Cyberpunk 2077 update promises 'up to 33%' higher frame rates on Arrow Lake and at the very least we gained 8% in the benchmark

Cyberpunk 2077's 2.2 update could be fantastic for anyone with a machine packing an Intel Arrow Lake chip—all five of you—thanks to announced CPU-threading optimisations for the game. The changes would reportedly put CPUs like the Core Ultra 9 285K up to the same performance figures of Ryzen 7 7800X3D, albeit in one game, but our own tests don't quite show that level of improvement.

In the Cyberpunk 2077 patch notes, the devs promise "Optimized threading system to improve performance by up to 33% on Intel Arrow Lake based CPUs." 

Not likely, our Nick reckons. According to our tests of the Cyberpunk 2.2 patch, we're looking at closer to 6-8% increases on average. Where previously we were getting 103 fps averages and 76 fps 1% lows with the Core Ultra 9 285K, the 2.2 patch only hits 110 fps averages, and 77 fps 1% low. That's closer to 8% than the projected 33%.

That's running at 1080p, which puts the most stress on the CPU, and using DLSS Balanced. Our test system uses an RTX 4070. It's possible that other configurations or presets will return better results, and we stuck with the in-game benchmark for our brief testing. There may be other spots in the game with greater CPU utilisation and a larger subsequent improvement.

It's also worth noting that our ARL test rig has had a couple of BIOS updates since those first results were taken. Without reverting back to the original BIOS used in the review (which wasn't very stable), we can't be sure the performance uplift is even down to the patch itself. Either way, we're not entirely sure where those apparent 33% projections came from, but there's maybe some niche case where you'll see such an uplift.

When we first got our hands on Intel's Core Ultra 9 285K, we noted that while its E-cores were smashing, it was notably slower than 14th Gen chips like the Core i9 14900K, particularly where Cyberpunk 2077 benchmarks were concerned. On the whole, Arrow Lake was a little lacking against the competition.

In our testing we found that limiting the chip to its P-cores at least wasn't that much of a detriment in Cyberpunk 2077. This is down to CD Projekt's engine for the game (REDengine 4) being primarily designed around an eight-core CPU that will make use of more cores if available. When running on mostly E-cores, however, it suffers a 14% drop in performance compared to the full sweep of E- and P-cores available, and an even more dire 22% deficit in the 1% lows. Basically, Cyberpunk 2077 had issues with the chip's P-cores. 

We reckoned that 22% could be pulled back by upping the E-core's clocks, though. 

Either way, we thought maybe the CD Project devs had executed some E-core improving threading magic that would make Arrow Lake look a lot better in Cyberpunk 2077 benchmarks, even pushing it past the kind of performance we've seen from the Core i9 14900K, but the numbers are a little less impressive on our end.

That said, even without a massive increase in performance, we did note the game now feels at the very least smoother to play on Core Ultra post-patch. That's still a win.