaming card market, professional cards are in a sense unique. For example, ATi's only partner in the Asia-Pacific region for professional cards is Elsa, so all professional cards in the region that utilize ATi's graphics chips are Elsa products. Of course, the fact that Elsa is the only one with a professional card is just to show that the chip maker will only work with one vendor in a certain region for technical and quality reasons.
For professional cards, simply emphasizing on technical parameters doesn't make much sense for the average gamer, and here I would like to briefly talk about the misconceptions that consumers have about the relationship between professional cards and gaming cards. Currently, many high-end gamers and even some enterprise-level users use their gaming cards as professional cards, which not only does not achieve the desired purpose, but also affects their work efficiency.
A large portion of consumers believe that the display cores used in professional cards are the same as the cores used in high-end gaming cards, but this is not the case. Generally speaking, the game only requires about 50,000 polygons of processing power, and professional software often need to millions of polygons of oversized files for effective and fast processing, so professional graphics creation software for professional card requirements are harsh. In order to meet this requirement, the core architecture of professional graphics must be completely based on Maya, Solidworks and other professional graphics creation software and design, give full consideration to the professional graphics creation software for professional card geometry and light processing, double-sided illumination, supersampling capabilities and other aspects of the higher requirements. Therefore, the design should not only meet the hardware requirements of professional software, but also optimize for the current mainstream graphic creation software to meet the need for long and stable operation, and these characteristics will not be integrated into the design of the game card. So with the game card to run professional software will often appear splash screen, can not complete the display and so on. This is why we often hear that professional cards are the same as high-end gaming cards, which is very wrong, and that there is more to professional gaming cards than just speed.
In addition, another obvious difference is the added value of the product. Most gaming cards we're familiar with give away mainstream games like CS, and some cards give away video editing software. However, most professional cards, including Elsa's FireGL series products, generally come with commonly used software, such as HYDRAVISION multi-monitor management software for building dual-monitor displays, and so on, which not only brings great convenience to professional graphic creators, but also greatly improves their work efficiency. It's easy to see that the ProCard offers a full range of support for graphic creators. This all-round support is also expressed in the service, in addition to the necessary product warranty, Elsa and other professional card manufacturers will have a professional pre-sales, after-sales service team to help users to solve the purchase and use of a variety of thorny issues.
Through the above introduction, we believe that most readers will have a further understanding of the graphics card, and this is exactly where we are. At the same time, we realize that for those who are eager to learn more, the above is only a fraction of what is needed. For this reason, Elsa will be focusing on more knowledge-based articles in the future, so that more consumers can buy the products they want.
Recently, a variety of graphics card review articles are increasingly appearing around us, many of which, in addition to professional analysis, will introduce the product's workmanship, "compared to the XX product's workmanship is better," such words and phrases are common, but a more in-depth analysis of the rare, which is the reason why this article was created.
"Workmanship" is actually a very general concept for products. It is mainly divided into: 1, design 2, materials 3, manufacturing process three major aspects, we will analyze these three points one by one.
1, design The design of the product is to determine the "workmanship" good or bad premise. It directly determines the material and manufacturing process of the product later. Graphics card design compared to the motherboard is much simpler, in addition to driver optimization and software debugging, in terms of hardware, wiring is the key to determine the quality of the card. Good wiring not only ensures that the distance from each piece of memory to the display chip is the same, but also should have good resistance to electromagnetic interference and very little electromagnetic radiation. Reflected in the appearance of the card, you should see from the memory to the display chip with a large number of serpentine lines to ensure that the length of each line is consistent, thus enhancing the stability of the card. Serpentine line also eliminates long straight wiring in the current through the inductive phenomenon, greatly reducing the crosstalk between the line, of course, by reducing the density of the wiring can also play the same role. Electromagnetic interference and electromagnetic shielding has been the design of graphics cards to overcome the problem, generally using 4-layer or 6-layer board design graphics cards, and with a large area of copper grounding can be a good solution to this problem.
From the above, it is not difficult to analyze the surface area of a well-designed graphics card will not be too small, must be made larger to facilitate wiring, especially those with more memory particles of the graphics card. The disadvantage of increasing the area of the card is that the cost increases, but the use of double-sided chip technology can be a good solution to this contradiction, that is, the design can be installed on the reverse side of the card chip components PCB board, making full use of the surface area of the card. Although this design will also relatively increase the cost, but far less than increasing the surface area. However, there are drawbacks to this approach, such as higher technical requirements and more difficult to design.
2. The material used is the most direct reflection of a graphics card's workmanship. The material is the easiest way to reflect the workmanship of the card. As mentioned above, the material is determined by the design, the use of 4-layer or 6-layer board design is actually a material problem, the general European and American manufacturers of graphics cards are 6-layer board design, the advantage is that the design is easy to little consideration of the length of the wiring is consistent with the problem of EMC and electromagnetic shielding is good, easy to pass the CE safety regulations. Disadvantages. I do not need to say it, so Europe and the United States large manufacturers designed graphics cards than the average graphics card is much more expensive. In fact, Europe and the United States large manufacturers designed graphics cards sold expensive there is another reason, is the use of a large number of chip components, gallium capacitors and metal chip resistors are very expensive electronic components. A penny for a penny, good components to ensure that the quality of these cards and stable performance, the appearance of the product also looks neat, beautiful. On the contrary, some Taiwan design and production of graphics cards more cheap electrolytic capacitors, also known as upright capacitors, and most of the 4-layer board design, and even some cards or 2-layer board. From the appearance of the obvious difference between the design of Europe and the United States, the surface of the card appears to be cluttered and untidy, the two performance comparisons of European and American products are usually slightly superior, but Taiwan's products have a greater advantage in terms of price.
3, manufacturing process now graphics card manufacturers are already using machines to arrange materials and welding, so the components on the board are generally very neatly arranged, but this is limited to SMD components, electrolytic capacitors, these plug-in components will inevitably be upside down, affecting the appearance of the neat. In this regard, the advantages of the all-patch design of the graphics card are fully realized. In addition, Europe and the United States of America's graphics cards are similar to the milling machine method to cut the PCB board, so that the graphics card edges are very smooth and beautiful. While Taiwan's graphics card PCB boards are cut, folded board method of production, although it does not affect the performance, but the appearance is a bit rough. In the manufacturing process there is also a place that can clearly see the good or bad workmanship is the thickness of the gold plating of the gold finger. A good quality PCB board should see the gold fingers have a certain thickness that can withstand repeated plugging and unplugging to ensure that the graphics card makes good contact with the slot.
Simply put, the standard for determining a graphics card's workmanship is that the components on the card's PCB should be neatly arranged, with clean and even solder joints, electrolytic capacitors with both feet inserted into the low, not to the east and west, and the gold fingers plated thick, not easy to barge, and the card's smooth edges indicate that the manufacturer's manufacturing process is excellent, and the graphics card will not be poorly made.
Finally, I'd like to say a few things to keep in mind when shopping for a graphics card. Low-end graphics cards generally prioritize price, so its materials will be poorer, a large number of electrolytic capacitors, using ordinary resistors instead of metal-chip resistors, the use of 2-layer board design, and the production of old machines, and even the use of artificial plug-ins. Choose this type of graphics card as long as the quality is good, the warranty can be fixed. After all, the product has been cheap to this point. Europe and the United States of America's high-grade graphics cards, to pay attention to its origin, if it is the original cut price is. The most difficult to choose and buy should not be on the Taiwan video card, a few famous video card workmanship is certainly not bad, but it is very expensive, far beyond its cost price, if you are willing to spend more money to buy peace of mind, I have nothing to say. Because it defeats the purpose of my writing this article in the first place. In fact, in the lower price of these cards can be found in the product of performance, workmanship is not inferior to them at all. The difficulty lies in the fact that there are a lot of mixed products and their propaganda is so strong that you can be fooled if you're not careful.
2. Try to buy products from companies that have their own manufacturing plants, at least in terms of quality control.
3. Try to buy a graphics card made by a motherboard manufacturer, as they are generally well equipped to test the compatibility of motherboards and graphics cards, and motherboard manufacturers often get their hands on new or even unannounced motherboard chips very early, so their graphics cards will have fewer compatibility issues with future motherboards and will be easier to fix if something goes wrong.
4. Some of the smaller aspects of the workmanship reflect the level of care that went into designing the product. For example, whether to use a fan or a heatsink, and what is the padding between the fan or heatsink and the display chip. Needless to say, a fan with a heat sink filled with thermal adhesive is much better than a heatsink that is felted on with double-sided tape. Unless you believe the claim that our products, due to their excellent design, generate very little heat and don't need to be fitted with a fan.
5. Pay attention to the gold finger part of the graphics card, the workmanship of the material varies greatly, from the side of the good graphics card gold finger plating thick, there are obvious protrusions. Plated good by repeated plugging and unplugging is not easy to barge off. Tell you a little trick: pay attention to the samples in the window of the gold finger, general samples placed on the longer time, often inserted to pull to try, plus oxidation, very easy to make the gold finger barge down.