A graphics card is the heart of any gaming computer, so it rightly receives a lot of attention from consumers and media alike. Nvidia has been the most dominant player in the graphics card market for years now, whether you look at desktop cards or the mobile versions used in laptops. While the company has quite a few spots in the worst graphics cards on the desktop side, its laptop GPUs have also had their fair share of flops, especially after the debut of the laptop RTX GPUs. Today, we look at five of the most disappointing laptop GPUs from Nvidia's RTX lineup.

1 RTX 2060

There used to be a time when Nvidia's laptop GPUs were appropriately named to differentiate them from their desktop counterparts. Remember the MX mobile GPUs and the ones with the M suffix even before that? This nomenclature clearly conveyed to the customer that they were buying a different GPU altogether. But, with the launch of the RTX graphics cards and their arrival on the best gaming laptops, Nvidia quickly dispensed with this courtesy. Laptop RTX cards were now being named identically to the more powerful and altogether different desktop GPUs.

One of the first GPUs with this egregious practice in action was the RTX 2060 (80W). This laptop GPU used the same TU106 die used on the desktop RTX 2060 and featured the same memory bandwidth, but that's where the similarities ended. The laptop RTX 2060 was clocked way below the desktop card (960/1200MHz vs 1365/1680MHz).

Now, it's obvious that laptop GPUs have to be cut-down versions of their desktop variants, but during the Pascal microarchitecture era, laptop GPUs like the GTX 1070 were clocked and performed quite similarly to their desktop counterparts. They were logically named differently to avoid any confusion on the part of the average consumer in the cases they weren't. The RTX 2060 laptop, however, was as much as 30% slower than the desktop RTX 2060 on average in 1080p gaming. Even if you chose the higher-powered 115W laptop 2060, the difference was still 18% on average.

In fact, the situation was so dire that the laptop variants of the RTX 2060 and the GTX 1660 Ti performed virtually the same, making the supposedly more powerful card essentially worthless. Moreover, the performance difference between the laptop 2060 (80W) and the desktop 1660 Ti was laughably huge. Essentially, the laptop RTX 2060 was similar only in name to the full-power desktop RTX 2060.

2 RTX 2070 Max-Q

The trend of misrepresenting laptop GPUs continues with the RTX 2070 Max-Q. The Max-Q branding was used for low-power cards that were meant to deliver better power efficiency than even already-cut-down laptop GPUs. To be fair, the RTX 2070 Max-Q has the same TU106 GPU as the desktop RTX 2070, complete with the same memory configuration. But, in actual gaming performance, the two GPUs are in two completely different performance classes.

For obvious reasons, Nvidia enforced a strict 80W power limit for the RTX 2070 Max-Q, a far cry from the desktop 2070's 175W TDP. Even the base clock of the Max-Q variant was nearly half that of the desktop RTX 2070. All of these compromises were probably necessary to accommodate the GPU in thin laptops with cooling constraints. Still, the average buyer would expect an RTX 2070 Max-Q to perform somewhere in the ballpark of its desktop counterpart, and that wasn't the case.

On average, the desktop RTX 2070 was around 40% faster than the RTX 2070 Max-Q in 1080p gaming. If you chose to substitute the Max-Q GPU with the higher-powered laptop RTX 2070, the difference wasn't as stark, but it was still around 20%. Consumers who chose to go with the RTX 2070 Super Max-Q in hopes of better value would also have been disappointed, as the desktop RTX 2070 Super was around 35% faster.

3 RTX 3070

The RTX 3070 desktop GPU was one of the best value-for-money cards of its time (if you could get it during the big GPU shortage of 2020). It was priced at $500 MSRP and was similar in performance to the RTX 2080 Ti. But was the laptop RTX 3070 based on the same GPU die? Yes. Did it perform similarly to its desktop sibling? Not even close.

With the RTX 3000 series of mobile GPUs, Nvidia decided to no longer use the Max-Q moniker to differentiate between GPU variants with different power limits. This task was left to the OEMs to be displayed on the laptop's spec sheet. Sure, it was a mandate, but you can't really expect an uninformed consumer to see a TDP figure next to the RTX 3070 and understand the performance they're getting. Naming different variants differently would have been the right thing to do.

Let's discuss the performance numbers then. If you compare the RTX 3070 laptop (115W) with the desktop RTX 3070, it is around 30-35% slower, on average, in 1080p and 1440p gaming benchmarks. While this is already an abysmal showing, it gets even worse when picking the RTX 3070 laptop (80W), as the low-power variant is around 50% slower, on average, in both 1080p and 1440p gaming.

It was undoubtedly disappointing to see an upper-mid-range laptop GPU perform so badly as compared to its desktop variant. Nvidia's shady naming scheme was back in full force, and customers were left to decipher the various TDP ratings of different RTX 3070 laptop variants by themselves.

4 RTX 3080 Ti

Now, we're moving to a serious performance category. The RTX 3080 Ti laptop GPU was a refresh of its previous most powerful laptop GPU, the RTX 3080. Nvidia even created a new GPU die, the GA103, to replace the GA104 seen in the laptop GPU. This new GPU saw an increase in CUDA cores while most other specs remained the same. The clock speeds did see a small decrease, though. The laptop RTX 3080 Ti was available only with 16GB VRAM (no 8GB variant anymore). Things looked good on paper.

But, as you've probably guessed, the real-world gaming performance paled in comparison to not just the desktop RTX 3080 Ti but even the desktop RTX 3080. The laptop RTX 3080 Ti GPU ended up being as much as 30% slower than the desktop RTX 3080 and around 40% slower than the desktop RTX 3080 Ti. While these figures might be in line with what we've seen in other laptop and desktop comparisons, it hurts especially hard in this case because consumers expect no-compromise performance from a GPU named RTX 3080 Ti.

Being significantly slower than its desktop counterpart wasn't the only issue with the RTX 3080 Ti laptop GPU. It simply didn't need to exist. RTX 3080 Ti laptops delivered only around 5-10% more FPS than those with the laptop RTX 3080 while costing significantly more. Put simply, the RTX 3080 Ti was one of the worst mobile RTX GPUs ever seen despite being a great product in isolation.

5 RTX 4090

We've arrived at the big daddy of 'em all. The desktop RTX 4090 still reigns at the top of the most powerful consumer graphics cards on the market. Gamers can finally run Cyberpunk 2077 in RT Overdrive at playable framerates! Jokes aside, when the RTX 4090 came to laptops, consumer expectations were enormous. And while 4K 60FPS gaming (sometimes more than 100FPS in some titles) was impressive on a laptop, the performance chasm compared to the desktop RTX 4090 was significant. Factor in the price of RTX 4090 laptops, and you'd find it difficult to find reasons to go for the laptop.

The laptop RTX 4090 featured the AD103 GPU used in the desktop RTX 4080, while the RTX 4090 desktop had the full-fledged AD102 die. This severely cut down the specs of the laptop card vs. the desktop one, with the added blows of reduced clock speeds, decreased VRAM amount and memory speed, and overall slashed memory bandwidth. The laptop card was also power limited to 150W, a far cry from the desktop 4090's 450W TDP.

Now, the benchmarks: Based on several reviewers, the desktop RTX 4090 was 56% faster than the laptop RTX 4090 at 4K resolution, with the 1440p and 1080p performance gaps being around 35% and 20%, respectively, on average. Gamers buying an RTX 4090 laptop were probably not gaming at 1080p or even 1440p resolutions, so the 4K performance gulf becomes the most glaring drawback here. Seeing the fastest laptop GPU in the world deliver only around 60% of the performance of its desktop counterpart was downright depressing.

Future of laptop GPUs

Laptop GPUs have always been catching up to desktop GPUs, and there are many practical reasons for this. What happens with the next GPU generations is hard to tell. Still, there are promising signs, such as the fully integrated liquid cooling system seen in the upcoming Lenovo Legion 9i, which will allow the RTX 4090 inside it to better flex its muscle.

Innovations like these will no doubt help bring laptop gaming closer to desktop gaming. How affordable and accessible these innovations are, and when they will become widely available across gaming laptops, is something we'll have to wait for.