Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Our $100M Series B

oxide.computer

Post reply on HN
Print view View original

341–350 of 522 posts

Re: Our $100M Series B

#341

Post by steveklabnik »

keeda wrote: post #334
zer00eyz wrote: post #279

Earlier quoted context omitted.

> Do the people making buying decisions at this scale care if their sysadmins have better tools? Look at who oxide is selling to and for what reasons. It's about compute + software at rack scales. It does not matter if it is good it matters that it's integrated. Gear at this level is getting sold with a service contract and "good" means you dont have to field as many calls (keeping the margins up). > Everything we he…

As someone only tangentially familiar with this domain, I have questions about this: > Look at their CPU density and do the math on power. It's fairly low density. Look at the interconnects (100gb per system). Also fairly conservative. It's the perfect product to replace hardware that is aging out, as you wont have to re-plumb for more power/bandwidth, and you still get a massive upgrade. It sounds like the CPU densi…

I am not saying that I fully endorse the characterization of the parent, but it is true that we started selling these systems two years ago, and new hardware comes out with better stats all the time.

Given how small we are, new designs and refreshes take a while. Part of growing as a company is being able to do this more often. We'll get there :)

Re: Our $100M Series B

#342

Post by yencabulator »

TZubiri wrote: post #96
master_crab wrote: post #74

Earlier quoted context omitted.

I don’t disagree that there are some fat margins in the cloud, but how is vendor lock-in any different here? Companies could end up paying fat margins to oxide too while still managing physical gear and plant.

Well, their servers are mostly hypervisors, so the interface is mostly any virtual machine. You can "just" migrate by exporting or importing the vms.

You're literally ignoring that the Oxide management stack is very much custom and effectively vendor-locking the purchase to be maintained by them. They are not general purpose PC servers.

You can "just" migrate away from Oxide but that would mean throwing away the hardware you now own. That's the grandparent's point; if you're migrating out of a cloud to avoid the margins demanded by the cloud vendor, now you're at the mercy of whatever Oxide thinks your support contract is worth.

Sure, the convenience may be worth it, but watch how many companies are now struggling to get off of VMWare after Broadcom moves.

Re: Our $100M Series B

#343

Post by steveklabnik »

jeffbee wrote: post #303
bcantrill wrote: post #212

Earlier quoted context omitted.

We didn't use anything from OCP. When we first started the company, we thought we might use the enclosure (and considered ourselves "OCP inspired"), but there ended up being little value in doing so (and there was a clear cost). And on the stuff that we really cared about (e.g., getting rid of the traditional BMC), we were completely at odds with OCP (where ASPEED BMCs abound!). So in the end, even the mechanical and…

How does the Oxide 48V architecture compare to what is known about the Google/OCP architecture? Does Oxide use single stage conversion, intermediate 12V buses, or ??

I'm not someone who works on this part of the product, but we talk a little about this stuff here:

* https://oxide.computer/blog/how-oxide-cuts-data-center-power...

* https://docs.oxide.computer/guides/introduction

I feel like we had a good Oxide and Friends on this too... https://oxide-and-friends.transistor.fm/episodes/bringing-up... has some info about our power setup.

Anyway, I barely know anything about this topic, but I think this answers your immediate question: we convert AC -> DC once at the rack level, and then use a bus bar to distribute that to each sled. Each sled also has a converter to convert that 54V down to 12V for its own bus within each sled.

Re: Our $100M Series B

#344

Post by 0xbadcafebee »

neom wrote: post #83

I'm a Bryan Cantrill fan so I'm glad this is working out, I was extremely skeptical of them at the beginning(on HN too), I think because I've built DCs for many years and was stuck in a mindset that served my use case, I've come around to Oxide. My main concerns originally were 2 fold: "this seems bougie", is there actually a market for this, and, is there a good interoperability story with mix and match. From what I…

Even if it makes no sense as a technical thing, businesses will buy it. Look at all the huge companies that keep spending millions trying to DIY their own datacenter for the 3rd time. Enterprise loves to buy big iron and self-hosting crap, so I'm sure they will be successful selling this. However, I think they're going to need to branch out to more services in order to continue increasing their revenue every year (after 5+ years lets say).

Re: Our $100M Series B

#345

Post by yencabulator »

throw0101c wrote: post #286
kortilla wrote: post #231

Earlier quoted context omitted.

But the majority of heat is going to come from the CPU and this is a product to run arbitrary customer workloads. If the customers leave these things idle, then oxide is going to shine. But a busy rack is going to be dominated by CPU heat.

According to Oxide Computer, they found that going from 20mm to 80mm fans dropped their chassis power usage (efficiency is to the cube of the radius): a rack full of 1U servers had 25% of its power going to the fans, and they were able to get down to 1.2%: * https://www.youtube.com/watch?v=hTJYY_Y1H9Q From their weblog post: > Compared to a popular rackmount server vendor, Oxide is able to fill our specialized racks…

This is not any different than the "blade" form factor that was popular in the 90s. Shared power and cooling that was not constrained by the height of a 1U rack chassis, with larger fans. Hell, even Supermicro has blade-style chassis with 80mm fans. This is not novel.

It's just plain old engineering, optimized to sell whole racks not individual servers or <=8U units, sprinkled with opinions about low-level firmware etc, with a bespoke OS and management stack.

Re: Our $100M Series B

#346

Post by steveklabnik »

derangedHorse wrote: post #299

After scrolling the site for 5 minutes, I still don't understand what they do. Do they offer on-prem servers? Then how does it bring cloud scalability? Is their selling point modular computer components so that you can order things online to achieve cloud scalability? Whatever it is, I expect it to have a poor user experience if I can't find a simple 'about' page.

> Do they offer on-prem servers?

Yes.

> Then how does it bring cloud scalability?

You buy an entire (or half) rack at a time, and then can treat the entire thing as one elastic pool of resources, you do not need to manage the individual sleds within the server yourself. Your interface to the rack is similar to a VPS provider, it's "give me a new vm with these specs."

Happy to elaborate if you have more questions.

Re: Our $100M Series B

#347

Post by steveklabnik »

htrp wrote: post #34
MOARDONGZPLZ wrote: post #9

Very bullish on this team! Congrats. I’ve been pushing my company to adopt their hardware, and we have!

Same! also curious about how you think about the potential of vendor lock-in?

(Not your parent) What kind of lock in are you worried about?

Re: Our $100M Series B

#348

Post by steveklabnik »

bix6 wrote: post #219
diggan wrote: post #110

Earlier quoted context omitted.

Not finding a ton of information about it, was curious about it too. But the "smallest system" seems huge, like 16 sleds and a HN comment from 2023 ( https://news.ycombinator.com/item?id=38498840 ) says: > An oxide rack has a minimum cost of something like 600k not including all the infra you need to run a rack, maintenance, and then needing to upgrade So maybe not super useful for people who are looking to do some h…

Wow so really for F500 and such

It's much broader than that, but you're right that we make sense for larger organizations, not smaller ones.

Re: Our $100M Series B

#349

Post by cortesoft »

transpute wrote: post #272
sethops1 wrote: post #238

Earlier quoted context omitted.

I was skeptical as well, if only because just being a better product isn't enough to win the market. Everything we hear about Oxide sounds like an impressive green field implementation of a data center, but is that enough? Do the people making buying decisions at this scale care if their sysadmins have better tools?

What percentage of enterprise IT compute has not moved to a public cloud?

I know I have been involved in multiple efforts to move the same workloads into and then out of the cloud, as corporate budgeting requirements prioritized either capex or opex at different times.

Re: Our $100M Series B

#350

Post by steveklabnik »

wmf wrote: post #333
steveklabnik wrote: post #289

Earlier quoted context omitted.

What would you like to see open that isn't? The major thing I can think of is not opening up like, our CAD files for the hardware. But that's pretty far away from your characterization.

It's open source (that effectively requires the rack as a dongle) but not flexible or modular in any way. There's not even a single PCIe slot AFAIK.

I see. I wouldn't describe that as "fully closed" but I appreciate the perspective!
Post reply on HN