From mboxrd@z Thu Jan 1 00:00:00 1970 X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on inbox.vuxu.org X-Spam-Level: X-Spam-Status: No, score=-1.0 required=5.0 tests=DKIM_SIGNED,DKIM_VALID, MAILING_LIST_MULTI,RCVD_IN_DNSWL_NONE,RCVD_IN_MSPIKE_H2 autolearn=ham autolearn_force=no version=3.4.4 Received: (qmail 31232 invoked from network); 23 Aug 2021 05:23:32 -0000 Received: from tb-ob20.topicbox.com (173.228.157.66) by inbox.vuxu.org with ESMTPUTF8; 23 Aug 2021 05:23:32 -0000 Received: from tb-mx0.topicbox.com (tb-mx0.nyi.icgroup.com [10.90.30.73]) by tb-ob20.topicbox.com (Postfix) with ESMTP id 9335324862 for ; Mon, 23 Aug 2021 01:23:29 -0400 (EDT) (envelope-from bounce.mM40ea45711a1551fd53807b84.r522be890-2105-11eb-b15e-8d699134e1fa@9fans.bounce.topicbox.com) Received: by tb-mx0.topicbox.com (Postfix, from userid 1132) id 4E65C332D5E7; Mon, 23 Aug 2021 01:23:29 -0400 (EDT) ARC-Authentication-Results: i=2; topicbox.com; arc=pass; dkim=none (no signatures found); dmarc=pass policy.published-domain-policy=none policy.applied-disposition=none policy.evaluated-disposition=none (p=none,d=none,d.eval=none) policy.policy-from=p header.from=sirjofri.de; spf=pass smtp.mailfrom=sirjofri+ml-9fans@sirjofri.de smtp.helo=oat.nine.sirjofri.de; x-internal-arc=fail (as.1.topicbox.com=pass, ams.1.topicbox.com=fail (message has been altered)) (Message modified while forwarding at Topicbox) ARC-Message-Signature: i=2; a=rsa-sha256; c=relaxed/relaxed; d= topicbox.com; h=date:from:to:message-id:in-reply-to:references :subject:mime-version:content-type:content-transfer-encoding :list-help:list-id:list-post:list-subscribe:reply-to :list-unsubscribe; s=sysmsg-1; t=1629696209; bh=fYjVAPPFLhiEX+Lp sUsNiNnZIutG7XfqiExI+spqSz8=; b=VnpRyokDILKzcGjXMVekXupMUtSBNwHf qgXgqlO5vyPedvWSkDOAyX8zpBBOIpER+N8EN3qY/Kpw9/HyucHPh/oDvEdnBV3v 7IqwDj8lVeHLsF3JSY73NUefyQv4fKfHehQ3OZ/USvgt0R09hVUfzzT/eQXzeQ3/ jxPBsmEnBSA= ARC-Seal: i=2; a=rsa-sha256; cv=pass; d=topicbox.com; s=sysmsg-1; t= 1629696209; b=m233DHjZVgr+jhTzXXTSX3Hi0BYhg3qtljEtCNV3QM21SSAdXy U+9qhkWdiNIhA12nZyxfcRuAnKLoKEMJ/63c2FmnScsnnmAq4TYRwDM7j7QxigLX pj7LYjkofMLe/CsxpB/ExOYZCXXrcobT5/UnNRwZxkzHVVqaaNz+fl4gE= Authentication-Results: topicbox.com; arc=pass; dkim=none (no signatures found); dmarc=pass policy.published-domain-policy=none policy.applied-disposition=none policy.evaluated-disposition=none (p=none,d=none,d.eval=none) policy.policy-from=p header.from=sirjofri.de; spf=pass smtp.mailfrom=sirjofri+ml-9fans@sirjofri.de smtp.helo=oat.nine.sirjofri.de; x-internal-arc=fail (as.1.topicbox.com=pass, ams.1.topicbox.com=fail (message has been altered)) (Message modified while forwarding at Topicbox) X-Received-Authentication-Results: tb-mx0.topicbox.com; arc=none (no signatures found); bimi=skipped (DMARC Policy is not at enforcement); dkim=none (no signatures found); dmarc=pass policy.published-domain-policy=none policy.applied-disposition=none policy.evaluated-disposition=none (p=none,d=none,d.eval=none) policy.policy-from=p header.from=sirjofri.de; iprev=pass smtp.remote-ip=5.45.105.127 (oat.nine.sirjofri.de); spf=pass smtp.mailfrom=sirjofri+ml-9fans@sirjofri.de smtp.helo=oat.nine.sirjofri.de; x-aligned-from=pass (Address match); x-me-sender=none; x-ptr=pass smtp.helo=oat.nine.sirjofri.de policy.ptr=oat.nine.sirjofri.de; x-return-mx=pass header.domain=sirjofri.de policy.is_org=yes (MX Records found: oat.nine.sirjofri.de); x-return-mx=pass smtp.domain=sirjofri.de policy.is_org=yes (MX Records found: oat.nine.sirjofri.de); x-tls=pass smtp.version=TLSv1.2 smtp.cipher=ECDHE-RSA-AES256-SHA smtp.bits=256/256; x-vs=clean score=0 state=0 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed; d=9fans.net; h=date:from :to:message-id:in-reply-to:references:subject:mime-version :content-type:content-transfer-encoding:list-help:list-id :list-post:list-subscribe:reply-to:list-unsubscribe; s=dkim-1; bh=fYjVAPPFLhiEX+LpsUsNiNnZIutG7XfqiExI+spqSz8=; b=ErWxz2DgJuYs yzsI2lU86G2+DNXndXiuEdjeNEbB69IZUkg91SW07PvJcXI9zV1BAKXypPGutslf Tw9QVjrf+JuYNdvgGUVeEK1T+SBC1BeyMBWUuEksj/SDLmHqH4Kdy0lVrT/UYs3M 8cZXKE31JdbFxEaTXGIeIpv0Luk1MsI= Received: from tb-mx0.topicbox.com (localhost.local [127.0.0.1]) by tb-mx0.topicbox.com (Postfix) with ESMTP id DE437332D1F1 for <9fans@9fans.net>; Mon, 23 Aug 2021 01:23:19 -0400 (EDT) (envelope-from sirjofri+ml-9fans@sirjofri.de) Received: from tb-mx0.topicbox.com (localhost [127.0.0.1]) by tb-mx0.topicbox.com (Authentication Milter) with ESMTP id 9D20A154677; Mon, 23 Aug 2021 01:23:19 -0400 ARC-Seal: i=1; a=rsa-sha256; cv=none; d=topicbox.com; s=arcseal; t= 1629696199; b=WEaiwVut7dCkxWYn6JTb7WnP4TcMJ67CaEkrms87hIAJ5gVRB/ mBB3D7A3tGBrRMyXj+AHDHjCg30FxHHWQoMRiFbIuDDC0Sz5Rm0MGeMNXOm1KMNa VdHtkqZ0Y04agS+2pp4vT5UBbWM/4PQ0KZ8qBLBDbjHIFzInkiYpssxHjsBFeQwo UIDIv+TQidpXWUu+DubxJabX7CoWUEfO+aeUxCkldzVi3gNmCLzj//jJRrBDY9Ov cOfQu979oVC3IyMXFDRcwroPWWJGRNUmzjj59yT5jof6UhAQCiuZJwF2XRSlwL1y XXZ0/pBx3dT/C7355AcjWjO1n89VEEo3+hQA== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d= topicbox.com; h=date:from:to:message-id:in-reply-to:references :subject:mime-version:content-type:content-transfer-encoding; s= arcseal; t=1629696199; bh=2vU6A7cB8LCDZs/SmCuNNF8K/P/GRQnSWn6r84 lNAhA=; b=Y1A+eD9lQwBWHEhlx9c1CWmMuJR2PQihUaKS/LECuhCXN8o+dKf/HW T95q1sF6aXG82fqeWfZKoSXEVArwlxIcViegcO83rBx9pzyfMP+kp6x5XvYFtiOt jiGh7Rlxh2otL0OhTdJNn4NphdD7R5o35ctcuMg6dMk9ewCBRy7nVp+EDZ91YoHh SOoLA45wUwdirEQvfCj1p7HqUJbRs3La+6Xy4alEn4ltHpLInzN+kfSLufBStrM2 UCVJNSJupmDl+sWD+txyTl5KjWR6dxiU3wX29ZJSph1+kpKoXjySuVuLECjpdAqp PRZ9KFJxj/eur50qHDwt2gndd/ZPK+IA== ARC-Authentication-Results: i=1; tb-mx0.topicbox.com; arc=none (no signatures found); bimi=skipped (DMARC Policy is not at enforcement); dkim=none (no signatures found); dmarc=pass policy.published-domain-policy=none policy.applied-disposition=none policy.evaluated-disposition=none (p=none,d=none,d.eval=none) policy.policy-from=p header.from=sirjofri.de; iprev=pass smtp.remote-ip=5.45.105.127 (oat.nine.sirjofri.de); spf=pass smtp.mailfrom=sirjofri+ml-9fans@sirjofri.de smtp.helo=oat.nine.sirjofri.de; x-aligned-from=pass (Address match); x-me-sender=none; x-ptr=pass smtp.helo=oat.nine.sirjofri.de policy.ptr=oat.nine.sirjofri.de; x-return-mx=pass header.domain=sirjofri.de policy.is_org=yes (MX Records found: oat.nine.sirjofri.de); x-return-mx=pass smtp.domain=sirjofri.de policy.is_org=yes (MX Records found: oat.nine.sirjofri.de); x-tls=pass smtp.version=TLSv1.2 smtp.cipher=ECDHE-RSA-AES256-SHA smtp.bits=256/256; x-vs=clean score=0 state=0 X-ME-VSCause: gggruggvucftvghtrhhoucdtuddrgedvtddruddtgedggeekucdltddurdegudehrddttd dmucetufdoteggodetrfdotffvucfrrhhofhhilhgvmecuhfgrshhtofgrihhlpdggtfgf nhhsuhgsshgtrhhisggvpdfurfetoffkrfgpnffqhgenuceurghilhhouhhtmecufedttd enucenucfjughrpeffhffvkfgjfhfugggtgfesthhqredttddtjeenucfhrhhomhepshhi rhhjohhfrhhiuceoshhirhhjohhfrhhiodhmlhdqlehfrghnshesshhirhhjohhfrhhird guvgeqnecuggftrfgrthhtvghrnhepffelffehvdetvdefleeludejueetkedtteehiedv ueevudekffehvdefhedvfeetnecuffhomhgrihhnpegrrhhiiihonhgrrdgvughupdhtoh hpihgtsghogidrtghomhenucfkphephedrgeehrddutdehrdduvdejpdelhedrledtrddv udekrddvudefnecuvehluhhsthgvrhfuihiivgeptdenucfrrghrrghmpehinhgvthephe drgeehrddutdehrdduvdejpdhhvghlohepohgrthdrnhhinhgvrdhsihhrjhhofhhrihdr uggvpdhmrghilhhfrhhomhepoehsihhrjhhofhhrihdomhhlqdelfhgrnhhssehsihhrjh hofhhrihdruggvqe X-ME-VSScore: 0 X-ME-VSCategory: clean Received-SPF: pass (sirjofri.de: 5.45.105.127 is authorized to use 'sirjofri+ml-9fans@sirjofri.de' in 'mfrom' identity (mechanism 'mx' matched)) receiver=tb-mx0.topicbox.com; identity=mailfrom; envelope-from="sirjofri+ml-9fans@sirjofri.de"; helo=oat.nine.sirjofri.de; client-ip=5.45.105.127 Received: from oat.nine.sirjofri.de (oat.nine.sirjofri.de [5.45.105.127]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) by tb-mx0.topicbox.com (Postfix) with ESMTPS for <9fans@9fans.net>; Mon, 23 Aug 2021 01:23:18 -0400 (EDT) (envelope-from sirjofri+ml-9fans@sirjofri.de) Received: from sirjofri.de ([95.90.218.213]) by oat; Mon Aug 23 07:23:15 +0200 2021 Date: Mon, 23 Aug 2021 05:23:14 +0000 (UTC) From: sirjofri To: 9fans <9fans@9fans.net> Message-ID: <0dff7dc8-4197-489c-8bb2-c1ef2dfca96f@sirjofri.de> In-Reply-To: <1007A30B-A109-41BA-BD4A-FF06AED8DFEC@iitbombay.org> References: <92764e35-f5cf-460a-91df-050ba471e6dd@sirjofri.de> <9352EE7C-AE94-4C1C-8738-5A1DA8ECE5A7@iitbombay.org> <1007A30B-A109-41BA-BD4A-FF06AED8DFEC@iitbombay.org> Subject: Re: [9fans] Drawterm GPU (was: Software philosophy) MIME-Version: 1.0 Content-Type: text/plain; charset="UTF-8"; format="flowed" Content-Transfer-Encoding: quoted-printable X-Correlation-ID: <0dff7dc8-4197-489c-8bb2-c1ef2dfca96f@sirjofri.de> Topicbox-Policy-Reasoning: allow: sender is a member Topicbox-Message-UUID: 3ce4b7a4-03d2-11ec-8b2f-f52be28fdc7d Archived-At: =?UTF-8?B?PGh0dHBzOi8vOWZhbnMudG9waWNib3guY29tL2dyb3Vwcy85?= =?UTF-8?B?ZmFucy9UYWQyOWJmYzIyM2RjNGZiZS1NNDBlYTQ1NzExYTE1NTFmZDUzODA3?= =?UTF-8?B?Yjg0Pg==?= List-Help: List-Id: "9fans" <9fans.9fans.net> List-Post: List-Software: Topicbox v0 List-Subscribe: Precedence: list Reply-To: 9fans <9fans@9fans.net> List-Unsubscribe: , Topicbox-Delivery-ID: 2:9fans:437d30aa-c441-11e9-8a57-d036212d11b0:522be890-2105-11eb-b15e-8d699134e1fa:M40ea45711a1551fd53807b84:1:mE9Crf_EBKt8yRSSQc3_qUZCVQKdf0anJk_jiD_f6cs Good morning, GPUs are SIMD processors. The hundreds of cores are great for highly=20 parallel calculation. In GLSL/HLSL I can write a program which is calculated for a very small=20 set of pixels (usually 2x2 or 1x1). So if you have=C2=A0 a resolution of 10= x10=20 the program is basically run 5x5=3D25 or 10x10=3D100 times in parallel. Boo= st=20 the resolution to more real values like 1080p you see how the many cores=20 benefit the whole calculation. This high parallelization can only really happen because most stuff is=20 independent. For example, when raytracing each ray is (almost)=20 independent of other rays. Still, the 2x2 matrix is often calculated dependently because for=20 texturing (and mip mapping) you need the "distance" between two pixels in=20 a fragment shader. This is why (for texturing) you may end up having=20 slower programs and some waiting time between some threads, because=20 sometimes you need the value of the neighbor thread and have to wait=20 until it's calculated. Well, these are very language-specific details that are important for=20 graphics, but apply similarly to other use cases. I can imagine that for=20 neural networks you can just write the code for one node nad execute it=20 500 times for 500 nodes in parallel. Imagine having this beast on the CPU=20 with just 4 cores... I hope this helps you understand how GPU cores ("shaders") work. Vulkan would indeed be interesting. Since we are only interested in the=20 compute part it might even make our programs really small, the "hello=20 world" part of drawing triangles would be the "client" side (writing a=20 rasterizer, raymarcher, tracer, whatever). It could still be a lot lines=20 of code, but maybe we still benefit from the 10% speedup. I still have to understand how all this "shader compilation" stuff works.=20 In webgl it's like, "here's my code, make a shader from it, then I tell=20 you it's a fragment shader". Shader compilation happens automatically. In=20 UE shader compilation takes a long time, and I believe also in blender=20 shaders are stored in a precompiled binaries. sirjofri 23.08.2021 06:13:53 Bakul Shah : > Don't high end GPUs have thousands of "cores"? Even high end CPUs don't=20 > have more than a few dozen cores to 128 or so. While each kind's cores=20 > are very different, seems to me GPU/CPU paths have diverged for good.=20 > Or we need some massive shift in programming languages + compilers. I=20 > lack imagination how. Still, the thought of the CPUs gaining the=20 > complexity of the graphics engine scares me! > > -- Bakul > > On Aug 22, 2021, at 12:09 PM, Paul Lalonde =20 > wrote: > > I'm pretty sure we're still re-inventing, though it's the CPU's turn to=20 > gain some of the complexity of the graphics engine. > > Paul > > On Sun, Aug 22, 2021, 12:05 PM Bakul Shah wrote: >> Thanks. Looks like Sutherland's "Wheel of=20 >> Reincarnation[https://www2.cs.arizona.edu/~cscheid/reading/myer-sutherla= nd-design-of-display-processors.pdf]"=20 >> has not only stopped but exploded :-) Or stopped being applicable. >> >> -- Bakul >> >> On Aug 22, 2021, at 9:23 AM, Paul Lalonde =20 >> wrote: >> >> It got complicated because there's no stable interface or ISA.=C2=A0 The= =20 >> hardware evolved from fixed-function to programmable in a commercial=20 >> environment where the only meaningful measure was raw performance per=20 >> dollar at many price points.=C2=A0 Every year the hardware spins and=20 >> becomes more performant, usually faster than Moore's law.=C2=A0 With 3D= =20 >> APIs hiding the hardware details there is no pressure to make the=20 >> hardware interface uniform, pretty, or neat.=C2=A0 And with the need for= =20 >> performance there are dozens of fixed function units that effectively=20 >> need their own sub-drivers while coordinating at high performance with=20 >> the other units.=C2=A0 >> The system diagrams for GPUs look complex, but they are radical=20 >> simplifications of what's really on the inside. >> >> Intel really pioneered the open driver stacks, but performance=20 >> generally wasn't there.=C2=A0 That might be changing now, but I don't kn= ow=20 >> if their recently announced discrete product line will be=20 >> driver-compatible. >> >> Paul >> >> >> On Sun, Aug 22, 2021 at 8:48 AM Bakul Shah =20 >> wrote: >>> The FreeBSD amdgpu.ko is over 3Mbytes of compiled code. Not counting=20 >>> the "firmware" that gets loaded on the GPU board. drm/amd/amdgpu has=20 >>> 200K+ lines of source code. drm/amd over 2M lines of code. Intel's=20 >>> i915 seems to be about 1/10th the amd size. AIUI, this is linux GPU=20 >>> driver code, more or less unchanged (FreeBSD has shim code to use=20 >>> it). How did the interface to an SIMD processor get so complicated? >>> >>> =E2=80=A6 >>> >>> >>> >>> -- Bakul >>> > > > *9fans[https://9fans.topicbox.com/latest]* / 9fans / see=20 > discussions[https://9fans.topicbox.com/groups/9fans] +=20 > participants[https://9fans.topicbox.com/groups/9fans/members] +=20 > delivery=C2=A0options[https://9fans.topicbox.com/groups/9fans/subscriptio= n]=20 > Permalink[https://9fans.topicbox.com/groups/9fans/Tad29bfc223dc4fbe-Me785= 13510ae4df2da186c73a] ------------------------------------------ 9fans: 9fans Permalink: https://9fans.topicbox.com/groups/9fans/Tad29bfc223dc4fbe-M40ea4= 5711a1551fd53807b84 Delivery options: https://9fans.topicbox.com/groups/9fans/subscription