paint-brush
The Evolution of Nvidia's Graphics Cardsby@DawoodKMasood
2,046 reads
2,046 reads

The Evolution of Nvidia's Graphics Cards

by Dawood Khan MasoodJanuary 29th, 2020
Read on Terminal Reader
Read this story w/o Javascript
tldt arrow

Too Long; Didn't Read

This isn't going to cover every single graphics card used in the whole world as there are so many. But we are going to be covering the one company that almost run the whole production and lead the market with a storm. Side note for the Nvidia side the cards I have added are all the main board cards from Nvidia themselves and not custom boards like which MSI and Asus make.

Companies Mentioned

Mention Thumbnail
Mention Thumbnail
featured image - The Evolution of Nvidia's Graphics Cards
Dawood Khan Masood HackerNoon profile picture

This isn't going to cover every single graphics card used in the whole world as there are so many. But we are going to be covering the one company that almost run the whole production and lead the market with a storm. Side note for the Nvidia side the cards I have added are all the main board cards from Nvidia themselves and not custom boards like which MSI and Asus make.

Nvidia NV1 (1995)

Nvidia started making graphics cards all the back in 1993. But they didn't release any card to the public until 1995. The first-ever graphics card they made was called the Nvidia NV1.

Card Information:

When this card was released it's market price was $299. This graphics card was used within the Sega Saturn a great console for its time. The NV1 had a PCI interface and could be used with most motherboards that had PCI slots in. This card supported DirectX 1.0. Built on the 500 nm process size. Die area of 90 mm² and 1 million transistors. Power draw was rated at 2 W maximum.

Card Specs:

  • 2 MB of EDO Memory
  • 480.0 MB/s Bandwidth
  • 75 MHz GPU Clock
  • 60 MHz Memory Clock
  • 1x VGA Output

Nvidia Riva 128 (1997)

The next card to be released by Nvidia was the NV3 (NVIDIA Riva 128) They started to make The NV2 as the successor of the NV1, but as Nvidia had a few issues with Sega. And so they when on there own path and two years later after the NV1 they released the NV3 in 1997.

Card Information:

Market price is unknown. This card supported DirectX 5.0. This card supported OpenGL 1.0. Built on the 350 nm process size. Die area of 90 mm² and 4 million transistors. Power draw was rated at 4W maximum.

Card Specs:

  • 4 MB of SDR Memory
  • 1.600 GB/s Bandwidth
  • 128 bit Memory Bus
  • 100 MHz GPU Clock
  • 100 MHz Memory Clock
  • 1x VGA Output, 1x DB13W3 Output

Nvidia Riva TNT (1998)

After a single year, 1998, Nvidia released the NV4 (NVIDIA Riva TNT). This card was a very well known card since the specs were insane for its time.

Card Information:

Market price is unknown. This card supported DirectX 5.0. This card supported OpenGL 1.0. Built on the 350 nm process size. Die area of 90 mm² and 7 million transistors. Power draw was not known.

Card Specs:

  • 16 MB of SDR Memory
  • 1.76 GB/s Bandwidth
  • 128 bit Memory Bus
  • 90 MHz GPU Clock
  • 110 MHz Memory Clock
  • 1x VGA Output

Nvidia Riva TNT2 (1999)

In 1999, Nvidia released the NV5. (NVIDIA Riva TNT)(NVIDIA Riva TNT2). They wanted to keep on top of all the competitors. This card was made with the AGP port & no longer used the PCI Port.

Card Information:

Market price is unknown. This card supported DirectX 6.0. This card supported OpenGL 1.2. Built on the 250 nm process size. Die area of 90 mm² and 15 million transistors. Power draw was not known.

Card Specs:

  • 16 MB of SDR Memory
  • 1.144 GB/s Bandwidth
  • 64 bit Memory Bus
  • 125 MHz GPU Clock
  • 143 MHz Memory Clock
  • 1x VGA Output

Nvidia GeForce 256 DDR (2000)

After all those cards they finally believed they had cracked the whole video card deal and started to call these video cards graphics cards. And so followed the Nvidia NV10 (NVIDIA GeForce 256 DDR). This graphic card was released on February the 1st, 2000. This card had two versions. (There is a 3rd but Quadro cards will not be on this list). The DDR version, and the SDR. The SDR came out a bit earlier than the main card in the late 1999's.

Card Information:

Market price is unknown. This card supported DirectX 7.0. This card supported OpenGL 1.2. Built on the 220 nm process size. Die area of 139 mm² and 17 million transistors. Power draw was not known. Port interface AGP 4x.

Card Specs:

  • 32 MB of DDR Memory / 32 MB of SDR Memory
  • 4.8 GB/s Bandwidth / 1.144 GB/s Bandwidth
  • 128 bit Memory Bus / 64 bit Memory Bus
  • 120 MHz GPU Clock
  • 150/300 MHz Memory Clock / 143 MHz Memory Clock
  • 1x VGA Output

Nvidia GeForce2 Chips (2001)

With another year gone by, in 2001, came the GeForce2 Chips (NV11, NV15 and NV16). These cards were pretty much the same as NV10 but with larger amounts of ram, and cores. They also started to use the PCI slot once again.

Card Information:

Market price is unknown. This card supported DirectX 7.0. This card supported OpenGL 1.2. Built on the 180 nm process size. Die area of 65 mm² and 20 million transistors. Power draw was not known. Port interface PCI.

Card Specs:

  • 64 MB of SDR Memory
  • 1.328 GB/s Bandwidth
  • 128 bit Memory Bus
  • 175 MHz GPU Clock
  • 166 MHz Memory Clock
  • 1x VGA Output

Nvidia GeForce3 (2001)

The same year (2001) they also released the NV20 (GeForce3). They also went back to the AGP x4 port. This card was the one of the first to use DirectX 8.

Card Information:

Market price was 499 USD on release. This card supported DirectX 8.0. This card supported OpenGL 1.3. Built on the 150 nm process size. Die area of 128 mm² and 57 million transistors. Power draw was not known. Port interface AGP 4x.

Card Specs:

  • 64 MB of SDR Memory
  • 3.680 GB/s Bandwidth
  • 128 bit Memory Bus
  • 200 MHz GPU Clock
  • 230 MHz Memory Clock
  • 1x VGA, 1x DB13W3 Outputs

Nvidia GeForce4 (2002)

Moving to 2002 they started to release GeForce4. These cards were mostly copies of previous cards but with smaller boards and better cooling for the lower end cards. There was a lot of cards using this chipset and they made so many I couldn't list them all.

Card Information:

Market price is unknown. This card supported DirectX 8.0. This card supported OpenGL 1.3. Built on the 150 nm process size. Die area of 65 mm² and 29 million transistors. Power draw was not known. Port interface AGP 4x.

Card Specs:

  • 16 MB of DDR Memory
  • 3.2 GB/s Bandwidth
  • 64 bit Memory Bus
  • 200 MHz GPU Clock200/400 MHz Memory Clock
  • 1x VGA, 1x DB13W3 Outputs

Nvidia GeForce 6800 (2003)

The Nvidia GeForce 6800 (NV48) was released in 2003. This card was a bomb, honestly, because they bumped up the specs massively.

Card Information:

Market price was around $500. This card supported DirectX 9.1c. This card supported OpenGL 2.1. Built on the 110 nm process size. Die area of 287 mm² and 222 million transistors. Power draw was not known. Port interface PCIe 1.0 x16.

Card Specs:

  • 256 MB of DDR Memory
  • 22.40 GB/s Bandwidth
  • 256 bit Memory Bus
  • 325 MHz GPU Clock
  • 350/700 MHz Memory Clock
  • 1x DVI, 1x VGA, 1x S-Video
  • No Power Connectors

Nvidia GeForce 6600 (2004)

Since the success of the 6800 a year later they wanted to touch the whole market. So they released the 6600 a mid-range card that was cheaper than the 6800. This card was released in the middle of 2004.

Card Information:

Market price was around $199. This card supported DirectX 9.0c. This card supported OpenGL 2.1. Built on the 110 nm process size. Die area of 150 mm² and 146 million transistors. Power draw was not known. Port interface PCIe 1.0 x16.

Card Specs:

  • 256 MB of DDR Memory
  • 8 GB/s Bandwidth
  • 128 bit Memory Bus
  • 300 MHz GPU Clock
  • 250/500 MHz Memory Clock
  • 1x DVI, 1x VGA, 1x S-Video
  • No Power Connectors

Nvidia Geforce 7800 (2005)

Then in 2005, they replaced the 6800 with the all-new Geforce 7800. This card came in a few versions. The GT, GT Dual, and the GTX.

Card Information:

Market price was around $449. This card supported DirectX 9.0c. This card supported OpenGL 2.1. Built on the 110 nm process size. Die area of 333 mm² and 302 million transistors. Power draw was not known. Port interface PCIe 1.0 x16.

Card Specs:

  • 256 MB of GDDR3 Memory
  • 32 GB/s Bandwidth
  • 256 bit Memory Bus
  • 400 MHz GPU Clock
  • 500/1000 MHz Memory Clock
  • 2x DVI, 1x S-Video
  • 1x 6-pin Power Connector

Nvidia GeForce 8000 (2006)

Next came the GeForce 8000 Series. This card style was used for multiple GPUs that followed after its release.

Card Information:

Market price was around $399 - $599. This card supported DirectX 11.1. This card supported OpenGL 3.3. Built on the 90 nm process size. Die area of 484 mm² and 681 million transistors. Power draw was 155W. Port interface PCIe 1.0 x16.

Card Specs:

  • 768 MB of GDDR3 Memory
  • 86.40 GB/s Bandwidth
  • 384 bit Memory Bus
  • 576 MHz GPU Clock
  • 900/1800 MHz Memory Clock
  • 2x DVI, 1x S-Video
  • 2x 6-pin Power Connector

Nvidia Geforce 9000 (2008)

Now we move to the Geforce 9000 Series. This card was pretty much the same as the 8000 series but with shrunk down chip sets.

Card Information:

Market price was around $160 - $299. This card supported DirectX 11.1. This card supported OpenGL 3.3. Built on the 65 nm process size. Die area of 324 mm² and 754 million transistors. Power draw was 125W. Port interface PCIe 2.0 x16.

Card Specs:

  • 512 MB of GDDR3 Memory
  • 86.40 GB/s Bandwidth
  • 384 bit Memory Bus
  • 600 - 675 MHz GPU Clock
  • 900/1800 - 1100/2200 MHz Memory Clock
  • 2x DVI, 1x S-Video
  • 2x 6-pin Power Connector

Nvidia GeForce 280 GTX (2008)

The 9000 series was re-branded as the 100 series. The 200 series was released in 2008. All these cards were improved from the 9000 series and mainly had higher memory.

Card Information:

Market price was around $649. This card supported DirectX 11.1. This card supported OpenGL 3.3. Built on the 65 nm process size. Die area of 576 mm² and 1,400 million transistors. Power draw was 236W. Port interface PCIe 2.0 x16.

Card Specs:

  • 1024 MB of GDDR3 Memory
  • 141.7 GB/s Bandwidth
  • 512 bit Memory Bus
  • 602 MHz GPU Clock
  • 1107/2214 - 1100/2200 MHz Memory Clock
  • 2x DVI, 1x S-Video
  • 1x 6-pin + 1x 8-pin Power Connector

Nvidia GeForce 480 GTX (2010)

The 400 series was released in 2010. This new card was based on the Fermi architecture.

Card Information:

Market price was around $499. This card supported DirectX 12.0. This card supported OpenGL 4.6. Built on the 40 nm process size. Die area of 529 mm² and 3,100 million transistors. Power draw was 250W. Port interface PCIe 2.0 x16.

Card Specs:

  • 1536 MB of GDDR5 Memory
  • 177.4 GB/s Bandwidth
  • 384 bit Memory Bus
  • 701 MHz GPU Clock
  • 924 to 3696 MHz Memory Clock
  • 2x DVI, 1x mini-HDMI
  • 1x 6-pin + 1x 8-pin Power Connector

Nvidia GeForce 580 GTX (2010)

The 500 series was released late 2010. This new card was based on the Fermi architecture once again. These cards were pretty much the same as the 400 series but took less power to run and were a little faster.

Card Information:

Market price was around $499. This card supported DirectX 12.0. This card supported OpenGL 4.6. Built on the 40 nm process size. Die area of 520 mm² and 3,100 million transistors. Power draw was 244W. Port interface PCIe 2.0 x16.

Card Specs:

  • 1536 MB of GDDR5 Memory
  • 192.4 GB/s Bandwidth
  • 384 bit Memory Bus
  • 772 MHz GPU Clock
  • 1002 to 4008 MHz Memory Clock
  • 2x DVI, 1x mini-HDMI
  • 1x 6-pin + 1x 8-pin Power Connector

Nvidia GeForce 680 GTX (2012)

The 600 series was released 2012. This new card was based on the Kepler architecture. These new cards were almost ahead of there time as they doubled and tripled most stats that other cards had at the time.

Card Information:

Market price was around $499. This card supported DirectX 12.0. This card supported OpenGL 4.6. Built on the 28 nm process size. Die area of 294 mm² and 3,540 million transistors. Power draw was 195W. Port interface PCIe 3.0 x16.

Card Specs:

  • 2 GB of GDDR5 Memory
  • 192.3 GB/s Bandwidth
  • 256 bit Memory Bus
  • 1006 MHz GPU Clock
  • 1502 to 6008 MHz Memory Clock
  • 2x DVI, 1x HDMI, 1x Display Port
  • 2x 6-pin Power Connector

Nvidia GeForce 780 GTX (2013)

The 700 series was released 2013. Kepler architecture was used once again. Once again Nvidia bought these cards based off there 600 series, but they were able to boost the speeds really well. Sadly the price wasn't very consumer-friendly at the time.

Card Information:

Market price was around $649. This card supported DirectX 12.0. This card supported OpenGL 4.6. Built on the 28 nm process size. Die area of 561 mm² and 7,080 million transistors. Power draw was 250W. Port interface PCIe 3.0 x16.

Card Specs:

  • 3 GB of GDDR5 Memory
  • 288.4 GB/s Bandwidth
  • 384 bit Memory Bus
  • 863 MHz GPU Clock
  • 1502 to 6008 MHz Memory Clock
  • 2x DVI, 1x HDMI, 1x Display Port
  • 2x 6-pin + 1x 8-pin Power Connector

Nvidia GeForce 980 GTX (2014)

The 900 series was released in late 2014. I have skipped the 800 series as it was dedicated for mobile chip sets. these new cards were based on the Maxwell architecture. This architecture was based on power consumption.

Card Information:

Market price was around $549. This card supported DirectX 12.0. This card supported OpenGL 4.6. Built on the 28 nm process size. Die area of 398 mm² and 5,200 million transistors. Power draw was 165W. Port interface PCIe 3.0 x16.

Card Specs:

  • 4 GB of GDDR5 Memory
  • 224.4 GB/s Bandwidth
  • 256 bit Memory Bus
  • 1127 MHz GPU Clock
  • 1753 to 7012 MHz Memory Clock
  • 1x DVI, 1x HDMI, 3x Display Port
  • 2x 6-pin Power Connector

Nvidia GeForce 1080 GTX (2016)

The 1000 series was released in late 2016. This time Nvidia didn't use the same architecture but used a new one called Pascal.

Card Information:

Market price was around $599. This card supported DirectX 12.0. This card supported OpenGL 4.6. Built on the 16 nm process size. Die area of 314 mm² and 7,200 million transistors. Power draw was 180W. Port interface PCIe 3.0 x16.

Card Specs:

  • 8 GB of GDDR5X Memory
  • 320.3 GB/s Bandwidth
  • 256 bit Memory Bus
  • 1607 MHz GPU Clock
  • 1251 to 10008 MHz Memory Clock
  • 1x DVI, 1x HDMI, 3x Display Port
  • 1x 8-pin Power Connector

Nvidia GeForce 1080 GTX (2018)

The 2000 series was released in late 2018. Nvidia took their own path and brought in ray tracing and took it by storm at the start since these cards were the first for consumers. Once again the price was not the greatest even for the lower end cards. Later down the line in 2019 Nvidia updated the cards and added the word Super to the end of each model, these models were once again the same cards but with a little better clocking speeds and power draw. These RTX cards are based on Turing architecture.

Card Information:

Market price was around $699. This card supported DirectX 12.0. This card supported OpenGL 4.6. Built on the 12 nm process size. Die area of 545 mm² and 13,600 million transistors. Power draw was 215W. Port interface PCIe 3.0 x16.

Card Specs:

  • 8 GB of GDDR6 Memory
  • 448 GB/s Bandwidth
  • 256 bit Memory Bus
  • 1515 MHz GPU Clock
  • 1750 to 14000 MHz Memory Clock
  • 1x HDMI, 3x Display Port, 1x USB Type-C
  • 1x 6-pin + 1x 8-pin