{"id":103157,"date":"2022-02-01T15:36:44","date_gmt":"2022-02-01T13:36:46","guid":{"rendered":"https:\/\/prohoster.info\/blog\/novosti-interneta\/metod-identifikaczii-sistemy-polzovatelya-na-osnove-informaczii-o-gpu"},"modified":"2022-02-01T15:36:44","modified_gmt":"2022-02-01T13:36:46","slug":"metod-identifikaczii-sistemy-polzovatelya-na-osnove-informaczii-o-gpu","status":"publish","type":"post","link":"https:\/\/prohoster.info\/hu\/blog\/news\/metod-identifikaczii-sistemy-polzovatelya-na-osnove-informaczii-o-gpu","title":{"rendered":"A felhaszn\u00e1l\u00f3i rendszer azonos\u00edt\u00e1s\u00e1nak m\u00f3dszere a GPU inform\u00e1ci\u00f3i alapj\u00e1n","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"
A Ben-Gurion University (Israel), Lille University (France), and University of Adelaide (Australia) researchers developed a new technique for identifying user devices through the analysis of GPU performance parameters in web browsers. The method is called 'Drawn Apart' and is based on the use of WebGL to obtain a performance profile of the GPU, significantly enhancing the accuracy of passive tracking methods that operate without the use of cookies or the retention of identifiers on the user's system. <\/p>\n
Methods that consider the rendering characteristics, GPU, graphics stack, and drivers for identification have been used previously, but they were limited to separating devices only at the level of different graphics card models and GPUs, meaning they could only serve as an additional factor to increase identification probability. A key feature of the new 'Drawn Apart' method is that it is not limited to distinguishing different GPU models but attempts to identify differences among identical GPUs of the same model, caused by inconsistencies in the manufacturing processes tailored for massive parallel computations. The variations encountered during manufacturing allow unique fingerprints to be formed for the same device models.
<\/center> <\/p>\n
It turned out that these differences can be revealed by counting the number of execution blocks and analyzing their performance in the GPU. Checks based on a set of trigonometric functions, logical operations, and floating-point computations were used as primitives to detect different GPU models. To identify differences in identical GPUs, the number of simultaneously executing threads during vertex shader execution was evaluated. It is presumed that the observed effect is caused by differences in the thermal regimes and energy consumption of different chip instances (a similar effect was previously demonstrated for CPUs \u2014 identical processors executing the same code showed different energy consumption). <\/p>\n
Mivel a WebGL m\u0171veletek aszinkron m\u00f3don futnak, a v\u00e9grehajt\u00e1si id\u0151 m\u00e9r\u00e9s\u00e9re nem lehet k\u00f6zvetlen\u00fcl a JavaScript API performance.now()-t haszn\u00e1lni, ez\u00e9rt h\u00e1rom tr\u00fckk\u00f6t javasoltak az id\u0151 m\u00e9r\u00e9s\u00e9re: <\/p>\n
\n
onscreen \u2014 rendering a scene in an HTML canvas while measuring the callback function's execution time set via the Window.requestAnimationFrame API and called after the rendering is complete.\n
offscreen \u2014 a worker and rendering scene in the OffscreenCanvas object with the execution time measurement for the convertToBlob command.\n
GPU \u2014 rendering in the OffscreenCanvas object, but using for measurement the timer provided in WebGL that considers the execution duration of a set of commands on the GPU side. <\/ul>\n
Az azonos\u00edt\u00f3 l\u00e9trehoz\u00e1s\u00e1nak folyamat\u00e1ban minden eszk\u00f6z\u00f6n 50 ellen\u0151rz\u00e9st v\u00e9geznek, amelyek mindegyike 16 k\u00fcl\u00f6nb\u00f6z\u0151 jellemz\u0151 176 m\u00e9r\u00e9s\u00e9t tartalmazza. Az 2500 eszk\u00f6z \u00e9s 1605 k\u00fcl\u00f6nb\u00f6z\u0151 GPU inform\u00e1ci\u00f3inak gy\u0171jt\u00e9se sor\u00e1n v\u00e9gzett k\u00eds\u00e9rlet 67%-os hat\u00e9konys\u00e1gn\u00f6veked\u00e9st mutatott a kombin\u00e1lt azonos\u00edt\u00e1si m\u00f3dszerek eset\u00e9n a Drawn Apart t\u00e1mogat\u00e1s\u00e1nak hozz\u00e1ad\u00e1s\u00e1val. K\u00fcl\u00f6n\u00f6sen a kombin\u00e1lt FP-STALKER m\u00f3dszer \u00e1tlagosan 17,5 napos azonos\u00edt\u00e1st biztos\u00edtott, m\u00edg a Drawn Aparttal p\u00e1ros\u00edtva az azonos\u00edt\u00e1s id\u0151tartama 28 napra n\u0151tt.
<\/center> <\/p>\n
\n
The separation accuracy for 10 systems with Intel i5-3470 (GEN 3 Ivy Bridge) chips and Intel HD Graphics 2500 GPU in the onscreen test was 93%, while in offscreen \u2014 36.3%.\n
A 10 Intel i5-10500 (GEN 10 Comet Lake) rendszer eset\u00e9n, NVIDIA GTX1650 grafikus k\u00e1rty\u00e1val, az pontos\u00e1g 70% \u00e9s 95,8% volt.\n
For 15 systems with Intel i5-8500 (GEN 8 Coffee Lake) and Intel UHD Graphics 630 GPU \u2014 42% and 55%.\n
For 23 systems with Intel i5-4590 (GEN 4 Haswell) and Intel HD Graphics 4600 GPU \u2014 32.7% and 63.7%.\n
For six smartphones Samsung Galaxy S20\/S20 Ultra with Mali-G77 MP11 GPU, the identification accuracy in the onscreen test was 92.7%, while for smartphones Samsung Galaxy S9\/S9+ with Mali-G72 MP18 \u2014 54.3%. <\/ul>\n
<\/center> <\/p>\n
Megjegyezt\u00e9k, hogy a GPU h\u0151m\u00e9rs\u00e9klete befoly\u00e1solta a pontos\u00e1got, \u00e9s egyes eszk\u00f6z\u00f6k eset\u00e9n a rendszer \u00fajraind\u00edt\u00e1sa torz\u00edtotta az azonos\u00edt\u00f3t. A m\u00f3dszer m\u00e1s k\u00f6zvetett azonos\u00edt\u00f3 m\u00f3dszerekkel val\u00f3 kombin\u00e1l\u00e1s\u00e1val a pontoss\u00e1g l\u00e9nyegesen n\u00f6velhet\u0151. A pontoss\u00e1got a WebGPU \u00faj API-j\u00e1nak stabiliz\u00e1l\u00e1sa ut\u00e1n a sz\u00e1m\u00edt\u00e1si shaderek alkalmaz\u00e1s\u00e1val is tervezik n\u00f6velni. <\/p>\n
Az Intel, ARM, Google, Khronos, Mozilla \u00e9s Brave c\u00e9geket m\u00e1r 2020-ban \u00e9rtes\u00edtett\u00e9k a probl\u00e9m\u00e1r\u00f3l, de a m\u00f3dszer r\u00e9szletei csak most ker\u00fcltek nyilv\u00e1noss\u00e1gra. A kutat\u00f3k JavaScript \u00e9s GLSL nyelven \u00edrt m\u0171k\u00f6d\u0151 p\u00e9ld\u00e1kat is k\u00f6zz\u00e9tettek, amelyek k\u00e9pesek megjelen\u00edteni az inform\u00e1ci\u00f3t, de an\u00e9lk\u00fcl is m\u0171k\u00f6dhetnek. Emellett az Intel GEN 3\/4\/8\/10 GPU-alap\u00fa rendszerekhez adathalmazokat is kiadtak az inform\u00e1ci\u00f3k oszt\u00e1lyoz\u00e1s\u00e1ra a g\u00e9pi tanul\u00e1s rendszereiben. \n Forr\u00e1s: opennet.ru<\/a> <\/p>","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"excerpt":{"rendered":"
![\"A](\"\/wp-content\/uploads\/2022\/02\/947fc4477ff99c14ca7854e36e4ad976.jpg\")
<\/center> <\/p>\n![\"A](\"\/wp-content\/uploads\/2022\/02\/c57096ca8d5833df92cd7f992f19a289.jpg\")
<\/center> <\/p>\n![\"A](\"\/wp-content\/uploads\/2022\/02\/e5462f5b761dc6e16a43ac00d8f0b7aa.jpg\")
<\/center> <\/p>\n