As of today, Nvidia’s long Turing refresh cycle has come to an end. The GTX 1650 — the last GPU we’re aware of in the Turing family, based on the slimmed-down TU117 GPU — drops in as a replacement for the GTX 1050 / GTX 1050 Ti.
As one might expect, Nvidia has cut-down the TU116 used for the GeForce 1660 and GeForce 1660 Ti to create TU117. The new GPU packs roughly 2/3 the ROPs, cores, and memory bandwidth of the GTX 1660. A comparison between the Pascal and Turing family GPUs in this price range is shown below.
As with other Pascal cards, the label on the 1650 doesn’t match the price branding. This GPU is more a replacement for the 1050 Ti rather than the 1050. In the budget category, these relatively small absolute price differences stack up rather quickly, as anyone who has ever had to bring a build in under a tight price point can attest. Compared against the 1050 Ti, we’d expect the GTX 1650 to offer significantly higher performance. The base clock on the GPU is 1.15x higher, which should translate to at least a 1.1x performance improvement overall, if not more. Boost clocks are 1.19x higher, which means the gains could be larger on clock alone before the improvements of the Turing core are taken into account.
Nvidia Refuses to Sample
We can’t actually tell you anything about how the GTX 1650 performs, however, because Nvidia opted not to sample any GPUs to anyone. It’s not unusual for Nvidia or AMD to put less emphasis on its lower-end products, but in this case, Nvidia deliberately held back the GTX 1650 driver to prevent reviewers who secured cards from third-parties from testing the GPU, even if they had the hardware in-hand. This is not how desktop GPU reviews are typically handled. But we’ve got a pretty good idea why Nvidia did it.
On paper, the GTX 1650 should be at least 15 percent faster than the GTX 1050 Ti. We wouldn’t blink if you told us the actual gains were in the 1.2x – 1.25x range — 15 percent is the minimum gain we’d expect based solely on clock.
The problem is, AMD’s RX 570 is a lot more than 1.25x faster than the GTX 1050 Ti. TechSpot did an exhaustive comparison of the two GPUs and found that the RX 570 is, on average, 1.43x faster than the GTX 1050 Ti at 1080p. The fact that the RX 570 currently enjoys a 0.34 percent market share according to the Steam Hardware Survey compared with the 9.68 percent share for the 1050 Ti is a testament to Nvidia’s market power — but not the competitiveness of its hardware in that particular price bracket.
The RX 570 is certain to burn far more power than the GTX 1650 — we noted during the 1660 launch that the RX 590 burned almost literally twice as much power as the GTX 1660, and the RX 570 could well hit 2x the power draw of the GTX 1650. But launching a brand-new GPU that loses to your competitor’s two-year-old part isn’t a particularly good look either, and it’s likely why Nvidia refused to sample the GTX 1650 driver until today, thereby ensuring that even if hardware sites got their hands on a GPU, they wouldn’t be able to test it.
Given this positioning, we’d recommend waiting for reviews to tip up before pulling the trigger on a GTX 1650. We expect it to be a stronger card than the 1050 Ti, but don’t want to speculate on its performance against the RX 570.