Massive, unapologetic, firepower: 2TB write in 73 seconds

A 1.2PB single mount point Scalable Informatics Unison system, running an MPI job (io-bm) that just dumps data as fast as the little Infiniband FDR network will allow. Our test case. Write 2TB (2x overall system memory) to disk, across 48 procs. No SSDs in the primary storage. This is just spinning rust, in a … Read moreMassive, unapologetic, firepower: 2TB write in 73 seconds

Doing what we are passionate about

I am lucky. I fully admit this. There are people out there whom will tell you that its pure skill that they have been in business and been successful for a long time. Others will admit luck is part of it, but will again, pat themselves on the back for their intestinal fortitude. Few will … Read moreDoing what we are passionate about

We had a record setting, knock the barn doors down year last year

… and believe it or not, I forgot to mention it. This is the first time in company history that we had a backlog going into Q1. Orders being built and tested on the last work day of the year. We grew, not the amount we had originally forecast, but we understand why (and sadly … Read moreWe had a record setting, knock the barn doors down year last year

… and the positions are now, finally open …

See the Systems Engineering position here, and the System Build Technician position here. I’ll get these up on the InsideHPC.com site and a few others soon (tomorrow). But they are open now. For the Systems Engineering position, we really need someone in NYC area with a strong financial services background … Doug made me take … Read more… and the positions are now, finally open …

Massive. Unapologetic. Firepower. 24GB/s from siFlash

Oh yes we did. Oh yes. We did. This is the fastest storage box we are aware of, in market. This is so far outside of ram, and outside of OS and RAID level cache … [root@siFlash ~]# fio srt.fio … Run status group 0 (all jobs): READ: io=786432MB, aggrb=23971MB/s, minb=23971MB/s, maxb=23971MB/s, mint=32808msec, maxt=32808msec This … Read moreMassive. Unapologetic. Firepower. 24GB/s from siFlash

Updated DeltaV4 quick benchies

Streaming reads and writes. Far beyond memory/cache/… all spinning disk. Remember, this is our “slow” storage. [root@dv4-1 ~]# df -h /data Filesystem Size Used Avail Use% Mounted on /dev/md2 55T 65G 55T 1% /data Run status group 0 (all jobs): WRITE: io=65505MB, aggrb=1467.7MB/s, minb=1467.7MB/s, maxb=1467.7MB/s, mint=44633msec, maxt=44633msec Run status group 0 (all jobs): READ: io=65412MB, … Read moreUpdated DeltaV4 quick benchies

We built that: 10 years in business

[warning: longer post]
I mentioned this on twitter (@sijoe). The day job has been in business for 10 years. We’ve not taken outside investment to date, and we’ve not sold the company yet. We’ve been profitable and growing continuously during our lifetime. The preceding 3 years have seen growth, accelerating hard.
The company was built starting with a conviction that practitioners and users of HPC systems needed better designs, better systems than were being pushed out by traditional vendors in the early 2000’s. As beowulf systems took off, we saw some of the most egregious designs … 4 years ago, I wrote a post decrying some of these issues I had seen. And the points I raised rubbed a number of people the wrong way, as you can see in the comments.
We started the company in part to make sure the high performance side was represented in these cluster systems. We haven’t always been successful in this endeavor.
In the late 90’s, I had been an advocate within SGI for building and selling clusters. I pointed out benchmarks to colleagues whom had insisted that RISC machines would never lose in performance to CISC machines. Its very hard to explain away benchmarks that show otherwise. I had argued, in a very Clayton Christenson manner, that we should should creatively destroy our own products, lest our competitors do this for us.
Which they did.
The clustering effort at SGI was canceled when SGI exited the x86 market for the first time in 2001. I left within weeks of this, as I was convinced at that time, even though it was by no means certain, that Linux clusters would come to dominate HPC (for at least a while). I thought the management (then) at SGI had made a mistake.
My failure at that time was not to wait for a layoff and take a severance package, so I could start this company. That was a mistake. Not a fatal one though.
I had limited tolerance for risk at that moment, and felt I needed a larger company around me.
16 months later, having experienced what I needed to experience at another firm … I decided I needed to start this company, if for no other reason than to prove myself right. I spoke to my wife and we talked about the issues relative to our 1 year old. This was a huge risk.
Monumental.
I was putting out a shingle at the height of a recession.
I was doing this

(a mouse “flipping the bird” to an bird … ) to the recession.
We had our first customer in 15 days.
Our second in 30 days after starting.
Now here we are 3652 days later. Going from a skim over the water at slow speed, to 45 degree attack angle and full throttle about 5 years ago, then 60 degree attack angle and afterburners lighting off, pressing us back hard in our seats just a few years ago … to low earth orbit and warp cores spinning up now.
Yeah, we built that.

Read moreWe built that: 10 years in business