FWIW, even the “original” #NVIDIA #Thrust has been a continuous source of issues for us, so much so that I have a GitHub repository just to collect test cases for my bug reports to upstream.
The most notable issues?
Certain versions of Thrust with certain versions of the NVIDIA drivers led to GTX TITAN X (Maxwll) GPUs stalling or deadlocking after thousands of iterations.
https://github.com/NVIDIA/thrust/issues/742
You could imagine how fun this was for our PhD student whose workstation had that hardware.
one is a build failure against my GPU (already reported, with a fix ready and pending release), and the other is … slow performance in one of the #Thrust API calls that we use!
Turns out, `sort_by_key`, at least in the way we use it, is somewhere between 25% and 30% slower on my #AMD iGPU when using the latest #rocThrust (from the 5.6.0 software stack) than it is on the *CPU* when using the latest #Thrust with the OpenMP backend!
So, one of the reasons why we could implement the #HIP backend easily in #GPUSPH is that #AMD provides #ROCm drop-in replacement for much of the #NVIDIA #CUDA libraries, including #rocThrust, which (as I mentioned in the other thread) is a fork of #Thrust with a #HIP/#ROCm backend.
This is good as it reduces porting effort, *but* it also means you have to trust the quality of the provided implementation.
#thrust #rocthrust #CUDA #nvidia #ROCm #amd #GPUSPH #hip
There's another important different I haven't mentioned, between the CA and GPUSPH. All of the GPU code in the CA is “custom”, compute kernels written by yours truly. In GPUSPH, instead, there are a few instances where we rely on an external library: #Thrust.
I've already complained a bit about how this affects us <https://fediscience.org/@giuseppebilotta/110283708975056091> especially in terms of backend support, but things are *even worse*, and I'll take the opportunity here to complain a bit!
Corporate #FLOSS at its worst: #NVIDIA controls the #Thrust library and its #CUDA, #OpenMP and #TBB backend. #AMD provides rocThrust, that is just Thrust with the CUDA part stripped an a new backend for #ROCm / #HIP. Nobody* is working on a backend for #SYCL
#Intel provides its own #oneAPI alternative as #oneDPL, which is NOT a drop-in replacement.
This is why we can't have nice things.
*there's a dead project here
https://github.com/wdmapp/syclthrust
#onedpl #oneAPI #intel #sycl #hip #ROCm #amd #tbb #OpenMP #CUDA #thrust #nvidia #floss
A #photograph of a #UnitedLaunchAlliance #ULA #DeltaIVHeavy #rocket at #CapeCanaveral’s LC-37B.
The #Delta IV #Heavy is a variant of the Delta IV with 3 #CommonBoosterCores #CBC’s mounted side-by-side with the #DeltaCryogenicSpperStage #DCSS mounted on the center CBC. The center CBC throttles its #Rocketdyne RS-68 #engine down to 55% rated #thrust while the “outboard” CBC’s remain at 100% until #burnout.
The center CBC then throttles up until it burns its fuel, then releases the DCSS and #spacecraft.
https://heronfox.pixels.com/featured/t-1-heron-and-fox.html
#photograph #unitedlaunchalliance #ula #deltaivheavy #rocket #capecanaveral #delta #heavy #commonboostercores #cbc #deltacryogenicspperstage #dcss #rocketdyne #engine #thrust #burnout #spacecraft
The most important is definitely our dependency on #Thrust in a couple of places, but the extensive use of constant and shared memory is also a problem, since the syntax for their use in #SYCL is very different from the one in #CUDA, which would pose a problem for our plans to keep the device source code unified across APIs.
A #photograph of a United Launch Alliance #ULA #AltasV #rocket #liftingoff with a #GPS BlockII-F from #SLC40 in #CapeCanaveral Air Force Station #CCAFS. This version of the Atlas V is carrying its payload in a 4 meter fairing, no solid rocket boosters, and a single RL-10 on its #Centaur upper stage.
The first stage’s Russian-made RD-180 #engine is beginning to develop its full #thrust, and slowly lifting the 401 off the pad.
https://heronfox.pixels.com/featured/atlas-v-with-gps-iif-liftoff-heron-and-fox.html
(Edited to include link)
#photograph #ula #altasv #rocket #liftingoff #gps #slc40 #capecanaveral #ccafs #centaur #engine #thrust
Random Old Comic: Enemy Within https://www.toyboxcomix.com/2020/11/13/enemy-within/ #xp #buzzsaw #frenzy #laserbeak #ramjet #ratbat #ravage #rumble #soundwave #thrust #transformers
#xp #buzzsaw #frenzy #laserbeak #ramjet #ratbat #ravage #rumble #soundwave #thrust #transformers
#SpaceX has never ignited more than 42% of #SuperHeavy’s 33 Raptor #engines at once. When it ignites 🔥 all 33 engines, its maximum #thrust could leap to 7590 tons, beating the next #most #powerful #rocket in history – the Soviet N1 – by nearly 60% 💪 https://www.teslarati.com/spacex-sets-stage-starship-booster-33-engine-static-fire-2023
#Starship test probably 📆 January 30th, 31st, or February 1st
#SpaceX #SuperHeavy #engines #thrust #most #powerful #rocket #starship
#Gimbal testing new #electric #Raptor #TVC (see 0:40) https://youtu.be/OOol0ZF0a-U
Gimbaled #thrust https://en.wikipedia.org/wiki/Thrust_vectoring#Gimbaled_thrust
#RocketEngine #RocketScience #SpaceX #ThrustVectorControl #RaptorEngine
#gimbal #electric #Raptor #tvc #thrust #rocketengine #rocketscience #SpaceX #thrustvectorcontrol #raptorengine
RT one of Walter Deemer
@WalterDeemer
The stock market generated Breakaway Momentum today for the 25th time since 1945: https://walterdeemer.com/bam.htm
This is a genuine breadth thrust. It means (IMHO) we're in a bull market.
How long it lasts, and how far it carries, is something we will know only in the fullness of time.
#Deemer #Breakaway Momentum #SPX #thrust
#deemer #breakaway #spx #thrust
Holy crap how did I miss that Lidia Yuknavitch had a new book this past summer! Thrust.
Book of Joan was one of the most fantastic books I've ever read. Has anyone read this latest of hers??
#books #thrust #bookofjoan #yuknavitch
Holy crap how did I miss that Lidia Yuknavitch had a new book this past summer! Thrust.
Book of Joan was one of the most fantastic books I've ever read. Had anyone read this latest of hers??
#books #thrust #bookofjoan #yuknavitch
@Time 📆 December 2, 2022 #SpaceX successfully test-fired 11 of the 33 engines on its #SuperHeavy #rocket that will produce more than 7.25 million kg of #thrust, nearly double 💪 that of the #SLS. The #SuperHeavy stands 69 m tall and serves as the first stage carrying the 50 m #Starship #spacecraft 🚀. The entire stack stands nearly 40 stories tall—again easily beating out the #SLS’s 32 stories. https://time.com/6238248/spacex-starship-nasa-sls-rocket-launch
#SpaceX #SuperHeavy #rocket #thrust #sls #starship #spacecraft
I designed the #engine controllers and worked on the #thrust #vector controller too along with the overall #avionics architecture of #terran1 . Very proud to see this vehicle March towards flight #3dpriting #space #rockets #rocketscience #engineering #electricalengineering #electric #circuitdesign
#engine #thrust #vector #avionics #terran1 #3dpriting #space #rockets #rocketscience #engineering #electricalengineering #electric #circuitdesign
https://www.instagram.com/reel/Cl7HtAousbH/?igshid=MDM4ZDc5MmU=
Not sure if the link will work well here. I designed the #engine controllers and worked on the #thrust #vector controller too along with the overall #avionics architecture of #terran1 . Very proud to see this vehicle March towards flight #3dpriting
#engine #thrust #vector #avionics #terran1 #3dpriting
...the extreme #thrust of the #Space Launch System #rocket caused some #damage to the mobile launch tower. The violent shaking from the #launch also #broke the tower's access #elevator and blew its #doors off https://arstechnica.com/science/2022/11/nasas-new-rocket-blows-the-doors-off-its-mobile-launch-tower
#thrust #space #rocket #damage #launch #broke #elevator #doors