Skip to main content

We Finally Have Our First GeForce GT 1010 GPU Benchmark

NVIDIA GT 1010
(Image credit: Shutterstock)

As shared by @BenchLeaks on Twitter, we have our first 3rd party benchmark of Nvidia's GeForce GT 1010, showcasing who the GPU performs in Geekbench's CUDA benchmark. The GPU stealthily launched nearly a year ago, and strangely, it's taken that long for someone to get their hands on the entry-level graphics card.

We don't know what has taken Nvidia so long to produce the GT 1010, but we can safely assume the silicon shortage had something to do with it. The new GPU is one of the last graphics cards to be produced on Nvidia's now ancient but very successful Pascal architecture and will fill out the bottom end of Nvidia's GT lineup.

Specs for the GT 1010 include a cut-down version of the GP108 die, featuring a grand total of just 256 shader cores and 2GB of GDDR5 memory. Thankfully the GT 1010 will only come with one "speedy" GDDR5 variant, and not include a slower DDR4 model like Nvidia did with the GT 1030 (which should help keep its performance consistent).

However, its greatest advantage is its incredibly low 30W TDP, making it a good option for small pre-built PCs with power supplies under 200W.

See more

Still, this won't make the GT 1010 a great gaming GPU by any stretch of the imagination, as demonstrated by this Geekbench benchmark. The system utilizing the GT 1010 in the benchmark is equipped with a Core i9-12900K, Asus ROG Strix Z690-F Gaming Wifi and 64GB of memory. The operating system used was Ubuntu 20.04.3 LTS.

The rig allowed the GT 1010 to reach an underwhelming Geekbench CUDA score of 7730 points, which equates to just 5% of the performance of Nvidia's GeForce RTX 3070 GPU. So don't expect to do any hardcore gaming on this GPU at all.

But, the GT 1010 wasn't really designed for gaming in the first place. Nvidia's GT lineup has always been focused on hardware acceleration for basic tasks on low-end machines. It can also provide a plethora of additional display outputs for multi-monitor setups.

Aaron Klotz
Aaron Klotz

Aaron Klotz is a freelance writer for Tom’s Hardware US, covering news topics related to computer hardware such as CPUs, and graphics cards.

  • King_V
    Odd that it has a TDP rating the same as the GT 1030.
    Reply
  • drtweak
    Ummm is it even worth it? I feel like most intergraded graphics in the past 5 years can do just as good of a job XD

    Since there is nothing on Passmark, and going by the 95% slower, 3070 scores 21934. 5% of that is 1096. The HD 630, which came with the 7th Gen 5 years go, scores 1140 in passmark. If it is simply for things like 4th Gen intel and older i can see, but still. Why? lol
    Reply
  • King_V
    I figure just a replacement for the old GT710 and 730 cards, maybe. Anything that needs the extra outputs.
    Reply
  • spongiemaster
    King_V said:
    I figure just a replacement for the old GT710 and 730 cards, maybe. Anything that needs the extra outputs.
    Exactly. When we buy Dell's for our office, if we need a VGA port, the desktop comes with some fanless AMD graphics card. Not sure of the exact model, but you certainly wouldn't attempt to game on it. This card is for OEM's that want to add ports to their office desktops.
    Reply
  • Eximo
    Looks like spare MX230 mobile chips that didn't meet the power requirements to be mobile.
    Reply
  • InvalidError
    drtweak said:
    Ummm is it even worth it? I feel like most intergraded graphics in the past 5 years can do just as good of a job XD
    If you need extra video outputs or just need something to stuff in an IGP-less system as a placeholder or otherwise head-less system, the cheapest GPU possible that is still in-production and actively supported makes sense.
    Reply
  • Olle P
    King_V said:
    Odd that it has a TDP rating the same as the GT 1030.
    Not at all. In the lower segment most GPUs have the same nominal TDP to allow use of the same cooling solution. (Just like there's a bunch of CPUs with 65W TDP.)
    Reply
  • salgado18
    95% slower than an RTX 3070 means absolutely nothing. A few clicks into the benchmark and I could get a score for the GT 1030: around 10300 points on CUDA benchmark. So, the card is around 75% of a GT 1030. Not quite that bad, probably is slightly worse than the Ryzen 2200G GPU. Maybe it can do some light gaming at 720p.
    Reply
  • mitch074
    So... It's slower than a RadeonHD 7770 (itself a bit slower than a GT 1030). And since pretty much all GT 730 variants have gone out of support, I guess Nvidia needed something even cheaper than the GT 1030 DDR that still had official support for OEMs that needed more than the 2 or 3 outputs you can find on most embedded graphics - or who have a surplus of either non-F Intel or -X Ryzen processors.
    Reply
  • King_V
    salgado18 said:
    95% slower than an RTX 3070 means absolutely nothing. A few clicks into the benchmark and I could get a score for the GT 1030: around 10300 points on CUDA benchmark. So, the card is around 75% of a GT 1030. Not quite that bad, probably is slightly worse than the Ryzen 2200G GPU. Maybe it can do some light gaming at 720p.
    What I did to guesstimate that is go to the GPU hierarchy chart.

    The RTX 3070 is 76.3% on that chart. So, 5% of that is 3.815%. If that estimate is correct, that puts it above the Intel Iris Plus on the i7-1065G7, and below Intel Iris Xe DG1.

    Vega 8 in on the R3 3200G does better.

    Again, that's if we can use that 5% of the GTX 3070 as an exact measure. Obviously actual testing would be required. Though, I imagine that would hardly be a priority. Am I curious? Yes. But I also suggested testing the Vega 3 on the Athlon 2x0 APUs, so I'm hardly on the level of "sane" here. :LOL:
    Reply