Top all-time · #9085
Stream HPC
Company blog · streamcomputing.eu/bloghardwareprogramming
Stream HPC is the company blog at streamcomputing.eu, writing about hardware and programming. Since 2011, 9 different Hacker News readers have linked it 9 times. It is still writing: the last post was on 2026-10-28, and you can follow it by feed.
Too few sampled comments (7) to say how often HN recommends it.?
Why HN links it
Stream HPC, an OpenCL consultancy's blog with posts on GPU performance per watt, CUDA vendor lock-in and driver issues. HN commenters cite its posts as sources in GPU and supercomputing threads.
Summarised by Claude from a sample of 7 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
Q&A
Most-linked posts?
- /blog/2012-08-27/processors-that-can-do-20-gflops-watt/3 people · 3 mentions · first linked 2013-04-16
What Hacker News says?
With regards to your thoughts about the hardware being expensive, I found the following [1] article interesting. I'm mostly referring to the vendor lock-in part with CUDA. [1]: https://streamcomputing.eu/blog/2014-08-05/7-things-nvidia-d...
Comment linking /blog/2014-08-05/7-things-nvidia-doesnt-want-know/
on “Keylogger in Hewlett-Packard Audio Driver”Submitted An introduction to AccelerEyes ArrayFire - a library on top of OpenCL and CUDA as a story · 3 points
HN discussionPer dollar, AMD performs much better than nVidia for a significant number of cases[1] and they have much larger memory. However, this is comparing an equally optimized implementation. Unfortunately nVidia simply has much more optimized content availible. https://streamcomputing.eu/blog/2016-05-06/noticeable-proces...
Comment linking /blog/2016-05-06/noticeable-processors-nvidia-amd-intel/
on “Nvidia adds telemetry to latest drivers”Submitted Difference between CUDA and OpenCL 2010 as a story · 1 points
HN discussion"I wonder if that's intentional on NVIDIA's part." I think that's a reasonable guess. NVIDIA only supports OpenCL 1.2 (and it took them about 6 years to get there from 1.0, while other vendors were at 2.x). While I don't think NVIDIA ever outright stated OpenCL was on the back burner, their support clearly waned, which devs noticed ( https://streamcomputing.eu/blog/2012-09-10/nvidias-industry-... ). As for why...wel…
Comment linking /blog/2012-09-10/nvidias-industry-leading-support-for-opencl/
on “Blender 2.79: OpenCL on par with CUDA”It is 90 GFLOPS at less that 5 Watts. That's not too shabby. Adapteva claims that the Epiphany cores have a GFLOPS/W ratio of 50. See here: http://streamcomputing.eu/blog/2012-08-27/processors-that-ca... Also, the boards for the backers feature a ZYNQ-7020 SOC by XILINX which sports a 1.3M Gate FPGA, available to the user. This ain't bad either.
Comment linking /blog/2012-08-27/processors-that-can-do-20-gflops-watt/
on “Parallella, a $99 Linux Supercomputer”HFT optimizes for low latency rather than high performance, and FPGA is currently the state of the art there to my knowledge [1]. When it comes to performance, you should look at FPGA in terms of performance per Watt. Generally speaking they outperform GPUs by an order of magnitude in FLOP/W*s [2], which in turn already have ~ 3x-5x advantage over Xeons [3]. This measure is the most important one when it comes to th…
Comment linking /blog/2012-08-27/processors-that-can-do-20-gflops-watt/
on “Intel Buys Altera for $16.7B”20+ GFLOPS/W for a single part isn't something new: http://streamcomputing.eu/blog/2012-08-27/processors-that-ca... (2012) A huge amount of overall system power is spent in data transport. Plus, double everything for cooling. That brings the total system efficiency way down from what the actual computational components spend.
Comment linking /blog/2012-08-27/processors-that-can-do-20-gflops-watt/
on “Supercomputers: Obama orders world's fastest computer”We're a company in the Netherlands specialised in OpenCL-consultancy. Check http://streamcomputing.eu
Comment linking /
on “Do you use or know someone who uses OpenCL?”