Viewing profile — benchess
benchess
HN member- Joined
- Thu, Apr 25, 2013, 5:37 PM UTC
- HN karma
- 62
- Public activity
- 25 items
- HN profile
- View on Hacker News ↗
About benchess
No profile information was provided.
Recent public activity
-
comment
Comment #45699388
> - It's not possible to set screen time restrictions for Safari. I thought this too. I discovered it actually is possible though, just doesn't appear in the list. Go "Screen Time"…
-
comment
Comment #45596079
Heya, wanted to show off a project of mine to identify the scaling limits of a large Kubernetes deployment
- story
-
comment
Comment #44413189
The irony of posting this on GitHub which remains shamefully without IPv6
-
comment
Comment #42189194
It’s mentioned
-
comment
Comment #41287770
My kids like reading this over and over: Now & Ben: The Modern Inventions of Benjamin Franklin: https://a.co/d/jhm2SvM
-
comment
Comment #40530606
Use a dns proxy and https://github.com/froggeric/DNS-blocklists/blob/main/NoAppl... You may also need to disable Private Relay
-
comment
Comment #39430971
They’re for sale on Mouser for $20625 each https://www.mouser.com/ProductDetail/BittWare/RS-GQ-GC1-0109... At that price 568 chips would be $11.7M
-
comment
Comment #38740141
This isn't running on one chip. It's running on 128, or two racks worth of their kit. https://news.ycombinator.com/item?id=38739106 This doesn't mean much without comparing $ or wa…
-
comment
Comment #38738306
If I understand correctly, Groq chips have 220MB SRAM and the next best level is DDR4? How many chips are needed to run Llama2-70B at those speeds?
-
comment
Comment #27493810
The key constraint here is that the author has no access to persistent disks and can only use object storage for persistence. Otherwise Thanos would be extreme overkill for this nu…
- comment
-
comment
Comment #25909093
Hi, co-author here. Yes we are excited about the potential of this!
-
comment
Comment #25908996
Hi, co-author here! We use a pretty standard tech stack of PyTorch + NCCL + MPI. We've used both OpenMPI and MPICH to varying degrees. Kubeflow is interesting, but it solves a slig…
-
comment
Comment #25908350
Hi! Co-author here. We do keep the nodes running 24/7, so Kubernetes still provides the scheduling to decide which nodes are free or not at any given time. Generally starting a con…
-
comment
Comment #24380092
It's all pretty well covered in the Methodology section of the about page: https://takecongress.org/about.html
-
comment
Comment #20972222
Utilize SNI and serve up a fake cert when someone scans you without a matching hostname. Censys is scraping you by IP, so it'll just see the fake cert. You can do this in nginx by …
-
comment
Comment #16554173
Not all bandwidth is created equal. Different providers can have widely different levels of performance. Higher-priced bandwidth _can_ also be better bandwidth. Cloud providers hav…
-
comment
Comment #10595042
What an amazing story. Great to hear of people like Katherine!
-
comment
Comment #10025733
So when you hit the limit you want AWS to stop your spending... how? You're spending $/hr on compute, so terminate your instances. Plus your EBS volumes, so delete your drives. Plu…
- comment
- comment
-
comment
Comment #8785352
Also Stratos https://stratoscard.com/
-
comment
Comment #8529258
"No planned downtime: Local SSD data will not be lost when Google does datacenter maintenance." Only unplanned downtime then :) Live migration maybe adds another '9' to reliability…
-
comment
Comment #7644219
It's both a product and a platform. plus.google.com and its "What's New?"-style newsfeed is a tangible product. The news that Google+ integration is no longer required for other Go…