• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

PCI Express 2.0 question

you have been helpful i feel more clued up now :) i currently have a asus 8800 gts 640 mb so as forsted nipples has said even if its limted its still a lot better then what i have. However when your paying 400 quid plus for a card you dont ant it to be limited.

What i think i am gonna now do is sell some of my old parts, (i have just about enough to build a pc) think i will get a 260 gtx since my monitor only goes up to the 1620 res i think a 280 will be wasted on me anyways. The money i will save i can put towards a new mobo, the in time maybe get another 260 gtx.

Hmmmm unless it turns out i dont need a new mobo after all.

Arrrrh :P This uprading lark is stressfull.
 
Last edited:
PCI-E 1.1 is ok with PCI-E 2.0 cards as is the other way around without perfromance loss.

And you base this info on what? SHow me some results please of a GTX280 in a pci-e 1.1 slot and a pci-e 2.0 slot with no performance difference. I'm not saying that there will be some (see my other post) but I would like to the evidence.

Yes they work but if the card needs more than 2.5gb/s bandwidth then it will be limited. Actually, due to overheads, you lose 20% of the available bandwidth anyway.
 
Last edited:
When my card arrives what test should i do to find out if it is the same or not?

All you can do is find a review with the same/similar cpu and a pci-e 2 board and run the same benchmarks.

If you get similar results then the GTX280 is not using all the bandwidth of the pci-e 1 slot.

If your results are much lower than expected, then I guess that proves it is limited.

By how much though is the key, it might be none or it might be a lot.
 
I will be running a 4850 on a PCI-E 1.1 slot P35 as a stop gap.

I will be making a comparison directly with my X38 mobo.

The perfromance increse with PCI-e 2.0 will be negible .
 
hmmm getting confused now as some say the 4850 will work with a 1.0a and some say not :confused:this is only reason i havent brought one yet as i dont want too shell out my cash to find it dont work.

Well you have 7 days to return it. Or just wait for somebody with the same 1.0a slot as yours to try it first ;)
 
I will be running a 4850 on a PCI-E 1.1 slot P35 as a stop gap.

I will be making a comparison directly with my X38 mobo.

The perfromance increse with PCI-e 2.0 will be negible .

Maybe so but if there is any increase at all then it is obviously hitting the limit of the pci-e 1 board.

If that is the case then the faster GTX280 will have an even bigger differential between pci-e slots.

And if that is big enough, then even less reason to fork out £400 for something which might not be any quicker than a 4850 if it's bandwidth limited.
 
The PCI Express interface has been around for almost four years, and is in the middle of transitioning to its second generation. PCIe 2.0 effectively doubles bandwidth and offers better flexibility, while maintaining compatibility with PCIe 1.1. When PCI Express was first introduced, it provided more throughput, scalability and flexibility than the aged AGP bus, and it allowed the graphics companies ATI and Nvidia to create multi-card solutions for further 3D acceleration and quality improvement. The second generation of PCI Express was introduced with Intel’s X38 enthusiast chipset, and is being carried on by Intel’s X48, AMD’s 790/770 family and Nvidia’s nForce 7 series. But do we really need PCI Express 2.0 today?

Performance differences among technically similar platforms have almost disappeared, making feature differences and overclocking capabilities the main differentiators. Unless a motherboard manufacturer really messes up fine-tuning its product, two different motherboardshttp://en.wikipedia.org/wiki/motherboard based on the same chipset will perform alike when operated with comparable parameters. While only the latest chipset generation supports PCI Express 2.0 today, there are still many other chipsets available, which provide equal performance and comparable features. However, PCI Express 2.0 offers twice the bandwidth of PCI Express 1.1, by doubling the throughput per lane from 250 MB/s to 500 MB/s. A x16 PCI Express link hence offers the same bandwidth as PCI Express 2.0 at x8.

In order to benefit from the higher throughput, you’ll need both a motherboard and a graphics cardhttp://en.wikipedia.org/wiki/Graphics_processing_unit that comply with PCIe 2.0. While ATI has been quick with deploying PCIe 2.0 to its Radeon HD 3000 family, only the latest GeForce 9 graphics cards are PCIe 2.0 compliant.

We took an ATI Radeon HD3850 mainstream card and an Nvidia GeForce 9800 GX2 dual GPU high end board and limited their PCI Express 2.0 interfaces to x8, x4 and x1 link speeds in an effort to find out whether or not PCI Express 2.0 is really necessary to achieve the expected performance. Regardless of the outcome, PCI Express 2.0 does have one major advantage: new PCIe 2.0 graphics cards are compatible with PCIe 1.1 motherboards, meaning that they will run on them using the available bandwidth of PCI Express 1.1.

Eight PCI Express 2.0 lanes equal the performance of x16 PCI Express 1.1, providing 4 GB/s upstream and 4 GB/s downstream bandwidth to and from the graphics board.


The dual GPU GeForce 9800 GX2 card requires at least x8 PCI Express 2.0 to deliver high frames.


This is like running a GX2 in a 16x PCI-E 1.1 slot.

Crysis shows a noticeable impact when running on a reduced number of PCI Express 2.0 lanes, but only at very high quality. It’s amazing to see that you’ll get almost full performance even at x4 link mode.

So in conclusion If you have a PCI-E 1.1 slot on a P35 mobo you will not notice!
 
Last edited:
I have no clue if i have pci-e 1.1 or 1.0, i will try find out.

I have another question.

The 9800 gx2 out performs the 260 in benchmarks, the gx2 is bascily 2 graphics cards in one slot so is more powerfull then a single 260 gtx. Surely the gx2 would use more bandwidth then the single 260 gtx. Yet theres no mention of the gx2 been compatible with the pci-e 2.0, so what i am saying if a gx2, a dual graphics card in one slot doesnt need the extra bandwidth of pci-e 2.0 why does a 260 gtx?

Hmmmm?
 
So running a 4870 in my Neo2-FR (PCIE v1.0a) would be stupid?

Dammit, only bought the mobo 7 months ago :(
 
I have no clue if i have pci-e 1.1 or 1.0, i will try find out.

I have another question.

The 9800 gx2 out performs the 260 in benchmarks, the gx2 is bascily 2 graphics cards in one slot so is more powerfull then a single 260 gtx. Surely the gx2 would use more bandwidth then the single 260 gtx. Yet theres no mention of the gx2 been compatible with the pci-e 2.0, so what i am saying if a gx2, a dual graphics card in one slot doesnt need the extra bandwidth of pci-e 2.0 why does a 260 gtx?

Hmmmm?

I presume it depends on what chipset you have. I'm not sure and would like to know the answer too.

I have a p35c Ds3R and i assume that it has pci express 1.1 as its a p35 chipset?

Is this correct, and if so will it be fine with a 4870?
 
We took an ATI Radeon HD3850 mainstream card and an Nvidia GeForce 9800 GX2 dual GPU high end board and limited their PCI Express 2.0 interfaces to x8, x4 and x1 link speeds in an effort to find out whether or not PCI Express 2.0 is really necessary to achieve the expected performance. Regardless of the outcome, PCI Express 2.0 does have one major advantage: new PCIe 2.0 graphics cards are compatible with PCIe 1.1 motherboards, meaning that they will run on them using the available bandwidth of PCI Express 1.1.

Eight PCI Express 2.0 lanes equal the performance of x16 PCI Express 1.1, providing 4 GB/s upstream and 4 GB/s downstream bandwidth to and from the graphics board.


The dual GPU GeForce 9800 GX2 card requires at least x8 PCI Express 2.0 to deliver high frames.


This is like running a GX2 in a 16x PCI-E 1.1 slot.

Crysis shows a noticeable impact when running on a reduced number of PCI Express 2.0 lanes, but only at very high quality. It’s amazing to see that you’ll get almost full performance even at x4 link mode.

So in conclusion If you have a PCI-E 1.1 slot on a P35 mobo you will notice!

Is that all from Wiki? Does that mean they did that comparison or somebody else did? I would like to see the charts to see the difference in each game.
 
I have no clue if i have pci-e 1.1 or 1.0, i will try find out.

I have another question.

The 9800 gx2 out performs the 260 in benchmarks, the gx2 is bascily 2 graphics cards in one slot so is more powerfull then a single 260 gtx. Surely the gx2 would use more bandwidth then the single 260 gtx. Yet theres no mention of the gx2 been compatible with the pci-e 2.0, so what i am saying if a gx2, a dual graphics card in one slot doesnt need the extra bandwidth of pci-e 2.0 why does a 260 gtx?

Hmmmm?

I think easyrider's quotes showed the GX2 is needs more tha 8x pci-e 2 and hence more than 16x pci-e 1 to be not limited at high resolutions.
 
I presume it depends on what chipset you have. I'm not sure and would like to know the answer too.

I have a p35c Ds3R and i assume that it has pci express 1.1 as its a p35 chipset?

Is this correct, and if so will it be fine with a 4870?

Yes will be fine.
 
I am on the n force 6 chipset, its says i have 2xPCI-E slots at full x16 speed in my manual, will i be fine ? :(
Says nothing about 1.1
 
Last edited:
Bah, looks like I'll need to update my "only 1 yr old" Asus Commando mobo then (P695, v1.0a).

Hmm, this upgrade from my 8800GTX is starting to get expensive.
 
Back
Top Bottom