PCI Express 4.0 Brings 16 GT/s And At Least 300 Watts At The Slot

[The Benjamins]

8 hours ago, patrickjp93 said:

It's not even that dirty. People seem to think reducing 90% of the emissions from coal and oil plants is insignificant. Heck, the biggest contributing factor in greenhouse gasses is water vapor.

You also cant have 1 type of power planet design. power distribution companies adjust output to demand and some types like solar, nuclear you have very little control over power output. so they mix the types they use to give them the best cost and control over the output.

[ScratchCat]

MOAR POWAR!!!!!!!

[Thony]

With that wattage support no card will ever need cables ! Exciting ! 

 

But with that much power going through the board the temps must be high. Watercooled mobo :-O

[ScratchCat]

9 minutes ago, Thony said:

With that wattage support no card will ever need cables ! Exciting ! 

 

But with that much power going through the board the temps must be high. Watercooled mobo :-O

Maybe they will increase voltage to reduce the current and temperature on the thin traces

[patrickjp93]

7 hours ago, Agost said:

Having nuclear plants helps a bit
 

Water vapour comes to play later, since water has a different cycle than other greenhouse gases and it's influenced a lot by temperature. Moreover it can have the opposite effect, since more water vapour leads to a higher cloud formation and that rises the Earth's albedo. The main "actual" contributor is CO2, followed by methane and nitrous oxide, since they don't have anything like the water cycle. CO2 is also very dangerous for seas and oceans because it tends to dissolve inside water and this lowers the overall pH, causing issues to underwater fauna and flora (which accounts for most of the Earth's oxygen production)

By volume ocean flora and fauna produce less than 1/8 the oxygen produced by the rainforests, and 1/2 the oxygen produced by the boreal forests in Russia and the U.S.

 

And where did you get that crap about water vapor being helpful? It is not. Water reflects 5x as much infrared energy as it does ultraviolet and visible. It's by far the largest contributor to global warming, with CO2 as a distant second. And actually, pound for pound, methane is far more problematic than CO2. There's just less of it.

[patrickjp93]

7 hours ago, Lord_Doge said:

a 90% reduction? over what? the last 50 years? Our energy is still dirty AF but global warming is just a myth so why bother?

In the last 15 actually. Second, while it's not a myth, we clearly don't understand it fully, as all predictions said both the Arctic and Antarctic ice sheets would melt, and yet the Antarctic is expanding, faster than the Arctic is shrinking actually.

 

It's infinitely more dirty than perfectly clean energy, but that's a useless metric. If anyone needs to clean up its act that is Russia and China. The U.S. is under no obligation to sacrifice its economic outlook for a treaty other parties won't honor. Heck even India has nearly reached our total CO2 emissions, and it doesn't have nearly as many cars or produce nearly as much electricity.

[Guest]

1 hour ago, patrickjp93 said:

In the last 15 actually. Second, while it's not a myth, we clearly don't understand it fully, as all predictions said both the Arctic and Antarctic ice sheets would melt, and yet the Antarctic is expanding, faster than the Arctic is shrinking actually.

 

It's infinitely more dirty than perfectly clean energy, but that's a useless metric. If anyone needs to clean up its act that is Russia and China. The U.S. is under no obligation to sacrifice its economic outlook for a treaty other parties won't honor. Heck even India has nearly reached our total CO2 emissions, and it doesn't have nearly as many cars or produce nearly as much electricity.

I dont believe those numbers. In the last 15 years? No way. Do you have any links?

 

To be clear,I was being sarcastic with the whole "global warming is just a myth." I know the various goings on about it

[patrickjp93]

25 minutes ago, Lord_Doge said:

I dont believe those numbers. In the last 15 years? No way. Do you have any links?

 

To be clear,I was being sarcastic with the whole "global warming is just a myth." I know the various goings on about it

EPA numbers, read 'em and weep.

[Agosto]

6 hours ago, patrickjp93 said:

By volume ocean flora and fauna produce less than 1/8 the oxygen produced by the rainforests, and 1/2 the oxygen produced by the boreal forests in Russia and the U.S.

 

And where did you get that crap about water vapor being helpful? It is not. Water reflects 5x as much infrared energy as it does ultraviolet and visible. It's by far the largest contributor to global warming, with CO2 as a distant second. And actually, pound for pound, methane is far more problematic than CO2. There's just less of it.

Phytoplankton oxygen production is estimated to be more than half of the world's total. Moreover there are tons of cyanobacteria on the surface of every lake and sea, and large forests of seagrass (like posidonia) on shallow seas.

 

 

I've never written that water vapour is only helpful, but it can both cool and warm up the atmosphere depending on some conditions, while CO2 rise does only harm; methane is stronger but much less abundant and quite unstable inside the atmosphere. 

[patrickjp93]

52 minutes ago, Agost said:

Phytoplankton oxygen production is estimated to be more than half of the world's total. Moreover there are tons of cyanobacteria on the surface of every lake and sea, and large forests of seagrass (like posidonia) on shallow seas.

 

 

I've never written that water vapour is only helpful, but it can both cool and warm up the atmosphere depending on some conditions, while CO2 rise does only harm; methane is stronger but much less abundant and quite unstable inside the atmosphere. 

According to whose study?! That's ranked in the mid 10s in most I'm familiar with.

 

CO2 accelerates plant growth which accelerates oxygen production. Methane is a harm-only case, but that can be ionized and condensed for fuel.

[Agosto]

On 23/8/2016 at 4:35 AM, patrickjp93 said:

According to whose study?! That's ranked in the mid 10s in most I'm familiar with.

 

CO2 accelerates plant growth which accelerates oxygen production. Methane is a harm-only case, but that can be ionized and condensed for fuel.

The sum of all phytoplankton and sea plants cover more than 50% of earth's oxygen production, at least according to National Geographics (I couldn't find any paper addressing this particular data; I seriously doubt they're making up things) and I've always heard about it being this way. Phytoplankton has always played a big role in earth's oxygen production since the very beginning of life itself; lack of it lead to some extinctions, as this paper reports (though I don't have full access to it). Are you sure that the ~10% you're talking about isn't the atmospheric CO2 absorption, rather than oxygen production?

Atmospheric concentration of CO2 has risen by ~30% during the last century (we recently hit the 400 ppm mark) while most of forest on Earth have shrunk down due to human activities; land flora will not be enough. Seawater is a HUGE carbon sink and that's where most of CO2 tends to go (and that dissolved CO2 is the main source for marine photosynthetic life forms). While it's true that more CO2 leads to higher oxygen production rate on land, it's also true that increasing temperatures make oceans less capable of storing gases like CO2 and O2, the former leading to an increase of atmospheric carbon concentration and lower amounts of it for marine photosynthesis, the latter being harmful for marine life forms (which are a lot).

Think about it: water covers 71% of our planet and its surface is rich in phytoplankton, even at higher latitudes. On the other hand, landmasses have large desertic areas, tropical forests extent is dropping while high latitude forests are slowly growing. It's not that difficult to accept that oceanic life plays a big role on our planet under every aspect.

Methane concentration is way lower than CO2 (about 2ppm), which makes it play a second role in this scenario; extracting it from the atmosphere would be very expensive and time consuming. Moreover its lifetime is limited due to a relative instability inside the atmosphere, while CO2 is much more stable.

[Clyne]

On 8/21/2016 at 10:34 AM, djdwosk97 said:

Yeah, they're loaded because they can afford an extra $50/year electricity bill. 

 

Let's assume the system is used at full load for 4 hours a day 365 days a year (which is pretty fair/generous assumption). Now the average electricity cost is $0.13/kw/h. So, 365days/yr * 4hrs/day = 1460 hours. A Fury X based system pulls about 400w at full load ( http://www.anandtech.com/bench/GPU16/1533 ). So a system with two Fury X's should be around 650w~. Now assume the PSU has a 90% efficiency, and your power consumption under load is about 720w. So math time: 

720w * 1460 hrs * $0.13kw/hrs / 1000 = $136/year.

 

Now if we were to use a two 980Ti, then it would be 40w~ less, which would be $129/year. 

 

Or if you were to use a single GTX 1060, then it would be 300w~, which would be $57/year. 

 

So yeah, enthusiasts must be loaded because they can afford to spend an extra 10 hours worth of minimum wage a YEAR. Also, anyone who would buy AMD anything must be a moron...I mean, $7/year is really bank breaking stuff. 

This reminds me of somebody saying this on another forum..like if you can't afford the additional money for your electricity bill then you got bigger problems to deal with than power draw on a video card...

Funny how the thread was related to power draw in the first place..

https://hardforum.com/threads/rx-480-asus-strix-review-techpowerup-and-bad-rx-480-overclocking-and-power-usage.1905781/page-4#post-1042443620

[Simon771]

Everything sounds great about that power, so we don't need additional 8pin from PSU to GPU.

But will todays GPUs support that?

 

Could you plug GTX 1070 in that PCI-e 4.0 slot and not plug any 8pin to it? Will be card able to acctually pull all the power just from PCI slot?

[Kunter]

25 minutes ago, Simon771 said:

Everything sounds great about that power, so we don't need additional 8pin from PSU to GPU.

But will todays GPUs support that?

 

Could you plug GTX 1070 in that PCI-e 4.0 slot and not plug any 8pin to it? Will be card able to acctually pull all the power just from PCI slot?

Nope, normal PCIe spec has a protection for that, you'd have to at least short the 8pin itself (like you short PSUs).

And then still the cards are probably not layed out for that.

RX480 is definitely going not to power of PCIe 4.0 because the power plane of the core is not connected to the slot.

[Simon771]

34 minutes ago, Kunter said:

Nope, normal PCIe spec has a protection for that, you'd have to at least short the 8pin itself (like you short PSUs).

And then still the cards are probably not layed out for that.

RX480 is definitely going not to power of PCIe 4.0 because the power plane of the core is not connected to the slot.

So if we want to use advantage of that power delivery via PCI lane, we will need to wait for new generation of GPUs.

I guess we can expect that from GTX 12xx in 2018

[themoose5]

On 8/21/2016 at 0:34 PM, djdwosk97 said:

Yeah, they're loaded because they can afford an extra $50/year electricity bill. 

 

Let's assume the system is used at full load for 4 hours a day 365 days a year (which is pretty fair/generous assumption). Now the average electricity cost is $0.13/kw/h. So, 365days/yr * 4hrs/day = 1460 hours. A Fury X based system pulls about 400w at full load ( http://www.anandtech.com/bench/GPU16/1533 ). So a system with two Fury X's should be around 650w~. Now assume the PSU has a 90% efficiency, and your power consumption under load is about 720w. So math time: 

720w * 1460 hrs * $0.13kw/hrs / 1000 = $136/year.

 

Now if we were to use a two 980Ti, then it would be 40w~ less, which would be $129/year. 

 

Or if you were to use a single GTX 1060, then it would be 300w~, which would be $57/year. 

 

So yeah, enthusiasts must be loaded because they can afford to spend an extra 10 hours worth of minimum wage a YEAR. Also, anyone who would buy AMD anything must be a moron...I mean, $7/year is really bank breaking stuff. 

Great post! 

 

I always found it a little ridiculous that people would bring power bill cost into pc component discussions. Like the numbers we're talking about here are pretty negligible considering everything else that contributes to a persons power bill per month. Even at $129/year that's $10 per month and that's a PC at full load for four hours a day every day, "real world" use case is going to be much less then that for gaming.

 

Side note, if you have the extra cash laying around to buy a Fury X, let alone two, you can certainly afford to pay the extra $10 per month in power bills.

[Kunter]

4 hours ago, Simon771 said:

So if we want to use advantage of that power delivery via PCI lane, we will need to wait for new generation of GPUs.

I guess we can expect that from GTX 12xx in 2018

I guess we have to wait, yes.

 

[Now a longer answer why it may be possible with current GPUs (Except RX480):]

However, the PCIe spec says that one pin pair of 6/8pin (12V to ground) shall be used as sensing, when the card is supposed to start up it tries sense wehter it can pull 12V through the sensing pin, if that's not possible it will immediately  shut down. This is to prevent the load balancer to pull all the required energy through the PEG slot which could literally fry your mobo with something like a Titan X.

This suggests that its technically already possible to pull all power through PEG if you manage to connect the PEG 12V and Ground to the sensing pair. 

Practically the card may refuse to start or fry itself because the engineers designing the power planes probably do that with the expected power draw in mind, so yeah, maybe not.

RX480 is a little different in that point, because AMD did not implement the sensing the card can not determine wehter a 6pin is connected. Instead the core is exclusively fed through the 6pin which means without it the card will try to fire up but core wont start and everything is fine. If you wanted to power it exclusively through PEG you'd have to bridge the power plane of the PEG into the power plane of the core and yeah, I don't now, it'll likely be a mess and blow up instantly.

[Sintezza]

That article is a bit missleading as far as the pci-e power numbers are concerned i guess.

Its most likely 300W combined power for all 16x slots.

Which is basicly the same for the current pci-e 3.0 standards.

[zMeul]

Tom's Hardware got it wrong by assuming PCIe 4.0 will supply 300W from the connector

Quote

PT:PCI-SIG reached out to tell us that the power increase for PCI Express 4.0 will come from secondary connectors and not from the slot directly. They confirmed that we were initially told incorrect information. We have redacted a short passage from our original article that stated what were originally told, which is that the slot would provide at least 300W, and added clarification:

• PCIe 3.0 max power capabilities: 75W from CEM + 225W from supplemental power connectors = 300W total
• PCIe 4.0 max power capabilities: TBD

New value “P” = 75W from CEM + (P-75)W from supplemental power connectors.

 

[jasonvp]

On 8/21/2016 at 2:41 PM, Thread212 said:

i see.. so pcie gen 4 did give more performance for server computer.
i am sorry i just not very familiar when coming to server computer.

It'll make the biggest impact on servers first.  Then, as GPU and other mass market/consumer-level hardware catches up, the 4.0 spec will become more useful in PCs.

 

For servers, imagine massive RAID controllers.  Not the shit included on your motherboard by Intel, but the big money RAID controllers with their own caches and battery backups, etc.  They'll happily gobble up that bandwidth.  Higher bandwidth NICs will, as well.  Things like 40GigE, 50GigE, 100GigE, etc will all be possible at line rate on that bus.  In some cases, it may allow the NIC makers to double up on ports in a single PCI-E card, like they can with GigE and 10GigE today.

 

You're not going to use any of that at home, for a very long time.  But servers can use it now.

[patrickjp93]

I can't seem to paste links on Chrome for iOS right now, but there's a new topic. PCI-SIG clarified and said the slot power limit is still 75W.

[Aytex]

2 minutes ago, patrickjp93 said:

I can't seem to paste links on Chrome for iOS right now, but there's a new topic. PCI-SIG clarified and said the slot power limit is still 75W.

 

 

I gotchu

 

 

 

nvm somone else got it kek

[79wjd]

1 minute ago, patrickjp93 said:

I can't seem to paste links on Chrome for iOS right now, but there's a new topic. PCI-SIG clarified and said the slot power limit is still 75W.

you're about 9 hours late to the party (scroll about three posts)

[patrickjp93]

1 minute ago, Aytex said:

http://www.guru3d.com/news-story/pci-express-4-will-deliver-75-watt-over-pcie-not-300w.html

 

I gotchu

I mean I posted an LTT topic  but thanks all the same.

 

EDIT: 

 

EDIT2: well, rebooting fixed it.

 

[patrickjp93]

1 minute ago, djdwosk97 said:

you're about 9 hours late to the party (scroll about three posts)

Eh, zMeul doesn't count. You guys treat him like a second class citizen on here anyway