As an intel shareholder I am so glad they have successful hid the 2.5 gig ethernet IP from the consumer market for the past 5 years. Clearly that would put their data center business at risk .... History shows that this is the best way to maximize profits and it also involves no risk of losing core business to companies with 10% as many employees
Lol, but seriously, why hasn't Intel jumped on the NBASE-T bandwagon yet ? Obviously we need to get more switches and stuff supported but yeah 2.5/5GB is perfect for home use and should have been rolled out years ago. Sheesh.
WIFI is a complete non-starter for gaming. 1ms latency from PC to router becomes 30 ms on average and 90ms occasionally when there is re-transmission. slow and inconsistent. i'd rather be dead in a ditch than play csgo on wifi.
As for the 2nd part. I really don't know. Intel has had this IP in their library for many years. Its not a large piece of sillicon either. If I am not wrong Aquentia had pure CMOS 10GBASE-T controllers back in 65nm days. NBASE-T in 45nm or 28nm.
On bigger lithographies 10GBASE-T controllers ran hot enough they needed substantial heat sinks to keep cool. A few years back the estimate was that heatsinkless controllers would be possible in one or two process shrinks. I don't recall if that was 14 or 10 anymore. On Intel's side neither are really an option, 10 is apparently never going to be a volume tech for them and it's delays mean 14 is fully tied up on CPUs. For Samsung/TSMC with high performance/dollar parts moving to 7nm the capacity on 14/12/10 should be opening up for more cost sensitive customers.
Otherwise we seem to be stuck at the chicken/egg problem. No cheap 2.5/5/10g consumer routers means no demand for matching cheap 2.5/5/10g consumer network adapters and vice version. At least on Intel's side it's possible the plan was to have multi-gig ethernet support baked into chipsets by now; but that like everything else their tight coupling of process tech to hardware design and cpu to chipset models has resulted in the 10nm debacle preventing them from coming out. I'm really eager to see what the chipset for their upcoming 10nm Xeon's will support, since integration there would be a harbinger of things to come on the consumer side. AMD's only just gone from a fully outsourced chipset to a nominally in house one mostly full of 3rd party IP, so not having any sort of native multi-gig yet doesn't surprise me. Again it'll be interesting to see what happens with their next one; will they stay with needing a full up controller over PCIe or do what Intel's done and pull a large chunk of it into the chipset to encourage using their branded networking parts to finish the setup.
Intel 4 core Atom Xeon with passive cooling are SoC with 10Gb integrated into their 14nm. The dual 10Gb plus 4 cores running at max are passively cooled, not to mention the chip plus motherboard are only $200. You can get a 14nm SoC Xeon with integrated dual 10Gb NICs for the same price as a single 20nm class 10Gb NIC. That's the difference between 20nm class and 10nm class for networking.
The first generations of 10GBASE-T chips was actually produced on the 130 nm and later 90 nm nodes, certainly those that were used 2 process-generations earlier for actual CPUs.
Afaik Intel has moved to 65 nm (also a node known from CPU production), where they still had some issues.
Aquantia however uses the now-available and ubiquitous 28 nm facilities (I assume from former GPU production lines) and does seem to have gotten the power consumption to acceptable levels.
I'm fairly certain that Intel will be able to spare some 14 nm capacities in the near future and will likely move the NICs there.
Aquantia certainly made 10GBASE-T reasonably priced and it will only get better. The move of 10 Gbit/s into mainstream isn't far away anymore. Mid-range server boards already include it quite often, and I expect that we will see it becoming common in HEDT in 2020.
I went from a $300 Netgear router+AP to $100 Ubiquiti AP and my latency was cut in half, signal strength went up, packetloss and drop outs disappeared. I thought wifi just sucked. My roku express, located several walls and a floor different, kept losing signal and buffering. The ISPs have been pushing gateways with built in wifi and has increased the noise. It was getting too much to deal with. I gave Ubiquiti a try because I heard they were "enterprise" grade instead of "lemon" grade. Hoooolllllyyyy crap. It's a difference. Or at least it has been for me.
Why should they invest in 2.5/5g at all? 10gb gear is available for cheap now on Ebay and has been in the enterprise market for a decade or more now. They should be able to very cheaply put a few 10gb ports on a switch or high end router. I setup a 10gb *fiber* network at home nearly 10 years ago for cheap (infiniband is even cheaper). Its just crazy we haven't seen 10gb at least for soho stuff like drobo and similar storage devices.
Its so far behind now that with gigabit internet I can upload to cloud storage like amazon drive almost as quickly as I could upload to local NAS storage!
After reading the wikis for these standards, it seems to be due to the cabling requirements of 10GBASE-T. For clients who have large infrastructures already wired with Cat 5e, they can still use the intermediate speeds without rewiring everything.
This. A lot of cabling obstacles leave 5Gb as a good median. 10Gb requires ideal cable quality and short distances. I often see a lot of implantations handshake back down to 5Gb and that’s often the best that can be done within a budget without pulling new cable throughout an office.
In the end that’s still 5x the throughout so if you are just looking to speed up loading a 400mb quickbooks file or something there will be virtually no noticeable difference between 5Gb and 10Gb
I wish people would stop saying "10gb is cheap! Just buy used!" I don't want a 48 port switch that will sound like a 747 in my office. That is *not* a solution.
I doubt you need 48 ports in your office? ..Unless your office is in the server room? Mellanox ConnectX-3 adapters can be had on ebay now from $22, so yeah, even 40/56 Gbit is cheap. Good for a point-to-point with your local SAN/NAS/whatever.
Don't be an idiot. His entire point is that he CANNOT BUY a home-appropriate 10G (or hell, 2.5G) switch. They only come in industrial models -- LOTS of ports, big, strong fans, rack mounted.
If you know different, point us to a home-appropriate 10G switch, say 5 ports, no fan, basically size of a paperback book -- just like a 1G switch...
The point is not the inability to buy NICs (or, if you're living in the 21st century, TB or USB-C adaptors), the point is the SWITCH!
The closest I have found is the Netgear MS510TX ($200) and MS510TX-PP ($350) switches.
Each has SFP+: (1) 10G/1G port, and RJ45: (1) 10G/5G/2.5G/1G/100M port, (2) 5G/2.5G/1G/100M ports, (2) 2.5G/1G/100M ports, and (4) 1G/100M/10M ports.
The MS510TX-PP version has PoE+ available on all eight of the RJ45 ports.
I don't have one yet, but I've been looking at a purchase in the near future. Neither one is exactly quiet, with the non-PoE version rated at 21dB and the PoE version at 28.8dB. But they do fit in the small/home business price range.
Price is not the whole point. Yes, they're "affordable". But they're too large, and too loud, at least in part because they offer too much. And THAT is what people like me are complaining about.
Yeah it might be fine if you are living in your parents basement but wives and such like take a dim view to enterprise kit in the living room. Rightly so.
Won't sound like a 747 for long. Those high pressure fans will clog. I only go fan-less. No dust bunnies, just a fine coat of dust. Anything with a fan gets a winter coat of fur in no time.
Businesses tend to want warranties and support with their devices and thus the eBay route is generally not an option. Those businesses are generally the ones who are flooding eBay with equipment to sell: as soon as they're out of warranty/support contract, the hardware gets moved off to a recycler who often posts it on eBay.
As for 2.5 Gbit, that data rate can be had effectively 'for free' as it is changing the symbol space on the existing 1 Gbit encoding (only 100 of 256 are used). If 1 Gbit works on existing cabling, 2.5 Gbit will as well. Re-running cable is not fun, especially if moving from vanilla Cat5e to something more robust like Cat6A STP or Cat7.
Fiber costs have come down considerably but the cabling remain more fragile than copper and outside of the data center, power-over-cable is popular for many devices.
Because 10gb requires rewiring, either to fiber or cat6 cables. Plus most consumers don't have fiber internet and hence asynchronous connection speed with upload usually being mostly unusable for any large file transfers.
Because the average consumer (read most of them) doesn't use Ethernet at all. The market is too small. And myself that has a lot of media files, I really don't need it. it's not like I'm copying the files around between machines. The get consumed and 1gb is more than enough for that. not to mention that most hdds can't even continuously saturated 1gb.
Oh and i forogt with many providers you get an all-in one device which is modem, router, switch and wifi AP in one. often with config locked down and obviously a 1gb switch. To setup a 2.5/5 GB network you will net to change that device into a pure modem (if possible at all) and buy your own 2.5gb wifi router. here 95% of people would already fail...and then you need 2.5gb on all your devices or at least your PC and your NAS. Show me a consumer NAS with 2.5gb LAN.
Synology DS1620xs, and I believe other 2020 models will be coming with 2.5Gb as well. Along with an Asus GT-AX11000 router, or Netgear Nighthawk has 2.5Gb. This would work well if using wifi6. Modules are cheap and easy to buy.
Personally I would just jump to 10Gb as that can be put into Synology and many Qnap's already have 10Gb built in.
The problem with 10Gb though is that most adapters will only work at 10 or 1, and not step in-between, so planning is essential.
The best you can do today, as far as I can tell is - do your own cable modem - get a Nighthawk AX12 which will give you WiFi6, 2 USB3 ports (not C...), and on the LAN side + one 5/2.5G port and + 4 1G ports + of which two can be aggregated to give you a 2x1G
Not a great set of options there, but still MAYBE (depending on your house's circumstances) can be converted into something useful. The real shame is that they didn't work just a little bit harder to give two multi-gig ports rather than one, which would have been so much more useful...
"Why is the industry afraid of investing in 2.5/5G devices?"
Because the consumer market for them is utterly minuscule. There are vanishingly few home use cases where more than 1gbps is actally necessary, and the prices for even 'cheap' 2.5/5g kit is still many times (10x or more) the price of the commodity network gear people purchase in volume, and that companies like ISPs deploy at volume. That leaves the slice of the home enthusiast market that want more than 1gbps, but who also aren't willing to go to 10gbps.
You have to attack the chicken and egg problem from one end or the other. How many years was it that Intel was churning out PC chipsets with gigabit support before the first home gear appeared to take advantage of it? I seem to recall it was a good 4 or 5 years. Back then everyone said that gigabit switches were too expensive, too hot, too noisy, only for business, yada yada. Sound familiar? Now anything limited to 100M would be laughed out of the room.
If Intel has rights to (or owns) multi-gig IP, they should damn well start putting it in every chipset they put out. Same for AMD, Broadcom, or anyone else who puts Ethernet into their chips. A 250% to 500% link speed improvement, that works over existing cabling, for minimal cost, should be an obvious feature to support.
Intel was putting gbit into their chipsets for the enterprise market, where gbit core switches and routers were available (and provided a definite increase in performance for remote storage and remote sessions). The same is absolutely not the case for 2.5g/5g/etc: no business is swapping out their network kit for that capability, all the capacity increases are going on from the switch 'upwards' where 10g and up are the new norm. Putting 2.5g/5g into chipsets inflates the cost, but has no necessity and little to no uptake.
yes but... we can actually use the bandwith increase that came from 100M to 1G. Other than direct files transfers when would you use more than 1G of bandwidth? Hard drives can't keep up with it and most people have 50M or less internet connections. I'm all for overkill but we seem to be a long way from "needing" more than 1G for normal home use, and as mentioned the average user is wifi all around. Enterprise... different story of course.
A bit of an oddity but Intel has leveraged 2.5 Gbit Ethernet on various Atom chip for years now. The catch is that that 2.5 Gbit interface adheres to backplane spec, not Cat cabling. This was mainly for industrial/embedded usage.
Getting Lightbit onboard is good...that pretty much guarantees full investment at EMC. Love the cute heatsinks on the HBA ports too. I've noticed a lot of those media converters for hot af and can only imagine that'll be more of an issue as things go beyond 10Gb, so that's a creative solution.
Does it still grind to Kbps speeds when it hits tens of thousands of micro files? Oh sure you can transfer 8K video super fast but back in the real world...
This is the issue we need to address in data, not actual amounts of total bandwidth but how do we deal with modern software's need to populate itself with millions of micro files.
What you are describing is an application layer issue, due to start/ stop operation, probably because you used Windows explorer to try to copy files.
In the real world this is for servers running vm's with hundreds of connections, and also for storage connections. Not to transfer a lot of tiny files.
"supports up to a PCIe 4.0 x16 host interface—to be expected from a 100Gb NIC, but not something Intel PR is keen to highlight while their CPUs are still on PCIe 3.0." Lol quite we don't want people to know we support the competition because we are inferior until....
NVME over TCP? Why does this exist? To me, it sounds like a security breach waiting to happen. At NVME speeds. Add in Intel's recent track record on architectural security....
Well, it was only a few weeks ago that news broke about doing cache timing attacks over RDMA with Intel NICs that support DDIO (DMA to L3 instead of DRAM). So NVMe over TCP sounds like it might be more secure than NVMe over RDMA in some cases. But in practice, most deployments and all the really high-performance deployments of this will be on fairly isolated/locked-down networks.
What I would like to see is a way to jump between QSFP28 and SFP+ connections.
The cost per port to go from 1 GbE to 10 GbE is actually HIGHER than it is to go from 1 GbE to 100 Gbps (be it either GbE or IB).
The problem is that 10GbE is coming online for a lot more systems and devices (either RJ45 or SFP+) and right now, I don't have a way to bridge/jump between those speeds in a single switch.
Regarding a switch for home use, the Mikrotik CRS305 is doing pretty well for me. ~$140, fanless, 4 SFP+ ports for 10gb, and an RJ45 at 1gb for management and/or another switch port.
Gives me 10gb for a desktop PC, NAS, VM host, and an uplink to a different switch that handles my 1gb and PoE devices.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
48 Comments
Back to Article
azfacea - Tuesday, September 24, 2019 - link
As an intel shareholder I am so glad they have successful hid the 2.5 gig ethernet IP from the consumer market for the past 5 years. Clearly that would put their data center business at risk .... History shows that this is the best way to maximize profits and it also involves no risk of losing core business to companies with 10% as many employeesextide - Tuesday, September 24, 2019 - link
You forgot the /sLol, but seriously, why hasn't Intel jumped on the NBASE-T bandwagon yet ? Obviously we need to get more switches and stuff supported but yeah 2.5/5GB is perfect for home use and should have been rolled out years ago. Sheesh.
Ian Cutress - Tuesday, September 24, 2019 - link
I was told that current 2.5G solutions had a two year lead in the market. Three years ago. /shrugmooninite - Tuesday, September 24, 2019 - link
So the future of Ethernet appears to be dead. Lackluster Wi-fi 6 products are what we have to live with.Why is the industry afraid of investing in 2.5/5G devices?
azfacea - Tuesday, September 24, 2019 - link
WIFI is a complete non-starter for gaming. 1ms latency from PC to router becomes 30 ms on average and 90ms occasionally when there is re-transmission. slow and inconsistent. i'd rather be dead in a ditch than play csgo on wifi.As for the 2nd part. I really don't know. Intel has had this IP in their library for many years. Its not a large piece of sillicon either. If I am not wrong Aquentia had pure CMOS 10GBASE-T controllers back in 65nm days. NBASE-T in 45nm or 28nm.
Alistair - Tuesday, September 24, 2019 - link
dead in a ditch, hehebrunis.dk - Wednesday, September 25, 2019 - link
I'm guessing by Aquentia, you mean Aquantia?DanNeely - Wednesday, September 25, 2019 - link
On bigger lithographies 10GBASE-T controllers ran hot enough they needed substantial heat sinks to keep cool. A few years back the estimate was that heatsinkless controllers would be possible in one or two process shrinks. I don't recall if that was 14 or 10 anymore. On Intel's side neither are really an option, 10 is apparently never going to be a volume tech for them and it's delays mean 14 is fully tied up on CPUs. For Samsung/TSMC with high performance/dollar parts moving to 7nm the capacity on 14/12/10 should be opening up for more cost sensitive customers.Otherwise we seem to be stuck at the chicken/egg problem. No cheap 2.5/5/10g consumer routers means no demand for matching cheap 2.5/5/10g consumer network adapters and vice version. At least on Intel's side it's possible the plan was to have multi-gig ethernet support baked into chipsets by now; but that like everything else their tight coupling of process tech to hardware design and cpu to chipset models has resulted in the 10nm debacle preventing them from coming out. I'm really eager to see what the chipset for their upcoming 10nm Xeon's will support, since integration there would be a harbinger of things to come on the consumer side. AMD's only just gone from a fully outsourced chipset to a nominally in house one mostly full of 3rd party IP, so not having any sort of native multi-gig yet doesn't surprise me. Again it'll be interesting to see what happens with their next one; will they stay with needing a full up controller over PCIe or do what Intel's done and pull a large chunk of it into the chipset to encourage using their branded networking parts to finish the setup.
bcronce - Wednesday, September 25, 2019 - link
Intel 4 core Atom Xeon with passive cooling are SoC with 10Gb integrated into their 14nm. The dual 10Gb plus 4 cores running at max are passively cooled, not to mention the chip plus motherboard are only $200. You can get a 14nm SoC Xeon with integrated dual 10Gb NICs for the same price as a single 20nm class 10Gb NIC. That's the difference between 20nm class and 10nm class for networking.IamSorryDave - Wednesday, September 25, 2019 - link
Which CPU+mb for $200 were you referring to? The ones (with C3xxx) I am aware of are double that in price.thomasg - Tuesday, October 1, 2019 - link
The first generations of 10GBASE-T chips was actually produced on the 130 nm and later 90 nm nodes, certainly those that were used 2 process-generations earlier for actual CPUs.Afaik Intel has moved to 65 nm (also a node known from CPU production), where they still had some issues.
Aquantia however uses the now-available and ubiquitous 28 nm facilities (I assume from former GPU production lines) and does seem to have gotten the power consumption to acceptable levels.
I'm fairly certain that Intel will be able to spare some 14 nm capacities in the near future and will likely move the NICs there.
Aquantia certainly made 10GBASE-T reasonably priced and it will only get better.
The move of 10 Gbit/s into mainstream isn't far away anymore.
Mid-range server boards already include it quite often, and I expect that we will see it becoming common in HEDT in 2020.
bcronce - Wednesday, September 25, 2019 - link
I went from a $300 Netgear router+AP to $100 Ubiquiti AP and my latency was cut in half, signal strength went up, packetloss and drop outs disappeared. I thought wifi just sucked. My roku express, located several walls and a floor different, kept losing signal and buffering. The ISPs have been pushing gateways with built in wifi and has increased the noise. It was getting too much to deal with. I gave Ubiquiti a try because I heard they were "enterprise" grade instead of "lemon" grade. Hoooolllllyyyy crap. It's a difference. Or at least it has been for me.Bp_968 - Tuesday, September 24, 2019 - link
Why should they invest in 2.5/5g at all? 10gb gear is available for cheap now on Ebay and has been in the enterprise market for a decade or more now. They should be able to very cheaply put a few 10gb ports on a switch or high end router. I setup a 10gb *fiber* network at home nearly 10 years ago for cheap (infiniband is even cheaper). Its just crazy we haven't seen 10gb at least for soho stuff like drobo and similar storage devices.Its so far behind now that with gigabit internet I can upload to cloud storage like amazon drive almost as quickly as I could upload to local NAS storage!
Assimilator87 - Tuesday, September 24, 2019 - link
After reading the wikis for these standards, it seems to be due to the cabling requirements of 10GBASE-T. For clients who have large infrastructures already wired with Cat 5e, they can still use the intermediate speeds without rewiring everything.Samus - Wednesday, September 25, 2019 - link
This. A lot of cabling obstacles leave 5Gb as a good median. 10Gb requires ideal cable quality and short distances. I often see a lot of implantations handshake back down to 5Gb and that’s often the best that can be done within a budget without pulling new cable throughout an office.In the end that’s still 5x the throughout so if you are just looking to speed up loading a 400mb quickbooks file or something there will be virtually no noticeable difference between 5Gb and 10Gb
mooninite - Wednesday, September 25, 2019 - link
I wish people would stop saying "10gb is cheap! Just buy used!" I don't want a 48 port switch that will sound like a 747 in my office. That is *not* a solution.brunis.dk - Wednesday, September 25, 2019 - link
I doubt you need 48 ports in your office? ..Unless your office is in the server room? Mellanox ConnectX-3 adapters can be had on ebay now from $22, so yeah, even 40/56 Gbit is cheap. Good for a point-to-point with your local SAN/NAS/whatever.Namisecond - Wednesday, September 25, 2019 - link
Client adapters may be cheap, but what about the switches? I mean real switches with configurable ports. 16-port minimum.name99 - Wednesday, September 25, 2019 - link
Don't be an idiot. His entire point is that he CANNOT BUY a home-appropriate 10G (or hell, 2.5G) switch.They only come in industrial models -- LOTS of ports, big, strong fans, rack mounted.
If you know different, point us to a home-appropriate 10G switch, say 5 ports, no fan, basically size of a paperback book -- just like a 1G switch...
The point is not the inability to buy NICs (or, if you're living in the 21st century, TB or USB-C adaptors), the point is the SWITCH!
mark625 - Wednesday, September 25, 2019 - link
The closest I have found is the Netgear MS510TX ($200) and MS510TX-PP ($350) switches.Each has SFP+: (1) 10G/1G port, and RJ45: (1) 10G/5G/2.5G/1G/100M port, (2) 5G/2.5G/1G/100M ports, (2) 2.5G/1G/100M ports, and (4) 1G/100M/10M ports.
The MS510TX-PP version has PoE+ available on all eight of the RJ45 ports.
I don't have one yet, but I've been looking at a purchase in the near future. Neither one is exactly quiet, with the non-PoE version rated at 21dB and the PoE version at 28.8dB. But they do fit in the small/home business price range.
Here's the link: https://www.netgear.com/business/products/switches...
name99 - Wednesday, September 25, 2019 - link
Price is not the whole point.Yes, they're "affordable". But they're too large, and too loud, at least in part because they offer too much. And THAT is what people like me are complaining about.
What we want is something like
https://www.amazon.com/gp/product/B000N99BBC/
just with every port replaced with 2.5G
And the point is not the cost; I'd be happy to pay 10x the cost of the 1G version.
But I want that size, that power budget, that simplicity.
jabber - Wednesday, September 25, 2019 - link
Yeah it might be fine if you are living in your parents basement but wives and such like take a dim view to enterprise kit in the living room. Rightly so.bcronce - Wednesday, September 25, 2019 - link
Won't sound like a 747 for long. Those high pressure fans will clog. I only go fan-less. No dust bunnies, just a fine coat of dust. Anything with a fan gets a winter coat of fur in no time.Samus - Wednesday, September 25, 2019 - link
True, most of the 10Gb switches are pretty quiet unless they are also PoE+ and under heavy load. Any modern switch had load-sensitive fans anyway.Kevin G - Wednesday, September 25, 2019 - link
Businesses tend to want warranties and support with their devices and thus the eBay route is generally not an option. Those businesses are generally the ones who are flooding eBay with equipment to sell: as soon as they're out of warranty/support contract, the hardware gets moved off to a recycler who often posts it on eBay.As for 2.5 Gbit, that data rate can be had effectively 'for free' as it is changing the symbol space on the existing 1 Gbit encoding (only 100 of 256 are used). If 1 Gbit works on existing cabling, 2.5 Gbit will as well. Re-running cable is not fun, especially if moving from vanilla Cat5e to something more robust like Cat6A STP or Cat7.
Fiber costs have come down considerably but the cabling remain more fragile than copper and outside of the data center, power-over-cable is popular for many devices.
beginner99 - Wednesday, September 25, 2019 - link
Because 10gb requires rewiring, either to fiber or cat6 cables. Plus most consumers don't have fiber internet and hence asynchronous connection speed with upload usually being mostly unusable for any large file transfers.beginner99 - Wednesday, September 25, 2019 - link
Because the average consumer (read most of them) doesn't use Ethernet at all. The market is too small. And myself that has a lot of media files, I really don't need it. it's not like I'm copying the files around between machines. The get consumed and 1gb is more than enough for that. not to mention that most hdds can't even continuously saturated 1gb.beginner99 - Wednesday, September 25, 2019 - link
Oh and i forogt with many providers you get an all-in one device which is modem, router, switch and wifi AP in one. often with config locked down and obviously a 1gb switch. To setup a 2.5/5 GB network you will net to change that device into a pure modem (if possible at all) and buy your own 2.5gb wifi router. here 95% of people would already fail...and then you need 2.5gb on all your devices or at least your PC and your NAS. Show me a consumer NAS with 2.5gb LAN.Dug - Wednesday, September 25, 2019 - link
Synology DS1620xs, and I believe other 2020 models will be coming with 2.5Gb as well. Along with an Asus GT-AX11000 router, or Netgear Nighthawk has 2.5Gb.This would work well if using wifi6. Modules are cheap and easy to buy.
Personally I would just jump to 10Gb as that can be put into Synology and many Qnap's already have 10Gb built in.
The problem with 10Gb though is that most adapters will only work at 10 or 1, and not step in-between, so planning is essential.
name99 - Wednesday, September 25, 2019 - link
The best you can do today, as far as I can tell is- do your own cable modem
- get a Nighthawk AX12 which will give you WiFi6, 2 USB3 ports (not C...), and on the LAN side
+ one 5/2.5G port and
+ 4 1G ports
+ of which two can be aggregated to give you a 2x1G
Not a great set of options there, but still MAYBE (depending on your house's circumstances) can be converted into something useful.
The real shame is that they didn't work just a little bit harder to give two multi-gig ports rather than one, which would have been so much more useful...
edzieba - Wednesday, September 25, 2019 - link
"Why is the industry afraid of investing in 2.5/5G devices?"Because the consumer market for them is utterly minuscule. There are vanishingly few home use cases where more than 1gbps is actally necessary, and the prices for even 'cheap' 2.5/5g kit is still many times (10x or more) the price of the commodity network gear people purchase in volume, and that companies like ISPs deploy at volume. That leaves the slice of the home enthusiast market that want more than 1gbps, but who also aren't willing to go to 10gbps.
mark625 - Wednesday, September 25, 2019 - link
You have to attack the chicken and egg problem from one end or the other. How many years was it that Intel was churning out PC chipsets with gigabit support before the first home gear appeared to take advantage of it? I seem to recall it was a good 4 or 5 years. Back then everyone said that gigabit switches were too expensive, too hot, too noisy, only for business, yada yada. Sound familiar? Now anything limited to 100M would be laughed out of the room.If Intel has rights to (or owns) multi-gig IP, they should damn well start putting it in every chipset they put out. Same for AMD, Broadcom, or anyone else who puts Ethernet into their chips. A 250% to 500% link speed improvement, that works over existing cabling, for minimal cost, should be an obvious feature to support.
edzieba - Thursday, September 26, 2019 - link
Intel was putting gbit into their chipsets for the enterprise market, where gbit core switches and routers were available (and provided a definite increase in performance for remote storage and remote sessions). The same is absolutely not the case for 2.5g/5g/etc: no business is swapping out their network kit for that capability, all the capacity increases are going on from the switch 'upwards' where 10g and up are the new norm.Putting 2.5g/5g into chipsets inflates the cost, but has no necessity and little to no uptake.
Icehawk - Wednesday, October 2, 2019 - link
yes but... we can actually use the bandwith increase that came from 100M to 1G. Other than direct files transfers when would you use more than 1G of bandwidth? Hard drives can't keep up with it and most people have 50M or less internet connections. I'm all for overkill but we seem to be a long way from "needing" more than 1G for normal home use, and as mentioned the average user is wifi all around. Enterprise... different story of course.Kevin G - Wednesday, September 25, 2019 - link
A bit of an oddity but Intel has leveraged 2.5 Gbit Ethernet on various Atom chip for years now. The catch is that that 2.5 Gbit interface adheres to backplane spec, not Cat cabling. This was mainly for industrial/embedded usage.Phynaz - Tuesday, September 24, 2019 - link
You get dumber every day.Samus - Wednesday, September 25, 2019 - link
Getting Lightbit onboard is good...that pretty much guarantees full investment at EMC. Love the cute heatsinks on the HBA ports too. I've noticed a lot of those media converters for hot af and can only imagine that'll be more of an issue as things go beyond 10Gb, so that's a creative solution.jabber - Wednesday, September 25, 2019 - link
Does it still grind to Kbps speeds when it hits tens of thousands of micro files? Oh sure you can transfer 8K video super fast but back in the real world...This is the issue we need to address in data, not actual amounts of total bandwidth but how do we deal with modern software's need to populate itself with millions of micro files.
Billy Tallis - Wednesday, September 25, 2019 - link
Files only exist at a higher level of abstraction than NVMe, so doing NVMe over Fabrics doesn't change much about that issue.Dug - Wednesday, September 25, 2019 - link
What you are describing is an application layer issue, due to start/ stop operation, probably because you used Windows explorer to try to copy files.In the real world this is for servers running vm's with hundreds of connections, and also for storage connections. Not to transfer a lot of tiny files.
godrilla - Wednesday, September 25, 2019 - link
"supports up to a PCIe 4.0 x16 host interface—to be expected from a 100Gb NIC, but not something Intel PR is keen to highlight while their CPUs are still on PCIe 3.0." Lol quite we don't want people to know we support the competition because we are inferior until....The highlight of the story.
dumanfu - Wednesday, September 25, 2019 - link
As PCIe 4.0, it is missing a "designed for AMD Epyc" logoLOL
Namisecond - Wednesday, September 25, 2019 - link
NVME over TCP? Why does this exist? To me, it sounds like a security breach waiting to happen. At NVME speeds. Add in Intel's recent track record on architectural security....Billy Tallis - Wednesday, September 25, 2019 - link
Well, it was only a few weeks ago that news broke about doing cache timing attacks over RDMA with Intel NICs that support DDIO (DMA to L3 instead of DRAM). So NVMe over TCP sounds like it might be more secure than NVMe over RDMA in some cases. But in practice, most deployments and all the really high-performance deployments of this will be on fairly isolated/locked-down networks.A5 - Wednesday, September 25, 2019 - link
"I don't understand something so it is useless" is a really bad look.I know you were just setting up for your big lolz hot take, but you just look ignorant.
Dug - Wednesday, September 25, 2019 - link
You need to look at this-https://www.flashmemorysummit.com/English/Collater...
Maybe that will help you understand.
alpha754293 - Friday, September 27, 2019 - link
What I would like to see is a way to jump between QSFP28 and SFP+ connections.The cost per port to go from 1 GbE to 10 GbE is actually HIGHER than it is to go from 1 GbE to 100 Gbps (be it either GbE or IB).
The problem is that 10GbE is coming online for a lot more systems and devices (either RJ45 or SFP+) and right now, I don't have a way to bridge/jump between those speeds in a single switch.
MajesticTrout - Monday, September 30, 2019 - link
Regarding a switch for home use, the Mikrotik CRS305 is doing pretty well for me. ~$140, fanless, 4 SFP+ ports for 10gb, and an RJ45 at 1gb for management and/or another switch port.Gives me 10gb for a desktop PC, NAS, VM host, and an uplink to a different switch that handles my 1gb and PoE devices.