• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Pci-e 3.0

As far as I know, the speed is simply doubled. As for why PCIe 1.0 and 2.0 cards are comparable.. that must be related to the bus usage. That, I don't know.
 
Ha, what a bunch of crap.


Yeah, you could probably use a GF 580 on a PCI-E 1.0 MB and still be only losing < 5% performance. SLI/xFire seems to have some slight performance advantages on using higher bw pci-e slots though.


The only thing interesting about this to me is the fact that PCI-E 1x & 4x slots will get a jump in bandwidth. You could not have a USB3 card that runs at full speed.

WTB 10Gb Ethernet please =D
 
No, it will not make a difference in perceivable performance within the foreseeable future. We do not come close to saturating the 2.0 specification yet. But hey, nothing sells better than 'faster' to geeks, right?
 
some of you guys are ridiculous. the same crap is said every time we move to a faster standard. so are we supposed to just stop at 2.0 since that is sufficient for now? of course not. many people keep their mobos for several years so having a standard that will better support future gpu upgrades makes sense.
 
Last edited:
some of you guys are ridiculous. the same crap is said every time we move to a faster standard. so are we supposed to just stop at 2.0 since that is sufficient for now? of course not. many people keep their mobos for several years so having a standard that will support future gpu upgrades makes sense.

the idea is that pcie is backwards compatiable, and the difference in gpu performance difference between pcie 1.0 and 2.0 is minimal. if that holds true with pcie 3.0, then there's no reason to upgrade, allowing you to, exactly as you stated, keep your mobo for several years.
 
some of you guys are ridiculous. the same crap is said every time we move to a faster standard. so are we supposed to just stop at 2.0 since that is sufficient for now? of course not. many people keep their mobos for several years so having a standard that will better support future gpu upgrades makes sense.

No one is saying we should halt progress, I merely object to tech being marketed as something that will drastically improve your computing/gaming experience. I like shiny new toys as much as anyone BUT, I am unwilling to pay a premium or deal with software/firmware that hasn't been optimized for the sake of early adoption. I realize the last doesn't apply to the original topic but, it applies to the majority of 'new' tech.

Everyone should have a hobby but, there is WAY too much money being wasted IMHO on the latest and greatest that can only show an improvement when running obscure benchmarks.
 
No one is saying we should halt progress, I merely object to tech being marketed as something that will drastically improve your computing/gaming experience. I like shiny new toys as much as anyone BUT, I am unwilling to pay a premium or deal with software/firmware that hasn't been optimized for the sake of early adoption. I realize the last doesn't apply to the original topic but, it applies to the majority of 'new' tech.

Everyone should have a hobby but, there is WAY too much money being wasted IMHO on the latest and greatest that can only show an improvement when running obscure benchmarks.
and at which point should they implement it? 3 or 4 years from now when gpus may need more bandwidth or next year before it ever becomes an issue? its going to happen sometime and its not like anyone ever has to upgrade just for 3.0 since by the time we need it, it will already be on any mobo that's relevant.
 
Last edited:
and at which point should they implement it? 3 or 4 years from now when gpus may need more bandwidth or next year before it ever becomes an issue? its going to happen sometime and its not like anyone ever has to upgrade just for 3.0 since by the time we need it, it will already be on any mobo that's relevant.

it's not a matter of when they implement it, it's a matter of whether or not you need to upgrade to it.
 
and at which point should they implement it? 3 or 4 years from now when gpus may need more bandwidth or next year before it ever becomes an issue? its going to happen sometime and its not like anyone ever has to upgrade just for 3.0 since by the time we need it, it will already be on any mobo that's relevant.

I agree. Implement it at soon as possible, it's the marketing I hate. You see it's effects among enthusiasts all the time. 'Of course you have to have a 1000 W PSU, of course you have to have a quad core CPU, of course you have to have 6 + gigs of ram.'
 
it's not a matter of when they implement it, it's a matter of whether or not you need to upgrade to it.
what I am saying is by the time you do need it that it will already be there. they have to start sometime so implementing it before its even needed is just common sense.
 
I agree. Implement it at soon as possible, it's the marketing I hate. You see it's effects among enthusiasts all the time. 'Of course you have to have a 1000 W PSU, of course you have to have a quad core CPU, of course you have to have 6 + gigs of ram.'
but marketing is there for EVERY product. and honestly the vast majority of people have never even heard of pci-e express standards or know what they mean.
 
but marketing is there for EVERY product. and honestly the vast majority of people have never even heard of pci-e express standards or know what they mean.

so what's wrong with telling people that they might not need to upgrade?
 
so what's wrong with telling people that they might not need to upgrade?
tell who? people will be buying mobos because of new cpu sockets not because of pci-e 3.0. again any person that has relevant mobo by the time they need pci-e 3.0 will already have that feature since it will be implemented starting next year. there is nothing to complain about.
 
I'm on Toyota's side. Do any of you hate loading screens? Individually, CPU and GPU runs really fast, but there is a problem of communication. Intel has fitted GPU on die, meaning that communication is far better than it to discrete GPU. While is doesn't hurt discrete GPU, it is actually bias if you think about it. Rumors has it that the PCIe bandwidth affects the performance of PhysX. PhysX is an example where GPU off loads CPU, and if it is somehow bottlenecked by PCIe bandwidth, then it actually opens up a can of worms. At any rate, the controller is on CPU die, I don't think it is there interest to go for PCIe anytime soon.
 
and at which point should they implement it? 3 or 4 years from now when gpus may need more bandwidth or next year before it ever becomes an issue? its going to happen sometime and its not like anyone ever has to upgrade just for 3.0 since by the time we need it, it will already be on any mobo that's relevant.

I wish that the industry was more pro-active about the 2TB HD limit of MBR partitions and BIOS booting. Now that's a real problem that needs solving.

Someone needs to come out with a BIOS module that loads off of a bootable flash drive, and loads EFI OS installs off of GPT disks. Surprised MS hasn't come out with a USB flash drive bootloader for Win7, since it supports GPT partitions.
 
what I am saying is by the time you do need it that it will already be there. they have to start sometime so implementing it before its even needed is just common sense.

Looks like SATA III didn't get that memo as SandForce's 2nd generation controllers might push 500MB/sec reads/writes. Where is SATA IV? :biggrin:
 
No one has touched on this but will PCI-E 3.0 push more power through the socket than 2.0?

From PCI-SIG:

Q15: Does PCIe 3.0 enable greater power delivery to cards?
A15: The PCIe Card Electromechanical (CEM) 3.0 specification consolidates all previous form factor power delivery specifications, including the 150W and the 300W specifications.
http://www.pcisig.com/news_room/faqs/pcie3.0_faq/
So yeah, it looks like they quadrupled power through the bus...though it most likely won't allow that much power through the bus with current 2.x (or past) cards.
 
toyota, I *think* most people are commenting about how the article implies that PCI-e 3.0 is going to be some instantaneous leap forward due to this comment:
A lot of technobabble, sure, but one thing's for sure: your next graphics card is bound to murder your current one if paired with a PCIe 3.0 motherboard.
I think that part is what sparked all this. Probably just a miscommunication; although, if someone here is arguing that PCIe 3 is superfluous even years into the future, yea they are an idiot.

Q15: Does PCIe 3.0 enable greater power delivery to cards?
A15: The PCIe Card Electromechanical (CEM) 3.0 specification consolidates all previous form factor power delivery specifications, including the 150W and the 300W specifications.
http://www.pcisig.com/news_room/faqs/pcie3.0_faq/
I could be wrong because I haven't looked into it, but to me that sounds like it won't be able to deliver more power through the bus as pcie 1.0 supported 300W.
 
I did find presentation slides on the electronic updates...does seem that the bus still provides the same amount of max power (75W).
 
I dont see why everyone here is fighting this so much. We already need it today let alone tommorow, alot of 1156 boards wont even let you use SATA 3 and USB 3 at full speed if you fill both PCIe slots. A faster bus would allow that to work no problem. And thats just issues from today just think of the bottleneck PCIe 2 would have when SATA 4 and USB 4 come out let alone faster GPU's.
 
I dont see why everyone here is fighting this so much. We already need it today let alone tommorow, alot of 1156 boards wont even let you use SATA 3 and USB 3 at full speed if you fill both PCIe slots. A faster bus would allow that to work no problem. And thats just issues from today just think of the bottleneck PCIe 2 would have when SATA 4 and USB 4 come out let alone faster GPU's.
The problem you described roots from the P55 chipset and money saving measures, not a lack of PCIe speed. A 2x PCIe 2.0 slot could easily handle the bandwidth requirements of USB 3 or SATA 3.

The problem with P55 is that it doesn't actually have real PCIe 2.0 slots, they are limited to the same speed as PCIe 1.0. Note I am not talking about the x16 slots that come directly off the CPU, just the 8 "PCIe 2.0" from the Northbridge. Because these slots run at half speed, we need four of them to fully handle a USB 3 connection. Now the real problem shows itself:

A lot of 1156 boards do not physically have enough ports to handle a x16 video card, a x4 USB3 controller and a x2 SATA3 controller. Change platforms to X58 and its a non-issue because of SATA3 controllers only needing a 1x slot, and the massive amount of slots the chipset allows.
 
Back
Top