
graphics \ rtl \ hw sw co-design | GPU Arch @apexcompute
Joined January 2013
- Tweets51
- Following2.7K
- Followers306
- Likes8.3K
Selim retweeted
Apex Neural Rendering: Can Independent Devs Beat DLSS? digitalfoundry.net/features/…
ANR Frame Generation V.B16 is live with new model weights from your beta feedback:
- Fixed ghosting & lag in games with broken motion vectors
- Bloom & vignette compensation
- Signed DLLs
Next up: HW Flip Metering for RTX 5000 & RX 9000.
Thanks for 1k+ stars!
github.com/apex-compute/anr
Selim retweeted
Apex Neural Rendering Beats NVIDIA DLSS in Both Image Quality and Performance tpu.me/bttx
Selim retweeted
Introducing Apex Neural Rendering (ANR), our cross-platform Super Resolution, Ray Reconstruction, and Frame Generation technology, delivering up to 2.4× faster performance and higher quality rendering than NVIDIA DLSS.
ANR rebuilds neural rendering around a new architecture that outperforms current state of the art models.
Works on all GPU vendors.
- NVIDIA: GeForce RTX 20, 30, 40, and 50 series, plus GTX 16 series
- AMD: Radeon RX 5000, 6000, 7000, and 9000 series, plus Steam Deck
- Intel: Arc A and B series
Frame Generation available now:
apexcompute.com/neural-rende…
Our Super Resolution and Ray Reconstruction models are currently in training. Coming soon!
Early results: place & route running entirely on GPU (our Xilinx Vivado implementation flow replacement)
These 3 are designs we use to benchmark the tool
Example part is U50
Selim retweeted
We’re building the Apex GPU alongside our AI accelerator.
Our open-source Mesa driver just achieved Vulkan 1.3 conformance on our Alveo U50 FPGA prototype. This is the first FPGA-prototype GPU capable of running Vulkan while rendering the entire Ubuntu desktop.
We built the entire stack, from the GPU microarchitecture and RTL all the way up to the Mesa driver and Vulkan.
The entire Linux desktop is rendered and driven by the Apex GPU at up to 4K 120 Hz on FPGA.
@VulkanAPI
We’re building the Apex GPU, along with our AI accelerator.
Our open-source Mesa driver just achieved Vulkan 1.3 conformance on our Alveo U50 FPGA prototype.
This is the first FPGA-prototype GPU capable of running Vulkan on a commercially available FPGA platform. We implement the whole stack, from the GPU microarchitecture and RTL to the Mesa driver and Vulkan.
We also think there is a very interesting opportunity to combine the Apex GPU with our AI accelerator and program both through Vulkan, enabling graphics and AI acceleration through a unified software interface.
A lot more to share soon.
Selim retweeted
We actually have two products: our Unified Engine AI accelerator and a full GPU on FPGA. Our GPU has passed all Vulkan 1.3 conformance tests and we are going to upload the test results for @thekhronosgroup to review.
Apex Compute @apexcompute Developing Open-Source Mesa @VulkanAPI Driver For Their Hardware
Now that's super interesting unlike many of the other AI accelerator start-ups out there....
phoronix.com/news/Apex-Compu…
Apex Compute @apexcompute Developing Open-Source Mesa @VulkanAPI Driver For Their Hardware
Now that's super interesting unlike many of the other AI accelerator start-ups out there....
phoronix.com/news/Apex-Compu…
Selim retweeted
Connected 11 @AMD Kintex-7 FPGA cards running @apexcompute Unified Engine accelerator IP to test some big models on these cheap cards. 44 GB DDR3 with 22 Unified Engine cores totaling around 557 BF16 GFLOPs.
Selim retweeted
Quick recap from @hotchipsorg . We showed our FPGA prototype of @apexcompute Unified Engine running Qwen3, Llama 3.2, Gemma4, their VLM/LLM variants and pi 0.5 end to end. Some of those models are already approaching the performance of shipping edge AI ASICs. We also demonstrated multicore execution. As a bonus, we showed our GPU prototype supports Vulkan 1.4. Check our vkcube rendering.
Great technical discussions. Huge thanks to the Apex Compute team for all the work behind these demos. Looking forward to next year's conference with our own chip.
Unified Engine FPGA prototype specs on Alveo U50
- 8 Unified Engine cores at 366 MHz
- 1 MiB local SRAM per engine
- 1024-bit SRAM datapath
- BF16, FP8/INT8, FP4/INT4, TQ4
GPU prototype specs on Alveo U50
- 16 shader cores
- 200 MHz shader clock
- 1 DMA engine, 1 ROP, 1 texture unit
- 256 KiB L2 + 32 KiB L1
- Vulkan 1.4 / Mesa
- DirectX 12.2
- 1080p at 60 Hz
Selim retweeted
Replying to @apexcompute
@apexcompute will be at @hotchipsorg demonstrating our 8-core Apex Unified Engine V2 and Apex GPU V1 FPGA prototypes, running LLMs, VLMs, VLA models, and DirectX-supported games on the GPU prototype. Find us at Table B3!
Selim retweeted
Apex Compute side project: DirectX driver up and running on our GPU prototype. It supports Vulkan 1.4 and DirectX 12.1. Fully developed in-house
- 1 shader core (16-lane SIMD)
- 2GB VRAM (fpga board limited)
- GPU virtual addressing
- DMA and raw PCIe DMA engines
- Graphics front end
- Triangle, line, and point rasterization
- Conservative rasterization and VRS
- Texture sampling, filtering, gather, and BC decode
- Depth and stencil processing
- 8-MRT ROP with blending and MSAA resolve
- ROV and interlock support
- Ray intersection assist
- Independent display and scanout pipeline
- HDMI output with HPD and DDC
- Shader arithmetic and conversion units


























