faragon

The Intel D270 ([1]) has similar silicon size to an AMD Radeon R9700 ([2]) and the AMD Radeon XT 9070 ([3]), and about twice the matrix performance. So the price is mainly for the memory. Assuming "normal prices", e.g. 600 USD for 128GB, 2250 USD for 480GB, and the GPU chip itself being 600 USD if produced in enough volume, including industrial profit, its "normal price" could be less than 3000 USD. I.e., it is a slow chip with lots of RAM for covering the case of having many resident models, not for running very big models, which is a very interesting use case for AI hobbyists, and likely also for some use cases in the data centers.

[1] https://www.techpowerup.com/gpu-specs/d270.c4452

[2] https://www.techpowerup.com/gpu-specs/radeon-ai-pro-r9700.c4...

[3] https://www.techpowerup.com/gpu-specs/radeon-rx-9070-xt.c422...

Is it 100% deterministic?

New post by @nand2mario. This is the continuation of his work in his z486 FPGA core ([1]), now adding the implementation of the original Voodoo Graphics GPU. In addition to the linked blog post, it is also published a video ([2]), and a SD image for testing ([3]) with the Xilinx KV260 dev kit, as this time the FPGA used in the MiSTer FPGA project is not enough for fitting the CPU core and the GPU.

[1] https://news.ycombinator.com/item?id=49308269

[2] https://x.com/nand2mario/status/2096133116066992568

[3] https://x.com/nand2mario/status/2096454368929485184

Accurate or not, there are many risks other than those that could bring AI companies value to 0. E.g., systems not requiring passing over all the data continuously, but in small regions running at L2-L3 cache memory speed. Once someone figures out that, the need for high bandwidth memory would end.

Games, very few, if any. But software in general, quite a lot.

This is the closer we have for having a full 32-bit x86 system with FPU. Even with the lower precision in the x87 FPU by using the FPGA DSP blocks, eventually having a "486DX" (i486 instruction set + FPU) would enable running a massive amount of PC software (1980..2000's). Taking his z386 core (based in the i386 microcode) and adding 486-class optimizations, including a simplified dual I/D L1 cache, vs the unified but more complex L1 cache in i486's, really is a beautiful and imaginative approach.

8K decoding needs HW acceleration, unless having 8-16 cores with AVX-512.

1) open-weights

2) critical mass

3) de-facto monopoly

4) closed-weights on frontier models

5) profit

Thank you very much! Is the SSFv2 mapper supporting memory writes in the MiSTer FPGA, too? (Megadrive and/or 32x cores). Also, now that there's a SH3 core already available ([1] for replicating the Cave CV1000 system), it could be mind blowing having a system with 512MB of RAM on a SH3 at 100MHz running Linux.

[1] https://github.com/ika-musume/HS3

They may consider rewriting their web-based apps into native apps using IA.

Microsoft degrading user experience plus lower price on Mac computers it maybe their downfall. E.g., removing local accounts, unwanted advertising, arbitrary decisions (forced TPM requirement), etc.

"All your Copilot are belong to us"

Wasn't their target to replace humans with AI? Why is that not applicable to their business?

IIRC, the original CGA/EGA version was: 8x 360KB 5.25" disks, or 4x 720KB 3.5" disks.

Maybe it's time for creating simpler languages, easier to validate with a mix of LLMs, grammar validation, and test-covered aggregation/integration.

It's "fair": ARM takes 1% of the iPhone CPU, i.e., less than 0.1% of the total phone price, while Apple takes 30% of the apps in the iStore

With the workforce may happen like with DRAM and NAND flash memories: unexpected demand in one side leaving without enough offer in other sides.

My question is how they'll update the hardware, if AI processors get outdated every 4-5 years? Or they will keep outdated satellites just for routing and Internet-related services?

Is it working for phone calls when running Debian and Windows?

Why Microsoft stores the encryption keys of the users in their servers? Key recovery is convenient, but in my opinion it should exist the "opt out" option, without MS being involved in the key storage in their datacenters.

My prediction: Google will win. Rationale: cost/token + pre-existing ad knowledge + APIs for third-party apps. My concern is how an OpenAI collapse could hurt Microsoft and Nvidia.

Why Apple don't sponsor Asahi Linux? E.g., I would buy Apple computers gladly for running Linux.

That's beautiful. E.g., I have Chinese coworkers and now I realize what they meant with "that's not wrong". I felt those expressions like humbleness, but now I realize they were trying to keep their authenticity. Mind blowing.

Why Apple don't just buy Intel? That would allow better vertical integration.

Empathy and humility are ok when shown by behavior, not when used as a "manipulative card" against "ego". E.g., when the "empathy card" is used to manipulate others in "you're an ego-driven person that always knows better than me". In my opinion there are less confrontative ways, e.g., a team lead handling the frictions with procedures, diplomacy, and other soft skills.

Duralex were used and loved in Spain, too. Most houses still have some of them, although nowadays there's lot more competition. I would pay twice for their glassware, quality-wise is good stuff.

I used Dillo in 2001-2002 with the PlayStation 2 Linux kit. With 32MB of RAM for Linux, and 4MB for the graphics. It worked really well despite the CPU was 294MHz (MIPS R-5900, two-way in-order CPU, with SIMD unit, and only one hardware thread, having two auxiliary vector units as companions).

VT102 specs, e.g., 8085A 6MHz CPU [1]

[1] http://adb.arcadeitalia.net/dettaglio_mame.php?game_name=vt1...