Intel Skylake / Kaby Lake

[Arachnotronic]

Next to nobody needs it in consumer market.
Show me the consumer segment that uses 44 lanes?
How do they use the 44 lanes and point me to the actual persons doing so?

GTX 1080 in SLI - 32 lanes
Multiple M.2 SSDs and/or PCIe slot SSDs, using 4x PCIe lanes each - 4*n lanes for some positive integer n representing SSD count.

 

[MrTeal]

Next to nobody needs it in consumer market.
Show me the consumer segment that uses 44 lanes?
How do they use the 44 lanes and point me to the actual persons doing so?

I don't use 44, but I do use 36 out of 40 off the CPU. At one point I was using all 40 when I had quad 390s installed. It's not a typical install, but it's hardly outside the realm of possibility.

 

[Sweepr]

2017 and we still have games that rely on strong performance per core - including Microsoft's latest DX12 title Halo Wars 2. Once again Core i3-6100 results bode extremely well for Pentium G4560, basically the same performance for a much lower price ($64 MSRP). Also worth noting Halo Wars 2 (DX12 only title) supports Intel Gen 9 iGPUs - minimum requirements include HD 520.

That'd be nice.

I'd like to see:

6 core/28 lanes @ $379
6 core/44 lanes @ $499
8 core/44 lanes @ $799
10 core/44 lanes @ $1199

Then Intel could to a true "extreme edition" which would be a 12 core with all 48 lanes enabled @ $1649.

I guess it's possible, especially if Skylake-EP LCC packs 12C/24T.

 

[krumme]

Next to nobody in the consumer market today needs 8 high performance cores. But the argument for 8 cores is seldom about today, but the future, and I don't see why the same shouldn't be true for PCIe lanes.

And there are certainly people today running 1080s in sli at 16x. Throw in a PCIE ssd and/or even a third 1080 for physx and you'll be at 40-44. Is the net benefit huge? Probably not. But we are talking about a segment of the market that gets giddy pushing every bit of performance out of a CPU.

Huge?
The difference between 8x and 16x for gaming is at best 1% but more like 0%.
http://www.gamersnexus.net/guides/2488-pci-e-3-x8-vs-x16-performance-impact-on-gpus
So no even in the future it makes no sense.
Secondly- even in theoretical world where 16x even mattered witch it dont - the number of user using more than 2 gpu and one pci ssd is next to zero. Well. Not that it matters anyway. There is not a tiny bit of performance to get here. Get ram with one cycle less latency and you get more performance in gaming than that.

"Inconsequential Differences & Margins for Error
These numbers are close enough in some instances – like the GTA V 58.3 vs. 58 FPS output – that they're effectively within margin of test error and do not definitively show a performance gap. When a reasonable performance gap is shown – like the ~1% difference in Metro: Last Light numbers – it is imperceptible to the user but measurable with our tools. And we do mean imperceptible – we're talking 96FPS vs. 95FPS, for Metro.

Metro, by the way, is the most reliable FPS benchmarking tool we have ever used. The game produces almost precisely the same AVG, 1% low, and 0.1% lows with every single test pass, and so we trust these metrics as being outside of test variance.
"

 

[Ajay]

Huge?
The difference between 8x and 16x for gaming is at best 1% but more like 0%.
http://www.gamersnexus.net/guides/2488-pci-e-3-x8-vs-x16-performance-impact-on-gpus
So no even in the future it makes no sense.
Secondly- even in theoretical world where 16x even mattered witch it dont - the number of user using more than 2 gpu and one pci ssd is next to zero. Well. Not that it matters anyway. There is not a tiny bit of performance to get here. Get ram with one cycle less latency and you get more performance in gaming than that.
"

I've been running an x58 mobo for 8 years. I plan on a building new system later this year. I'll probably go with an x299 based board even if I just put KBL-X in it for now. I've run as many as 3 video cards in my system (SLI+PhyX card) - it's done everything I've asked of it. I suspect my next motherboard will last around the same amount of time. If that time is anything like the the last, I'll upgrade memory once, CPU once, storage* three time and video three times; I'll probably upgrade memory and storage less since I've learned to plan ahead better.


* I did lose one of my HDDs that was in a raid array, but I just moved some infrequently used stuff over to my home server rather than upgrading.

 

[dahorns]

Huge?
The difference between 8x and 16x for gaming is at best 1% but more like 0%.
http://www.gamersnexus.net/guides/2488-pci-e-3-x8-vs-x16-performance-impact-on-gpus
So no even in the future it makes no sense.
Secondly- even in theoretical world where 16x even mattered witch it dont - the number of user using more than 2 gpu and one pci ssd is next to zero. Well. Not that it matters anyway. There is not a tiny bit of performance to get here. Get ram with one cycle less latency and you get more performance in gaming than that.

"Inconsequential Differences & Margins for Error
These numbers are close enough in some instances – like the GTA V 58.3 vs. 58 FPS output – that they're effectively within margin of test error and do not definitively show a performance gap. When a reasonable performance gap is shown – like the ~1% difference in Metro: Last Light numbers – it is imperceptible to the user but measurable with our tools. And we do mean imperceptible – we're talking 96FPS vs. 95FPS, for Metro.

Metro, by the way, is the most reliable FPS benchmarking tool we have ever used. The game produces almost precisely the same AVG, 1% low, and 0.1% lows with every single test pass, and so we trust these metrics as being outside of test variance.
"

First, they were testing single cards, not SLI. Here is some SLI testing showing a significant difference: http://www.reddit.com/r/nvidia/comments/4wzpu0/titan_xp_sli_pcie_30_16x16_vs_16x8/?sort=new

Second, how can you possible support the statement "So no even in the future it makes no sense." GPUs and SSDs have been advancing fairly quickly. Why is it inconceivable that bandwidth demands may exceed the 8x limit 4 years from now?

Third, why does this matter to you so much? You don't think it will be valuable--for you. There are others who think it is valuable. Some of them have posted in this thread. Google 1080 SLI 16x and you'll find plenty of commentators on forums and reddit who use that setup.

Finally, I appreciate the irony of someone lamenting Intel's lack 8c/16t for the mainstream and simultaneously deriding the suggestion that additional PCIe lanes could be useful for the upper end mainstream/low end enthusiast. Either are essentially irrelevant to the mainstream user as 99.999% of use cases will see almost no benefit. Further, unlike the higher core count which can actually be a detriment to many mainstream use cases over a faster 4c, I don't think extra PCIe lanes is ever anything but a boon (from a performance perspective).

 

[krumme]

I've been running an x58 mobo for 8 years. I plan on a building new system later this year. I'll probably go with an x299 based board even if I just put KBL-X in it for now. I've run as many as 3 video cards in my system (SLI+PhyX card) - it's done everything I've asked of it. I suspect my next motherboard will last around the same amount of time. If that time is anything like the the last, I'll upgrade memory once, CPU once, storage* three time and video three times; I'll probably upgrade memory and storage less since I've learned to plan ahead better.


* I did lose one of my HDDs that was in a raid array, but I just moved some infrequently used stuff over to my home server rather than upgrading.

I guess you dont run an array of 4 pci ssd ?
8x for video is just plenty fast. If you ran out of gpu memory relying on pci is not getting you above 5fps. The future will not change that. The pci bus is not working here anyway.
And as you say most start to use their small server or nas for storage needs if they are a power user. And i mean in 3 years thats a cheap 10gb network attached solution beeing powered like something similar to a zen/core quad perf level be it arm or x86.

 

[krumme]

First, they were testing single cards, not SLI. Here is some SLI testing showing a significant difference: http://www.reddit.com/r/nvidia/comments/4wzpu0/titan_xp_sli_pcie_30_16x16_vs_16x8/?sort=new

Second, how can you possible support the statement "So no even in the future it makes no sense." GPUs and SSDs have been advancing fairly quickly. Why is it inconceivable that bandwidth demands may exceed the 8x limit 4 years from now?

Third, why does this matter to you so much? You don't think it will be valuable--for you. There are others who think it is valuable. Some of them have posted in this thread. Google 1080 SLI 16x and you'll find plenty of commentators on forums and reddit who use that setup.

Finally, I appreciate the irony of someone lamenting Intel's lack 8c/16t for the mainstream and simultaneously deriding the suggestion that additional PCIe lanes could be useful for the upper end mainstream/low end enthusiast. Either are essentially irrelevant to the mainstream user as 99.999% of use cases will see almost no benefit. Further, unlike the higher core count which can actually be a detriment to many mainstream use cases over a faster 4c, I don't think extra PCIe lanes is ever anything but a boon (from a performance perspective).

Ok a reddit link of a user who bought a 16x capable system and now makes a youtube vid to show he was right?
I can also find links from users that show regression in sli. Its difficult to meassure correctly. Ashes is a good bm as it gives more consistent results. As was noted in the review i linked. Even in ashes that showed good effect of 1% in that test i linked, a titan x in sli gives mimimal effect at 4k 3.5% and even a slight regression -2% at 1080 in this setup:
https://www.pugetsystems.com/labs/articles/Titan-X-Performance-PCI-E-3-0-x8-vs-x16-851/#Conclusion
Sure if cost is no issue its just a benefit and it gives benefit in cornercases.
Good testing is difficult to come by because no one want to show negative results.

Its the same issue and discussion we had when pci arived and agp was doomed old school. The benefit was way overrated back then and still is. Paper specs.

 

[lolfail9001]

I can also find links from users that show regression in sli.

In which case you would illustrate you have entirely missed the point.

 

[MarkizSchnitzel]

I like to lurk and read these topics, because i like technology. Though i don't know much about it forma technical standpoint.

Still, it's strange for me how little attention is given to the economic realities.

 

[AznAnarchy99]

Installed comp over the weekend with a format. Gave me an instant 40-50 FPS increase in Overwatch. Pretty good.

 

[Sweepr]

Some people asked when Pentium G4560 will be available in the US. ShopBLT updated their ETA, which now mentions February 13. Pre-order price remains $68.97, quite a bit cheaper than Pentium G4600 if you don't need HD Graphics 630.

 

[VirtualLarry]

Seriously, what is the point in delaying the release of the G4560 until the middle of Feb.? Obviously, Intel is playing supply games again. It's not like the Kaby Lake Celerons and Pentium G4600 aren't out already.

Edit: Are they seriously waiting for a supply of G4600 / G4620 CPUs, with defective iGPU EUs, so that they can be "cut down" to HD610 iGPUs?

 

[IntelUser2000]

640k....

Not comparable. The times are different.

Just in 1995, when I started using computers, my Dad saw the need to buy one. If you remember using old MacIntosh at school, they lagged behind your fingers when typing. Most people I know around me were nowhere near I would consider knowledgeable in computers. However, up to a certain point they did upgrade. It was slow for them too!

Core 2 in 2006 is when I started seeing enthusiast forum posters saying "our computers are enough".

Really, if you aren't working in an area related to developing any parts of a computer, or doing computer-aided design, or working on a workstation, or even a hardcore gamer, you really don't care about having a fast PC.

Now, the tech industry is trying to jam "IoT" devices down our throats. Last report I have read about them said people couldn't care less about them. It's a solution looking for a problem. It doesn't solve anything.

 

[richierich1212]

Are these new Kaby Lake pentiums compatible with DDR3 memory?

 

[coercitiv]

Are these new Kaby Lake pentiums compatible with DDR3 memory?

From the G4600 product page.

Memory Types DDR4-2133/2400, DDR3L-1333/1600 @ 1.35V

So it's a yes, just like Sky Lake they support DDR3 memory running at 1.35V

 

[CHADBOGA]

From the G4600 product page.

So it's a yes, just like Sky Lake they support DDR3 memory running at 1.35V

How many motherboards will be offering DDR3 memory slots with 1151 cpu socket?

 

[AtenRa]

How many motherboards will be offering DDR3 memory slots with 1151 cpu socket?

Newegg has close to 30

https://www.newegg.com/Product/Prod...07627 600567584 600036387 600036385 601115270

 

[LTC8K6]

The G4620 seems to be widely available.

 

[CHADBOGA]

Newegg has close to 30

https://www.newegg.com/Product/ProductList.aspx?Submit=ENE&IsNodeId=1&N=100007627 600567584 600036387 600036385 601115270

Thanks for that info.

In Australia, DDR3 motherboards for 1151 are currently very hard to find.

At two of Australia's biggest retailers there were none on offer, I had to check some smaller retailers to find them, after you posted your info.

 

[ShintaiDK]

Seriously, what is the point in delaying the release of the G4560 until the middle of Feb.? Obviously, Intel is playing supply games again. It's not like the Kaby Lake Celerons and Pentium G4600 aren't out already.

Edit: Are they seriously waiting for a supply of G4600 / G4620 CPUs, with defective iGPU EUs, so that they can be "cut down" to HD610 iGPUs?

They are out in Europe and Asia and been for a while. Its just the US market being behind.

 

[VirtualLarry]

They are out in Europe and Asia and been for a while. Its just the US market being behind.

Yep. Cheap chips for "developing markets", but soak the USA customers with more expensive chips. Same MO that Big Pharma uses, with medications half the price being sold in Canada.

 

[LTC8K6]

Yep. Cheap chips for "developing markets", but soak the USA customers with more expensive chips. Same MO that Big Pharma uses, with medications half the price being sold in Canada.

Except they don't even seem to be available from re-sellers looking to profit on the lack of availability in the USA.

 

[coercitiv]

Yep. Cheap chips for "developing markets", but soak the USA customers with more expensive chips.

There there, let's go back to happy thoughts about the Micro Center deals so ridiculously good they make people from developing countries bite their nails off in envy.

 

[ShintaiDK]

Yep. Cheap chips for "developing markets", but soak the USA customers with more expensive chips. Same MO that Big Pharma uses, with medications half the price being sold in Canada.

I dont think you understand the position of the US market.