jonny (good kind)
boosted
My #introduction needs a refresh.
I'm a #PhD student at the University of #Vermont, studying the #Evolution of #Evolvability. I'm into #AI, #ALife, #Biology, and #Philosophy, because I want to understand #life, #adaptation, and #intelligence using my native language of #ComputerScience. I share my musings and #research on my #blog. I love #science generally, and am full of bitchy #AcademicChatter.
I was a #SoftwareEngineer in #SiliconValley for many years, but left in 2021. I'm glad I did, and now I feel a bit betrayed by the #TechIndustry. I've been going back to my #FOSS roots, and gradually #DeGoogle ing my life. I still love to talk about #code #craft, #UX, and healthy #engineering #culture. Recently I've been enjoying #gpu #programming, mostly in #taichilang.
I have a wife and a #cat. I love #nature, #photography, #cooking, and #yoga.
All kinds of people are valid and worthy, but #trans people, folks on the #autism spectrum, and #bipoc get a shout out right now because they need our support.
My #introduction needs a refresh.
I'm a #PhD student at the University of #Vermont, studying the #Evolution of #Evolvability. I'm into #AI, #ALife, #Biology, and #Philosophy, because I want to understand #life, #adaptation, and #intelligence using my native language of #ComputerScience. I share my musings and #research on my #blog. I love #science generally, and am full of bitchy #AcademicChatter.
I was a #SoftwareEngineer in #SiliconValley for many years, but left in 2021. I'm glad I did, and now I feel a bit betrayed by the #TechIndustry. I've been going back to my #FOSS roots, and gradually #DeGoogle ing my life. I still love to talk about #code #craft, #UX, and healthy #engineering #culture. Recently I've been enjoying #gpu #programming, mostly in #taichilang.
I have a wife and a #cat. I love #nature, #photography, #cooking, and #yoga.
All kinds of people are valid and worthy, but #trans people, folks on the #autism spectrum, and #bipoc get a shout out right now because they need our support.
I am not worried about the alleged #aibubble
That occurs when there is over-investment. Right now the demand outstrips the supply.
#AI is not going away.
There will however be a problem when #bigtech moves over to #NPU instead of #GPU usage.
That will wipe out smaller providers as the cost per token will fall dramatically and that will shift power upwards.
Power companies also want to meet profit projections, so to make up for the shortfall in usage by increasing private #electricity cost.
alcinnz
boosted
📏 Text rendering and effects using GPU-computed distances
https://blog.pkh.me/p/47-text-rendering-and-effects-using-gpu-computed-distances.html
#graphics #gpu #programming #software #text #fonts #rendering #softwaredevelopment
📏 Text rendering and effects using GPU-computed distances
https://blog.pkh.me/p/47-text-rendering-and-effects-using-gpu-computed-distances.html
#graphics #gpu #programming #software #text #fonts #rendering #softwaredevelopment
Ulrike Hahn
boosted
🧠 New paper by Deistler et al: #JAXLEY: differentiable #simulation for large-scale training of detailed #biophysical #models of #NeuralDynamics.
They present a #differentiable #GPU accelerated #simulator that trains #morphologically detailed biophysical #neuron models with #GradientDescent. JAXLEY fits intracellular #voltage and #calcium data, scales to 1000s of compartments, trains biophys. #RNNs on #WorkingMemory tasks & even solves #MNIST.
Jaxley: differentiable simulation enables large-scale training of detailed biophysical models of neural dynamics
🧠 New paper by Deistler et al: #JAXLEY: differentiable #simulation for large-scale training of detailed #biophysical #models of #NeuralDynamics.
They present a #differentiable #GPU accelerated #simulator that trains #morphologically detailed biophysical #neuron models with #GradientDescent. JAXLEY fits intracellular #voltage and #calcium data, scales to 1000s of compartments, trains biophys. #RNNs on #WorkingMemory tasks & even solves #MNIST.
Jaxley: differentiable simulation enables large-scale training of detailed biophysical models of neural dynamics
Simulating a Planet on the GPU: Part 1 (2022)
https://www.patrickcelentano.com/blog/planet-sim-part-1
#HackerNews #Simulating #a #Planet #on #the #GPU #Part #1 #2022 #GPUComputing #PlanetSimulation #GraphicsProgramming
a Rust ray tracer that runs on any GPU – even in the browser
David Vasandani
boosted
Any of you sharing GPU's without SRIOV?
I have been "abusing" my host OS to run gpu demanding work.
Notably, steam with gamescope so servers aren't single use.
They can server as multi-media centers, but also as couch gaming stations, nas, etc.
Basically trying to push the envelope of HCI (Hyper Converged Infra).
All of this because GPU vendor are greedy and don't want to have SRIOV on consumer grade hardware.
Any of you sharing GPU's without SRIOV?
I have been "abusing" my host OS to run gpu demanding work.
Notably, steam with gamescope so servers aren't single use.
They can server as multi-media centers, but also as couch gaming stations, nas, etc.
Basically trying to push the envelope of HCI (Hyper Converged Infra).
All of this because GPU vendor are greedy and don't want to have SRIOV on consumer grade hardware.
Stefano Marinelli
boosted
🎁 Dev Box DPU Upgrade 🎁
More ZFS PoC for vGPU + DOCA blah blah NVMe-oF offloads
Left–to–Right (lotta slots)
- Broadcom CNA 2x10G
- Optane P3608 2x 2TB NVMe AIC
- Nvidia T10 16GB GPU (w/ 90mm blower)
- blank
- Nvidia T10 16GB GPU (w/ 90mm blower)
- Nvidia Bluefield2 DPU 16GB RAM, 2x 25G
- Optane 4x 512GB NVDIMM
- Samsung 4x 32GB ECC Volatile
- Intel Xeon 8370C (Azure) 32C/64T 270W
Chassis cooled by 2x 180mm, 1x 140mm, 1x 120mm. Idles very comfortably below 40dB, wind-tunnel at full rpm is more like whoooooshing than the usual 28K-RPM 40mm axial fan screamers people expect from enterprise hardware. Choose wisely.
#gpu #dpu #ai #dev #llm #nvidia #intel #optane #zfs #linux #freebsd
🎁 Dev Box DPU Upgrade 🎁
More ZFS PoC for vGPU + DOCA blah blah NVMe-oF offloads
Left–to–Right (lotta slots)
- Broadcom CNA 2x10G
- Optane P3608 2x 2TB NVMe AIC
- Nvidia T10 16GB GPU (w/ 90mm blower)
- blank
- Nvidia T10 16GB GPU (w/ 90mm blower)
- Nvidia Bluefield2 DPU 16GB RAM, 2x 25G
- Optane 4x 512GB NVDIMM
- Samsung 4x 32GB ECC Volatile
- Intel Xeon 8370C (Azure) 32C/64T 270W
Chassis cooled by 2x 180mm, 1x 140mm, 1x 120mm. Idles very comfortably below 40dB, wind-tunnel at full rpm is more like whoooooshing than the usual 28K-RPM 40mm axial fan screamers people expect from enterprise hardware. Choose wisely.
#gpu #dpu #ai #dev #llm #nvidia #intel #optane #zfs #linux #freebsd
alcinnz
boosted
I think regular programming-languages need to incorporate GPU programming capabilities.
For example, I would like to be able to program a GPU using the Go programming-language.
I think regular programming-languages need to incorporate GPU programming capabilities.
For example, I would like to be able to program a GPU using the Go programming-language.
Stefano Marinelli
boosted
Stefano Marinelli
boosted
Update for #nvidia #GPU #driver #ports on #FreeBSD to 580.95.05 landed as commit 7b1c88f42c7f4129bc96f405549f9e6b07bd6e3d.
https://cgit.freebsd.org/ports/commit/?id=7b1c88f42c7f4129bc96f405549f9e6b07bd6e3d
We were pending it as kmod pkg repo still didn't have nvidia things and bapt was looking into it, but as of this information
https://nvidia.custhelp.com/app/answers/detail/a_id/5703
the commit was done without waiting for the reply from bapt.
#FreeBSD is not entitled there, but as #FreeBSD native #drivers are known to shares most source codes with corresponding Linux versions, FreeBSD can be affected, too. So users of non-legacy versions of driver ports are strongly encouraged to upgrade.
With this reason, the commit is already merged into 2025Q4, as security fix is one of the branket approval by portmgr and secteam.
https://cgit.freebsd.org/ports/commit/?h=2025Q4&id=067be176997589baacd63c5e5db1981d3949cdb1
Note that oldest branch listed is 535.