Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 2603.28793

Toward a Universal GPU Instruction Set Architecture: A Cross-Vendor Analysis of Hardware-Invariant Computational Primitives in Parallel Processors

Ojima Abraham·Onyinye Okoli
View original source ↗Read source document ↗Export metadata ↓

Abstract

We present the first systematic cross-vendor analysis of GPU instruction set architectures spanning all four major GPU vendors: NVIDIA (PTX ISA v1.0 through v9.2, Fermi through Blackwell), AMD (RDNA 1 to 4 and CDNA 1 to 4), Intel (Gen11, Xe-LP, Xe-HPG, Xe-HPC), and Apple (G13, reverse-engineered). Drawing on official ISA reference manuals, architecture whitepapers, patent filings, and community reverse-engineering efforts totaling over 5,000 pages of primary sources across 16 distinct microarchitectures, we identify ten hardware-invariant computational primitives that appear across all four architectures, six parameterizable dialects where vendors implement identical concepts with different parameters, and six true architectural divergences representing fundamental design disagreements. Based on this analysis, we propose an abstract execution model for a vendor-neutral GPU ISA grounded in the physical constraints of parallel computation. We validate our model with benchmark results on NVIDIA T4 and Apple M1 hardware, the two most architecturally distant platforms in our study. On five of six benchmark-platform pairs, the abstract model matches or exceeds native vendor-optimized performance. The single outlier (parallel reduction on NVIDIA, 62.5% of native) reveals that intra-wave shuffle must be a mandatory primitive, a finding that refines our proposed model.

Explore related subjects

cs.DCcs.AR

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Ojima Abraham, Onyinye Okoli. 2026-03-22. Toward a Universal GPU Instruction Set Architecture: A Cross-Vendor Analysis of Hardware-Invariant Computational Primitives in Parallel Processors. https://arxiv.org/abs/2603.28793

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Reforge: Low-Latency Distributed GNN Serving with Selective Embedding Recomputation

Graph Neural Networks (GNNs) have been widely adopted for their ability to compute expressive node representations in graph datasets. However, serving GNNs on large graphs is challenging due to the high communication, computation, and memory overheads of constructing and executing computation graphs, which represent information flow across large neighborhoods. Existing approximation techniques in training can mitigate the overheads but, in serving, still lead to high latency and/or accuracy loss. To this end, we propose Reforge, a system that enables low-latency GNN serving for large graphs with minimal accuracy loss through two key ideas. First, Reforge employs selective recomputation of precomputed embeddings, which allows for reusing precomputed computation subgraphs while selectively recomputing a small fraction to minimize accuracy loss. Second, we develop computation graph parallelism, which reduces communication overhead by parallelizing the creation and execution of computation graphs across machines. Our evaluation with large graph datasets and GNN models shows that Reforge significantly outperforms state-of-the-art techniques.

cs.DC↗
arXiv2026

Agentic AI Workload Characteristics

Agentic AI shifts LLM serving from isolated prompt-generation requests to stateful, multi-turn executions that repeatedly invoke the model, call tools, and grow context over time. This paper characterizes ReAct-style agents from both the LLM-serving and tool-execution perspectives using an end-to-end tracing infrastructure across reasoning and non-reasoning Gemma and Qwen configurations on five agentic benchmarks. Our study shows that agentic workloads are not simply long-prompt workloads: with effective context caching, most input tokens are reused across turns, making execution decode-dominated while increasing dependence on long-lived KV-cache state. We also find that tool use has a clear temporal structure, with agents shifting from read/explore behavior early in execution to execute/write behavior later. These results show that efficient agentic serving must jointly manage repeated model re-entry, persistent context state, and workload-dependent tool behavior.

cs.DC↗
arXiv2026

Byzantine Causal Reliable Broadcast (BCRB) with Constant-Size Message Metadata

Asynchronous Byzantine Reliable Broadcast (BRB) is a fundamental primitive that guarantees agreement and validity in distributed systems subject to Byzantine faults, but it lacks ordering guarantees. In this paper, we address Byzantine Causal Reliable Broadcast (BCRB), which builds on BRB to enforce causal message ordering. We present a novel BCRB protocol that decouples causal ordering from the BRB layer, achieving constant-size $\mathcal{O}(1)$ message metadata overhead and $\mathcal{O}(n^2)$ communication word complexity as against $\mathcal{O}(n^3)$ communication word complexity of existing protocols; here $n$ is the number of processes. We present two variants of our protocol: a cryptographic version using a threshold encryption scheme and sequence gating, and its non-cryptographic version. In the cryptographic version, senders broadcast ciphertexts immediately, and decryption shares are piggybacked on out-of-band ACKs, preventing early decryption and front-running. In both versions, causal safety is achieved probabilistically. We evaluate the probability of causal safety violations using a random variable path analysis under independent exponential link delay distributions. We show that both variants satisfy liveness and the probability of weak safety violation is bounded by $\mathcal{O}(f^{-3}\cdot\ln^3 f)$, where $f$ is the upper bound on the number of Byzantine processes, and $f < n/3$ and $f=\mathcal{O}(n)$. Further, for the crypto version, we show that the probability of strong safety violation is bounded by $\mathcal{O}(f^{-1} \cdot \ln^2 f)$. We also show how to modify our two protocols to guarantee 100\% weak safety keeping $\mathcal{O}(1)$ message space overhead but with $\mathcal{O}(n^3)$ messages and $\mathcal{O}(n^3)$ communication word complexity.

cs.DC↗