一个AI模型展示了理解和解释复杂技术文档的惊人能力，特别是1978年产的Central Data 64K DRAM板。该AI不仅对电路板的操作进行了清晰的解释，还识别出了示意图中的一个小错误，并纠正了用户对某些组件的理解。这种先进的理解能力凸显了AI在技术分析和解决问题方面的潜力，即使是在历史计算领域。 AI
影响 展示了AI在理解和解释复杂技术文档方面日益增长的能力，可能有助于历史硬件分析。
排序理由 用户报告AI能够理解和解释技术文档，包括识别错误。
在 Mastodon — mastodon.social 阅读 →
AI 生成摘要 · Google Gemini · 来自 3 个来源。 我们如何撰写摘要 →
完整方法见我们的编辑标准。
And it caught onto a bit of design flaw: the board does the DRAM refreshes using full RAS+CAS read cycles instead of RAS-only refresh cycles. It just doesn't push the data onto the system bus during a refresh. Which is fine, as far as it goes; a read will refresh the memory just …
Last night, while chatting with a friend who's really into AI, he just handed off the whole reference manual to AI, and asked it to explain how it works. A short time later, it popped out this: https:// cdc64k-ram-lab.noddson.chatgpt .site/ Well, god damn. It's... really good! It…
I've been trying to figure out why my 1978-vintage Central Data 64K DRAM board isn't working quite right. (S-100 bus, for those retro enough to know what that means.) Writes to it have about a 50% chance of working. Half the time, a write just does nothing, or writes only some of…