New Innosilicon A4+ released 550Mh/s @ 700w

[vinylwasp]

Release note: http://www.innosilicon.com/html/news/19.html

Final price looks like it will be $2946 (based on the same $ per MH/s as the A4).
Read the release note for details.

[Eyedol-X]

Release note: http://www.innosilicon.com/html/news/19.html

Final price looks like it will be $2946 (based on the same $ per MH/s as the A4).
Read the release note for details.

Is there somewhere they've announced or confirmed the final price?

[InGeM]

haha that looks like a modified Antminer

[nukepower]

I have one of the A4 not the plus, I feel you all are gonna get fucked on this so called deal from what i saw in the main post on the bittalk forums they don't make the A4 any more and if you did a per order for this month batch release you may be forced to by a upgrade that they should upgrade you to at there cost for the per orders all ready made and the A4 was suppose to run at the plus version speeds at the price of the A4 i also feel they made the A4 run like it did for one reason money an they knew bitmain was coming out with a new version of there L1,which bitmain never sold to the public till the L3 ..I can't prove this but i get the feeling they did have the L1 .I'm' Aware of how they claimed they didn't make the L1 , any way I don't much believe any thing any more form any of those company's till i see it ...

[vinylwasp]

@Eyedol-X - As it says in the post the final price will be based on the final speed multiplied by by this factor (final speed/280) * $1500 (the current public price of an A4) which at 550Mh/s works out to be $2946 If they squeeze more out of it (they told me they expect 10% more in the final build) you'll still be paying $5.35 per MH/s using that formula.

@nukepower - The big design difference between an A4 and a L3+ is that Inno uses a parallel DC circuit design and Bitmain uses a serial where all the chips on a single board are daisy chained together. This is really efficient but has some weaknesses as all the current runs through one circuit with the most amps (and therefore heat ) at the start. With the A4 each chip has it's own power circuit, so there's loads of built in redundancy and hence reliability, but it's a lot more power hungry. I think the big difference with the A4+ will be that Inno have moved to a serial DC design, but have hopefully done it in such a way that they've retained their bullet proof reliability.

I'll let you know when mine arrive.

[nukepower]

well see and they did say at first the A4 was suppose to run @

A4 ASIC: 3Mhs per chip at 3.6W nominal, support DCDC less PCB
Extreme Low Power: daisy chained ASIC running at 0.7V, only in 14nm
Miner Performance: 600Mhs from 720W at the Wall or 850Mhs from 1020W
PSU Requirements: common PCI-E 8Pins/6+2Pins with 12V output
Included: Raspberry Pi running the most user friendly A2 terminator Software


your right the A2 did rock

[vinylwasp]

well see and they did say at first the A4 was suppose to run @

A4 ASIC: 3Mhs per chip at 3.6W nominal, support DCDC less PCB
Extreme Low Power: daisy chained ASIC running at 0.7V, only in 14nm
Miner Performance: 600Mhs from 720W at the Wall or 850Mhs from 1020W
PSU Requirements: common PCI-E 8Pins/6+2Pins with 12V output
Included: Raspberry Pi running the most user friendly A2 terminator Software


your right the A2 did rock

Yes those pre-production targets were based on the performance of a single chip in the lab, and yes they would have delivered on that promise possibly if they'd opted to use a serial DC design. As we know they didn't and the power consumption was 3x what they promised with the associated increase in heat whihc required lower chip density and less MH/s per miner.

I've got quite a few A4's. They rock too. If you have a fan failure and they get hot, the inbuilt thermal protection kicks in and the card will shut itself down until a full power cycle and the temp has dropped. Try that with an L3.

They may have missed those target specs but try finding one post anywhere where an A4 hashcard has died, they're totally bulletproof.

[nukepower]

I discovered something about my One A4 half tube I have if I run it though a proxy some thing like the one Mining Rig Rentals lets us use, .. i can run mine on any pool at the right speeds .. i do have issue with it on some pools but if I run it though a proxy it runs like it should on any pool ... and Innosilicon has told me they are a ware of some pools under hashing and were trying to fix it with software which might not happen now for the normal A4 or the first A4.. and I'm looking to buy one of the orange pi one controllers some got with there A4, i got one of the old RPI which might be some of the cause, no I'm sure it is ....


well any way i may be able to buy one or two A4 + or what ever is out some time next year i will have a lot of bills and CC's paid off by July 2018... can't wait to free up that money for mining gear ... and max out my houses power ... that's my limit and goal .

[hashingpro]

Is Innosilicon run by the same asshats that ran KNC Miners?
Coming out with a new hardware, forcing customers who prepaid to pay more for higher machine than what they ordered?

That sounds awefuly familiar to me.

[nukepower]

I don't think so Innosilicon makes more then miners,I don't think they want to control the network like KNC did which was knc's down fall ....
I know Innosilicon does not have there own Mining farm or didn't before the A4 ... but yea they are fast becoming jerks like the guys at bitmain and like KNC was .... even Avalon who sells a better BTC miner then bitmain are turning into ass holes.'

Can't wait til next year when we may get a 7 nm chip from Japan a Free nation which may be very good for the Free world, once again.

Very bad for china's miner market. we can only hope ..

Innosilicon is a china based company an been around longer then KNC,is was .

I think Innosilicon started in the memory making business first.