Department of Computer Science, University of #Manchester @OfficialUoM, one of the largest in the @RussellGroup π©πΏβπ»π©βπ»π¨πΏβπ»π¨βπ»
Manchester, UK
Joined July 2012
- Tweets7K
- Following584
- Followers12.4K
- Likes1.4K
Computer Science @ The University of Manchester retweeted
TornadoVM 8.0.0 is out πͺοΈ
After 10 years: thank you, Graal. Java kernels are now compiled by TC, our own pure-Java compiler to CUDA, OpenCL and Metal.
β’ AOT ready
β’ No JVMCI
β’ Now Apache 2.0
β’ Your code doesn't change
π tornadovm.org/blogs/tornadovβ¦
#OpenSource #Java #AI #GPU
Computer Science @ The University of Manchester retweeted
Great to see @tornadovm on the top of the @ycombinator today! The article has all of the latest work on accessing @nvidia libraries from Java!
Computer Science @ The University of Manchester retweeted
Honoured to receive the Best Paper Award at #INLG2026! ππΎ
Our survey, led by my student @tylerl404, reveals a striking gap between AI practitioners and the public: practitioners are more optimistic, yet nearly half still see AIβs societal impact as negative
@csmcr @inlgmeeting
Happy to announce that my paper with @chenghua_lin "Do as I Say, Not as You Do: Assessing Public Versus Practitioner Perspectives on the Development of Artificial Intelligence and Language Technologies" won Best Paper at @inlgmeeting! @sltcdt #NLP Paper: bit.ly/4iaeier
Computer Science @ The University of Manchester retweeted
Had so much fun at @JDD_Krakow 2026 in KrakΓ³w yesterday.
I presented βJava Goes CUDA-Nativeβ: new @tornadovm features that let @jitLLM run LLM inference at GPU-native speed, from Java.
tornadovm.org
jitllm.io
#opensource #Java #TornadoVM #GPU #AI
Computer Science @ The University of Manchester retweeted
96.65 tokens/sec for #Llama 3.2 1B β running through a pure-#Java GPU inference engine. @mikepapadim explores GPULlama3.java & #TornadoVM and how they bring GPU-accelerated #LLM inference to the #JVM.
See the code: javapro.io/2026/10/13/buildiβ¦
#AI @langchain4j @QuarkusIO
Computer Science @ The University of Manchester retweeted
LLM Inference on GPU in Java?
@tornadovm and @jitllm at @Devoxx Belgium!
* Java writing Java that runs on the GPU. Java for the GPU, on the GPU!.
* A @QuarkusIO @langchain4j application reviewing it's future port to Quarkus 4 locally, all in Java on GPU!
youtu.be/HO5CpETzywk?is=_ws2β¦
Computer Science @ The University of Manchester retweeted
TornadoVM has a new family member! #jitLLM provides blazing fast AI inference at native speedβ‘οΈπ
Can Java compete with llama.cpp/@ollama? YES! With @jitLLM!
jitLLM uses @nvidia CUDA libraries (cuBLAS, cuDNN) and reaches 0.90Γ to 1.12Γ of llama.cpp's throughput!
jitLLM is empowered by @tornadovm and will be soon in @langchain4j and @QuarkusIO.
github.com/beehive-lab/jitllβ¦
Computer Science @ The University of Manchester retweeted
Optimized GPU libraries (e.g. cuBLAS, cuDNN, etc.) by @nvidia, now one jar away for your Java - without caring about JNI.
We also show demos with @ApacheFlink and @jitllm.
Watch our talk in @Devoxx π
youtu.be/6SPnBmQrYEk?si=3DI_β¦
#opensource #AI #TornadoVM #Java #GPU
Computer Science @ The University of Manchester retweeted
Yesterday my colleagues @MXekalaki and @mikepapadim showed live demos about using @nvidia cuFFT and cuBLAS with @tornadovm for signal filtering and @jitllm!
NVIDIA performance is a jar away from Java.
#Java #CUDA #TornadoVM
Optimized GPU libraries (e.g. cuBLAS, cuDNN, etc.) by @nvidia, now one jar away for your Java - without caring about JNI.
We also show demos with @ApacheFlink and @jitllm.
Watch our talk in @Devoxx π
youtu.be/6SPnBmQrYEk?si=3DI_β¦
#opensource #AI #TornadoVM #Java #GPU
Computer Science @ The University of Manchester retweeted
We are @Devoxx, getting ready for our first hands-on #GPU lab for #Java!
We will be in BOF2 room at 16:50, with live demos!
@tornadovm @jitllm
Computer Science @ The University of Manchester retweeted
Can Java compete with llama.cpp/@ollama? YES! With @jitLLM!
jitLLM uses @nvidia CUDA libraries (cuBLAS, cuDNN) and reaches 0.90Γ to 1.12Γ of llama.cpp's throughput!
jitLLM is empowered by @tornadovm and will be soon in @langchain4j and @QuarkusIO.
github.com/beehive-lab/jitllβ¦
Computer Science @ The University of Manchester retweeted
Next week we will be in @Devoxx and speak about @tornadovm, @jitllm and how we accelerate Java on GPUs, using @nvidia native libraries!
Come and talk to us, to learn more!
#opensource #Java #AI #TornadoVM #jitLLM #GPU
Computer Science @ The University of Manchester retweeted
TornadoVM 7.0.0 is out!
New TileContext API: Write #GPU kernels from #Java in tiles, not threads, via @nvidia cuTile model. The compiler picks threads, shared memory and Tensor Cores.
Mixes with Parallel API, KernelContext & cuBLAS in one TaskGraph.
π github.com/beehive-lab/Tornaβ¦
Computer Science @ The University of Manchester retweeted
Java 27 is out and @tornadovm is already compatible!
xitter.cf/tornadovm/status/21002β¦
Java 27 is now available! #Java27 #JDK27 #OpenJDK
Download now: social.ora.cl/6015BGuseS
Release notes: social.ora.cl/6017BGuset
API Javadoc: social.ora.cl/6018BGuseQ
Features: social.ora.cl/6010BGusea
Blog: social.ora.cl/6013BGuseL
Computer Science @ The University of Manchester retweeted
A new TileContext API in TornadoVM's CUDA backend! Write tile kernels in plain Java and let @nvidia's cuTile compiler pick the threads, layouts and tensor-core instructions. No PTX, no warps.
PR is open - feedback welcome π
github.com/beehive-lab/Tornaβ¦
#Java #TornadoVM #GPU
@tornadovm now has a cuTile backend from Java.
Write a Java kernel with the new TileContext API β compiled to CUDA Tile C++ β tileIR picks the layout and HMMA
Then chain it with JIT'd CUDA kernels and a @nvidia libraries in One CUDA Graph.
github.com/beehive-lab/Tornaβ¦
Computer Science @ The University of Manchester retweeted
TornadoVM 6.0.0 is here - JVMCI-free and JNI-free.
β
Compatible with any JDK 21-27 distribution.
β
Cold start down 37%.
β
4.1Γ faster GEMM with Tensor-Core MMA.
β
~7,400 lines of C/C++ glue deleted.
π New blog: tornadovm.org/blogs/tornadovβ¦
All SDKs are pushed in @sdkman_ and docker images in DockerHub!
#Java #GPU #CUDA #TornadoVM #OpenSource #HPC #AI
Computer Science @ The University of Manchester retweeted
Big leap forward for Tornadovm, as it moves into a JVMCI and JNI free era by adapting FFM. All new updates will be @Devoxx in Belgium along how we enable more @nvidia features to the JVM!
TornadoVM 6.0.0 is here - JVMCI-free and JNI-free.
β
Compatible with any JDK 21-27 distribution.
β
Cold start down 37%.
β
4.1Γ faster GEMM with Tensor-Core MMA.
β
~7,400 lines of C/C++ glue deleted.
π New blog: tornadovm.org/blogs/tornadovβ¦
All SDKs are pushed in @sdkman_ and docker images in DockerHub!
#Java #GPU #CUDA #TornadoVM #OpenSource #HPC #AI
Computer Science @ The University of Manchester retweeted
Very excited that the latest @tornadovm work on exposing the @nvidia ecosystem to Java will hit the @Devoxx in Belgium this year!
Some exciting news from our team! This year we will give 3 sessions at @Devoxx Belgium!
π οΈ Tools in Action: NVIDIA ecosystem from Java via #TornadoVM's Hybrid API - CUDA/cuBLAS/cuFFT, no JNI
π€ Talk: Building LLM inference engines in Java with TornadoVM + CUDA-native acceleration
π§βπ» Hands-on Lab: From install to insight β a GPU lab for Java developers
Java meets GPUs. Stay tuned π
#Java #CUDA #LLM #AI #GPU #JVM #opensource
