Can someone explain the naming of the different graphic card models?

Solution 1:

The following information was found here. The main reason graphics cards naming is so confusing is because they like to cram a lot of information in the name including; GPU manufacturer and model number, card manufacturer and model number, memory interface, motherboard slot, what it support, and more. If you cram a bunch f information into a name, you will get confusion.

Overall Naming

Graphic card naming schemes can be a bit difficult to decipher until you know what each part means. It can vary somewhat from manufacturer to manufacturer, but for the most part is pretty universal. Let's look at a typical card and parse it out.

Here is a typical description:

• EVGA 640-P2-N829-AR GeForce 8800GTS SSC 640MB 320-bit GDDR3 PCI Express x16 HDCP Ready SLI Supported Video Card

Here's what that all means:

• EVGA: This is the manufacturer of the card, in this case, EVGA. There are many card manufacturers, but only 2 main GPU manufacturers. This is similar to motherboards and CPU's. There are many motherboard manufacturers, but only 2 main CPU manufacturers. This is usually the first part of the description.
• 640-P2-N829-AR: This is the manufacturer's model number, almost always following the manufacturer.
• GeForce: This is the manufacturer of the GPU. There are two main GPU manufacturers, Nvidia and ATI (owned by AMD).
• 8800GTS SSC: This is the GPU model number. The main part of this is the 8800GTS. This is the model number from Nvidia, and will give you the best indication of the speeds of the card. More information on this below.
• 640MB: This is the amount of memory on the video card.
• 320-bit GDDR3: This is the memory interface.
• PCI Express x16: This is the type of motherboard slot that will be required to plug in this card. The x16 is the speed of the PCIe slot. PCIe speeds include x1, 2, 4, 8, 16, and 32, with 32 being PCIe 2.0.
• HDCP Ready: HDCP is a digital copy protection developed by Intel. HDCP ready means that the card can read and play HD DVD and Blu-ray discs. Without this, you would not get a picture.
• SLI Supported: SLI is Scalable Link Interface. This is a technology designed to allow 2 or more graphics cards to be used in a single system.

GPU Naming

NVIDIA

The other 3 numbers refer to the market that the card is marketed for. They go in increments of 50.

• 000-450: These are Nvidia's mainstream cards. The price range of these cards is usually under $150.00 and they usually have less than 512MB of memory. Current games will be playable at low to medium settings.
• 500-750: These are Nvidia's performance cards. They are priced from $100 - $300, and for the most part will play the current releasing games, on medium settings. These cards will have from 256 - 512MB of memory.
• 800-950: These are Nvidia's Enthusiast cards. They are priced from $200 - $700. These cards will for the most part play the current releasing games at high graphic levels. Memory on these cards will range from 512MB - 1GB.

ATi

ATI's naming convention is similar. Their card numbers will relate to the different markets that they are aimed towards. The ATI naming scheme has changed over the years, this is their most recent. The first number in the name refers to the series of the graphic card. The next three will determine what market the card is aimed at.

• 350-590: This is ATI's budget line. These cards will cost less than $100 and have 64MB - 128MB of memory. These cards will usually need graphics set to low settings for current games.
• 600-790: This is ATI's Mainstream line. These cards will cost from $100 - $200, with 128MB - 512MB of memory. These cards will normally play today's games at medium settings.
• 800-990: This is ATI's Enthusiast line. These cards will cost over $150, with 512MB - 1GB of memory. These cards will play today's games on high settings.

Note: I've been contacted by a number of builders asking about the ATI radeon graphics cards; they are no longer available on the retail market but can be found through third-party board manufacturers, who build and sell the Radeon-based boards.

Solution 2:

The reason why graphics card naming is made deliberately confusing is that the names are invented by marketing people rather than engineers. For marketing people the priority is for the name to sound more exciting than the current competition.

There is no need for the name to be consistent with the makers other current products - although they do make a small effort to achieve this.

There is absolutely no need for the marketing name to bear any resemblance to the names that engineers might use.

If you want to compare cards you have to either investigate and understand their various capabilities or rely on simplified published benchmarks (such as frame-rates for various game demos)

Solution 3:

No, No I can't. I don't suggest you try, either. In my opinion, they are too confusing to bother with - instead decide on an approximate price point, and rank cards based on benchmark results. Pay no attention to generation or status as entry-level, enthusiast, whatever, as there's no standard rule for "Next generation's entry level is more powerful than last generation's mid-range", or so on, it varies wildly.

While David's answer is very comprehensive and worth knowing, it's important to note that both ATI and nVidia are making it up as they go along, and there are no proper rules or guidelines as to what performance level links to what model number.