New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
Careful! Someone may report you for BuyVM shilling. Are you @Traffic?
[/joke]
And 4gb stock is no more.
E5 v1/2 per CPU benchmark point is far superior to E3 up to v4 in power usage, density in 1U is also far higher plus the ability to use 384GB+ RAM (32GB on E3 pre-Skylake) along with up to DDR4 and 40+ PCIe 3.0 lanes (16-24 with E3 pre-Skylake) makes this an Apple and Oranges comparison that is neither correct nor applicable.
I think @Francisco was more concerned with pricing of the CPU, and E5-2690 which gives the desired >3GHz cores is $2000+ but does sport 10 core. Even eBay is ~$400, but I am guessing he has some E3's in inventory without a use so, I think the offering is fine as is and affordable from his marketing end of the test.
Correct, if I put no other customers on that E5 then speedstep would always be in play, but that isn't the case, so we're more likely to operate at the base clock. I could get a 3Ghz+ base clock on an E5 but it'll be 130W/processor and $2000/each. Even still I won't get much out of it since i'd only get maybe 128GB worth of clients on it and i'd be better off just getting E3's anyway since v3's are still very strong work horses. V2's were on my shopping list due to the power savings but the price difference is small enough I can likely wiggle it into our needs and still get some extra horses out of the offering.
Thanks for the interest.
Those are engineering samples which:
Back when we were using L5520's someone sold me a pair of engineering processors and I could never get them working properly. I could never get the 2 of them running in the same board and once I did anything KVM related on them they would catch fire.
Francisco
We'll sit down and crunch things. If we feel it's popular enough we'll get more up quickly.
Just incase someone wants a an answer to this:
If we decide we are happy with this product line then yes, this will be the new pricing/features for our > $10/month KVM's. This isn't introductory pricing if it goes full blast, it'll stay at this point from now on.
Francisco
Would have tried that, if I could, offer looks very interesting.
Have to pass though.
This means I can use the 2 threads 100% all the time?
You got it.
It's fully dedicated, you can use all you want 24/7. If/when we roll the product out officially we'll live migrate you to a new node that'll be the final configuration. A fair bump in CPU performance is on the way
Francisco
Bitcoin mining! :P
If you like.
The threads aren't over subscribed so you can rip them as hard as you want
Francisco
Gonna be that guy. If enough people max out CPU, performance impact will definitely be there since a thread isn't a core. So unless there's a pinning solution (not sure that helps since Linux doesn't really care core vs thread to my knowledge), entirely possible for a few users to be demolished by the actions of others.
Of course, there's possible vs likely, and it's probably very unlikely to be abused. However, I've never administered an environment where customers were told that they were sold dedicated resources in a shared environment, that admittedly may come with increased usage patterns over that of a standard shared environment. Time will give the answer
But that's why I said thread
On an 8 thread box we allocate 8 to customers total. If a node is full of users running infinite loops to be dumb then weights will still keep everything pretty fair.
Even still, the CPUs in place right now are just to prototype the product, I'll message you the proposed final cpu model. Trust me, these plans are going to be killer.
Francisco
Take my money @Francisco , moar moar.
no one complaint the price out of LEB range?
Well, yes - In posting 3 already:
Why would they complain, this is not an offer thread by BuyVM, it is another member passing info along, which has always been acceptable.
Looks like they restocked a few
Yeah, managed to grab 1.
@tommy - As Tim said, I didn't post this thread, someone else did. I know that if I had posted it as an offer it would've been removed since I don't have anything <= $7.00/month listed.
The discussion right now is which CPU we want to roll on the final product. I had one in mind but someone brought up a good point so I'm reevaluating things quickly.
Francisco
May need to use ffmpeg occasionally for 2-3 hours straight
So long as it isn't over 24 hours a day you won't cause any rifts in space.
Have at it.
Francisco
Those plans look like a very nice middle ground for projects which need something better than a tiny VPS where performance is unpredictable and smaller than a full E3 which would be too much.
At this pricing, I guess there will be demand for them.
It's good for people running applications that are heavy single thread based where having more threads means little
Thanks for the feedback.
Francisco
Why buyvm don't have xen? I have good experience with xen, I find their cpu scheduling to be really great. Or is xen a thing of the past and kvm is the future?
I just never liked XEN very much. KVM keeps things very simple to develop
It supports Windows nicely with VIRTIO and that was a big thing for us.
Francisco
I love XEN also but unfortunately Debian Jessie is still not supported as XEN-PV guest.
because someone has used that to bypass LEB restriction
I don't remember which post it might one or two month's ago.
@Francisco ^
My own opinion is that there's a difference between bringing the community's attention to a product that they otherwise would have never heard about (due to LEB pricing), and shilling for a company in order to bypass LEB restrictions. It is only in the latter case that the poster actually has something to gain and can thus be accused of bypassing LEB restrictions.
However, such "bring LET's attention to a cool offer" threads need to be taken on a case-by-case basis where the community should factor in at least the following before making accusations:
(1) the poster's account age e.g. is this a new account? or a veteran member whose been active for a while?
(2) posting history e.g. does the member usually shill? or does he provide valuable contributions elsewhere?
(3) whether the poster gains anything, financial or otherwise, from the offer e.g. are there affiliate links? or any evidence to suggest that the member is affiliated with the company being discussed?
(4) the appropriateness or usefulness of the offer being shared e.g. is this something that the community would honestly find useful or comparable to other good offers?
Unfortunately, LET's present attitude is often to simply get hostile with the poster and immediately accuse the person of shilling regardless of anything else. These days, it's cool to troll even if the poster makes a good logical argument. This basically reduces the discussion down to that which would take place between a logical adult and a toddler.
Unlike others, I don't try to hide the fact that there are certain providers that I enjoy using.
But, to some, my transparency is all the reason that they need to make silly accusations.
I'll be honest in that, due to LET's hostility, I'm highly unlikely to ever bother sharing interesting non-LEB offers with the community again. Fortunately, I don't really come across interesting non-LEB worth-sharing offers too often.
I'm very fond of this community but some of you are really {please refer to @nekki's favorite word}.
As a suggestion, maybe a separate category would be a good place for threads like this?
UPDATE: To be clear, my opinion is that no provider should make posts about >$7 offers anywhere, regardless of whether the offers are his or belong to another provider. However, if it's a normal non-provider LET member, then refer to my case-by-case analysis.
there are lots of history of pum n dump companies in the past, this is a primary reason of skepticism of people in here