My presentation in 2026 was about Accelerating machine vision and computation tasks with Sparsity, Vulkan and SPIR-V.<\/p>\n
SPIR-V Wave Intrinsics<\/strong> (subgroup operations) \u2022 Waves\/Subgroups act as a group of threads executed in lockstep (typically 32 or 64 lanes)<\/p>\n \u2022 The most critical part of wave organization is how the hardware handles Sparsity (if-statements)<\/p>\n \u2022 The registers are “sliced” across the wave. When you call a wave intrinsic like WaveActiveSum(), the Then I talked about the imaging approach for 3D micrscopy using sparsity.<\/p>\n Sparsity of Illumination<\/strong> \u2022 By activating a single, incoherent point-light source (LED) at a time, we ensure that the photon \u2018origin\u2019 is \u2022 This also results in an arrangement which is not light starved as the LED is able to illuminate the camera Sparsity of the Subject<\/strong> \u2022 In many biological contexts and areas like photonics there are large homogeneous areas that are punctuated \u2022 This capability is particularly evident in low-contrast biological samples. For example, visualizing the endpiece of a sperm flagella is typically too small and low contrast. However, because the feature is sparse even My presentation in 2026 was about Accelerating machine vision and computation tasks with Sparsity, Vulkan and SPIR-V. SPIR-V Wave Intrinsics […]<\/p>\n","protected":false},"author":1,"featured_media":2638,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1,14],"tags":[],"class_list":["post-2617","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-uncategorized","category-blog"],"_links":{"self":[{"href":"https:\/\/www.davidbrebner.com\/index.php?rest_route=\/wp\/v2\/posts\/2617","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.davidbrebner.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.davidbrebner.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.davidbrebner.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.davidbrebner.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=2617"}],"version-history":[{"count":5,"href":"https:\/\/www.davidbrebner.com\/index.php?rest_route=\/wp\/v2\/posts\/2617\/revisions"}],"predecessor-version":[{"id":2645,"href":"https:\/\/www.davidbrebner.com\/index.php?rest_route=\/wp\/v2\/posts\/2617\/revisions\/2645"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.davidbrebner.com\/index.php?rest_route=\/wp\/v2\/media\/2638"}],"wp:attachment":[{"href":"https:\/\/www.davidbrebner.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=2617"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.davidbrebner.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=2617"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.davidbrebner.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=2617"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}
\n\u2022 A powerful feature for SIMD (Single Instruction, Multiple Data) processing on GPUs, allowing threads
\nwithin a “wave” to exchange data without explicit barriers<\/p>\n
\nhardware doesn’t send data to\/from RAM, it uses an internal Crossbar Switch or Shuffle Network<\/p>\n
\n\u2022 Standard wide-field microscopy use K\u00f6hler bathing the subject in uniform light. This intentionally reduces
\nthe phase contrast of the system as it isolates the focal plane. The Light rays arriving from opposing angles
\nmutually cancel the shadow casting and phase gradients that define high-frequency surface details<\/p>\n
\nlargely spatially distinct. This sparsity results in maximum phase conversion at all depths of field. This allows
\nthe system to recover sharp edge responses even in regions that would be considered “out of focus” or lowcontrast in a traditional bright-field arrangement<\/p>\n
\nsensor and is only blocked where light is absorbed or refracted away allowing very high-speed capture with
\nminimal shot noise<\/p>\n
\n\u2022 The second governing principle is the sparsity of the subject relative to the detector. The system is designed
\nto significantly oversample the optical resolution limit (100 nm pixel pitch vs. \u2248 980 nm Rayleigh limit)<\/p>\n
\nwith “sparse” small features separated from neighbouring structures by distances exceeding the Rayleigh
\ncriterion (separated enough they do not blur together). When an isolated sub-resolution feature is imaged
\nwith this sparse illumination strategy, it behaves as a discrete signal source rather than part of a convoluted
\ntexture<\/p>\n
\nthe weak gradient of the photometric response of the flagella tip can be isolated and reconstructed. The
\ninverse rendering algorithm detects this sparse signal by its consistent orientation response across the
\ndifferent off axis lighting angles<\/p>\n","protected":false},"excerpt":{"rendered":"