[rayiner]

The top Mac Studio has six thunderbolt 5 ports, each of which is a PCIe 4.0 x4 link. Each is a 8GB/sec link in each direction, which is a lot. Going from x16 down to x4 has less than a 10% hit on games: https://www.reddit.com/r/buildapc/comments/sbegpb/gpu_in_pci...

[mrheosuper]

Your example uses GTX1080, which is a very old GPU. Current flagship consumer GPU will take a harder hit on low bandwidth PCIE.

[rayiner]

Here’s more recent HW: https://www.pugetsystems.com/labs/articles/impact-of-gpu-pci...

This is an RTX4080.

“In the more common situations of reducing PCI-e bandwidth to PCI-e 4.0 x8 from 4.0 x16, there was little change in content creation performance: There was only an average decrease in scores of 3% for Video Editing and motion graphics. In more extreme situations (such as running at 4.0 x4 / 3.0 x8), this changed to an average performance reduction of 10%.”

[AnthonyMouse]

A 10% performance reduction seems like a lot to be leaving on the table.

[rayiner]

Not really.

[mrheosuper]

The article is nearly 3 years old and the 4080 is not even top of the line at the written time.

Still, 10% in difference is still considerable, almost gen-to-gen difference

[zozbot234]

PCIe 4.0 x4 is going to be a huge bottleneck, even recent SSDs have more throughput (they use PCIe 5.0) never mind GPUs.

[washadjeffmad]

Gaming isn't what people are using Mac Studios for. Thunderbolt also isn't a substitute for OCuLink.

[rayiner]

Sure, but it’s probably reflective of the fact that GPUs generally aren’t PCIe-bandwidth bound. Also, TB5 and Oculink2 both use PCI 4.0 x4 links.

[angoragoats]

Oculink is generally faster than TB5 despite them both using PCIe 4.0, because Oculink provides direct PCIe access whereas Thunderbolt has to route all PCIe traffic through its controller. The benchmarks show that the overhead introduced by the TB5 controller slows down GPU performance.

[wtallis]

It's not just the controllers; the Thunderbolt protocol itself imposes different speed limits. The bit rates used by Thunderbolt aren't the same as PCIe, and PCIe traffic gets encapsulated in Thunderbolt packets.

[rayiner]

Apple Silicon has an integrated thunderbolt controller so that should have less latency than PCs that use a discrete thunderbolt controller.

[angoragoats]

Maybe; I'm unable to find any benchmarks that specifically compare PCs with TB to Macs to test this. But there is certainly still overhead with TB no matter what, and therefore it'll never be as fast as Oculink.

[rayiner]

Sure, but how big of a difference is there? Even inside a desktop PC, you typically have PCIe ports directly off the CPU and ones off the chipset, and the latency for the latter is double. But the difference is immaterial in practice.

[adrian_b]

Many recent laptop CPUs from Intel and AMD have integrated Thunderbolt controllers (i.e. USB 4), so that has not been a difference for a long time.

[izacus]

That's just blatantly wrong, the performance loss of GPUs is very well documented and gets worse as you go towards higher end models. We're talking 30%+ loss of performance here.

[spacedcowboy]

Um, I have an M3 Ultra 512GB on my desk for development. Love me some Baldur’s Gate 3, everything turned up to 11…

[washadjeffmad]

I should have said 'buying'. We've all got something fun we run on ours.

[matja]

Yeah 80GB/s total I/O bandwidth is a lot for a Mac, but desktop PCs have been doing 1TB/s (128x PCIe5) for years (Threadripper etc).

[rayiner]

Sure. And lots of people need all that I/O. But my point is that it’s not like the Mac Studio has no I/O. The outgoing Mac Pro only has 24 total lanes of PCIe 4.0 going to the switch chip that’s connected to all the PCI slots. The advent of externally route PCIe is a development in the last few years that may have factored into the change in form factor.