People buying companies just to get their hands on some more Nvidia GPUs.
Ignorant question, but where is ATI in the AI/ML space? I only ever hear about NVidia and it's Cuda.
ATI was acquired by AMD in 2006. :)
DigitalOcean acquires Paperspace (YC W15) for $111M in cash
61–70 of 98 posts
Re: DigitalOcean acquires Paperspace (YC W15) for $111M in cash
#62People buying companies just to get their hands on some more Nvidia GPUs.
That’s a pretty hysterical outlook. What’s Nvidia supply chain like with their AI GPUs? Is it constrained or is this a joke :p
It is so constrained that companies are literally buying smaller GPU cloud companies.
So, if by “hysterical,” you mean it is pretty funny and a reflection of the hype cycle, then yes. If you mean that I am being overwrought, I can assure you I am not. Even the hyperscalers do not have enough GPUs
Re: DigitalOcean acquires Paperspace (YC W15) for $111M in cash
#63Earlier quoted context omitted.
What is $35M here? Also, YC2015 would make it about 8 years run in 2023. I think exit is timely because there's the dire possibility of fading AI/LLM hype that's where GPU demand would fall off the cliff not only on the server side but also that many devices might have better inference hardware.
35m is what Paperspace raised according to crunchbase. 12 of which was raised in 2021, before the AI hype. So unless their valuation was absurd I think all investors did OK here.
But all things considered their exit is solidly average, not the insane valuations top tier unicorns usually get.
Re: DigitalOcean acquires Paperspace (YC W15) for $111M in cash
#64Earlier quoted context omitted.
The I pretend I know what I'm doing strategy. Most acquisitions fail. It's a difficult path unless it fits a very specific need. Yet a common management / product strategy is to buy random companies as though they are ingredients of a dish and think it'll just work. Digital Ocean to me has lost their core focus and appeal. They aren't upgrading their hardware, software and other crucial pieces of their infrastructure…
I love DO as a user and not an investor, but their main value to me is that they have public-facing IPs. I could easily host all my sites on a single AMD machine under my desk, if only I had a good way of routing from the internet to my desk. I would be closer to nine fives than five nines, but that's good enough even for many commercial sites, including mine.
Just contact your ISP to get a public IP. You might not get a static IP, but you can configure a dhcp client to request your old IP when reconnecting after a power outage. Set DNS TTL reasonably low like 12 hours so that you can recover if you do get another IP.
Another way is to signup for a cheap VPS and use a tunnel to your home server. There's lots of really cheap VPS's with 128MB of memory that is plenty if you just want a tunnel.
Re: DigitalOcean acquires Paperspace (YC W15) for $111M in cash
#65Earlier quoted context omitted.
Starting equity was a little less than 10%. Valuations went up as several rounds were raised up to $30+m and share value went up until at one point it was north of $20m paper dollars. After market re-valuation an additional $60+m was raised and value went down a little but was still substantial. Then the IPO market all but disappeared. Final sale price was around $150m and common shareholders including myself initial…
That VC deserves to be named.
Name any of them, they all work this way, more or less. This is standard practice
Re: DigitalOcean acquires Paperspace (YC W15) for $111M in cash
#66Earlier quoted context omitted.
The I pretend I know what I'm doing strategy. Most acquisitions fail. It's a difficult path unless it fits a very specific need. Yet a common management / product strategy is to buy random companies as though they are ingredients of a dish and think it'll just work. Digital Ocean to me has lost their core focus and appeal. They aren't upgrading their hardware, software and other crucial pieces of their infrastructure…
I love DO as a user and not an investor, but their main value to me is that they have public-facing IPs. I could easily host all my sites on a single AMD machine under my desk, if only I had a good way of routing from the internet to my desk. I would be closer to nine fives than five nines, but that's good enough even for many commercial sites, including mine.
One option you could try is using a CloudFlare tunnel.
Re: DigitalOcean acquires Paperspace (YC W15) for $111M in cash
#67Earlier quoted context omitted.
Starting equity was a little less than 10%. Valuations went up as several rounds were raised up to $30+m and share value went up until at one point it was north of $20m paper dollars. After market re-valuation an additional $60+m was raised and value went down a little but was still substantial. Then the IPO market all but disappeared. Final sale price was around $150m and common shareholders including myself initial…
Common getting zero makes sense if pref holders had liquidation multiples ("we get 3x our investment back first"), but not if they converted to common, unless their conversion to common had some weird mechanic attached to it that massively diluted the remaining common down to effectively zero % (which sounds sue-able). I'm really curious about the mechanics of how your common holding netted out to zero if prefs also…
Yes so what happened is they converted, then had majority common, then voted to recapitalize and sell, after which recapitalization they made out with a profit but common got nothing. The problem with this legally was they didn't have the right to convert without a common vote beforehand, essentially they did it in reverse order. Their leverage was the company was not cash flow positive (in no small part because of the massive management fees the VC loaded on as part of funding), and so needed funding to continue. They also offered the existing board members a $1m bonus as part of post-sale consulting (essentially a legal bribe) as part of agreeing to the sale. It's all not technically legal, but you know the golden rule, he who has the gold makes the rules.
Because of the settlement terms I'm not able to name the VC, but I can tell you this kind of behavior is by no means unique to this VC, in fact it's rather de rigueur for the VC and PE worlds. It's even defensible in a way, they owe fiduciary duty to their limited partners, NOT to the common shareholders of the company they invest in.
I currently advise startups seeking financing to either a) only go the VC route if you think you're in megagrowth into the next dropbox et. all or b) you have enough self/family wealth to take VC investment on favorable terms ala stripe.
Re: DigitalOcean acquires Paperspace (YC W15) for $111M in cash
#68Earlier quoted context omitted.
I love DO as a user and not an investor, but their main value to me is that they have public-facing IPs. I could easily host all my sites on a single AMD machine under my desk, if only I had a good way of routing from the internet to my desk. I would be closer to nine fives than five nines, but that's good enough even for many commercial sites, including mine.
Just contact your ISP to get a public IP. You might not get a static IP, but you can configure a dhcp client to request your old IP when reconnecting after a power outage. Set DNS TTL reasonably low like 12 hours so that you can recover if you do get another IP. Another way is to signup for a cheap VPS and use a tunnel to your home server. There's lots of really cheap VPS's with 128MB of memory that is plenty if you…
It's not really a silver bullet nowadays, residential plan ISPs are very wary of allowing inbound connections in my experience
Re: DigitalOcean acquires Paperspace (YC W15) for $111M in cash
#69Earlier quoted context omitted.
Starting equity was a little less than 10%. Valuations went up as several rounds were raised up to $30+m and share value went up until at one point it was north of $20m paper dollars. After market re-valuation an additional $60+m was raised and value went down a little but was still substantial. Then the IPO market all but disappeared. Final sale price was around $150m and common shareholders including myself initial…
I recall this was one of the plots from the HBO show Silicon Valley Thanks for providing insight to the process, very fascinating.
Hard to watch, not only because it rang so true, but also because I've known people personally who gave their all to the company and when it didn't work out in the end, commit suicide. It's a cautionary tale of putting your entire self concept into your startup / employer.
Re: DigitalOcean acquires Paperspace (YC W15) for $111M in cash
#70Earlier quoted context omitted.
I love DO as a user and not an investor, but their main value to me is that they have public-facing IPs. I could easily host all my sites on a single AMD machine under my desk, if only I had a good way of routing from the internet to my desk. I would be closer to nine fives than five nines, but that's good enough even for many commercial sites, including mine.
One option you could try is using a CloudFlare tunnel.
I love this option. Doubles as a way to VPN-into your home network remotely (Warp Client) without opening any ports