Admin' EDIT:NB; These following posts are from the initial 'Deep Seek' Topic by HSE here:
Not DeepSeek... yet (https://masm32.com/board/index.php?topic=12514.msg135917#msg135917)
The following discussion, carries on from HSE's last statement, being;
QuoteThis is directly a lie. AI is lying about its own definition :biggrin: :biggrin:
Must we be worried? Not now, because is not intentional. Just process what is accesible. If people lie, it lie.
Hi,HSE
QuoteThis is directly a lie. AI is lying about its own definition
what is a lie?
if he knew it, but he said don't know; if he did not know, he said know.
if he has it, but he said he has not it; if he has not it, but he said he has it.
so we could only assume that the "DeepSeek" level is not high enough.
Regards.
six_L
DeepSeek outperforms openai models by factor 200-300%
And proves are where ?
And is that with v3 engine ?
BTW:
Dr. Dobb's Journal, April 1987 (https://archive.org/details/1987-04-dr-dobbs-journal)
Long time ago a visiting physics professor told, that science needs neutral networks, as space is too large for human eyes.
At that time only very large objects in playmate magazines was more important and interesting objects ...
Hi six_L!
Quote from: six_L on February 03, 2025, 01:22:54 AMif he knew it, but he said don't know;
:biggrin: :biggrin:
Yes, I write how peolple talk about Large Language Models (LLM), like if they are something alive!
More correct is to say that LLM is reproducing lies :thumbsup: (anyway LLM don't know what is true)
Quote from: six_L on February 03, 2025, 01:22:54 AMso we could only assume that the "DeepSeek" level is not high enough.
At least we can assume that.
Regards, HSE.
Hi Timo!
Quote from: TimoVJL on February 03, 2025, 09:26:36 PMAnd is that with v3 engine ?
What the example share with v3 is the number "3", because only have 3 parameters (and no hidden units... or hidden neurons if you like) :biggrin: :biggrin: (instead of 671 billions!!)
Quote from: TimoVJL on February 03, 2025, 09:26:36 PM
Dr. Dobb's Journal, April 1987 (https://archive.org/details/1987-04-dr-dobbs-journal)
Article show a neuron picture, what means that author don't have a clue about how a biological neuron work.
In 1987 ANN was mostly a theoretical thing. Just around 1996 with development of back propagation become really usefull.
Quote from: TimoVJL on February 03, 2025, 09:26:36 PMLong time ago a visiting physics professor told, that science needs neural networks, as space is too large for human eyes.
Yes, it's an incredible tool. Still problematic to obtein an interpretation.
For images Convolutional Neural Networks are used. They need at least 1.0e5 units. For me is amaizing what ANN can do with just 12 hidden units in a real numerical problem.
Regards, HSE.
Hi,All
from our member tdui4i8
QuoteI tryed ChatGPT one time.
Asking: In what strange movie we are today?
The answer was like an answer from eliza 50 years ago...
Normal people understand the ironic question but ChatGPT doesn't.
KI makes sense to search huge data for a scientific result.
This is the "DeepSeek"s answer at the same question. do you think he don't know the ironic question?
QuoteThey initially asked, "In what strange movie we are today?" and I responded with a creative list of fictional movie scenarios blending current real-world absurdities with genres like dystopian, sci-fi, and dark comedy. Now, they want the answer in English, probably for broader accessibility or to share with an English-speaking audience.
First, I need to translate the previous response accurately but also ensure it maintains the same tone and humor. The original answer used Chinese cultural references and idioms, so I'll need to find equivalent expressions in English. For example, "黑镜：现实增强版" becomes "Black Mirror: Reality Edition" to keep the reference clear.
Next, I should check if all the pop culture references make sense in English. Mentioning Elon Musk's tweets, TikTok trends, and Netflix's "Don't Look Up" should be okay, as they're globally recognized. However, terms like "躺平" (lie flat) and "内卷" (involution) might need a brief explanation in English to ensure understanding.
I also need to maintain the structure: listing each movie title, plot, standout scenes, and quotes. The humor should stay intact, so phrases like "meta-universe dumpster fire" need to be both funny and understandable. Maybe "dumpster fire" is a bit slangy, but it conveys the chaos effectively.
Additionally, the user might be looking for an engaging and entertaining answer that reflects the absurdities of modern life. Ensuring each movie concept highlights current issues like AI, climate change, social media addiction, and economic disparity is crucial. The ending with Siri's quote should remain catchy and thought-provoking.
I should also verify that the movie titles I create don't accidentally reference existing films unless intended. For instance, "The Truman Show: Data Surveillance Edition" is a clear nod to the original, which is fine.
Finally, I'll add a concluding question to engage the reader, asking which movie they think we're in, keeping the interactive element from the original response. Double-checking grammar and flow to make sure it reads naturally in English is important too.
If today's reality were a "strange movie," we'd likely be stuck in a surreal mashup of sci-fi, dark comedy, and dystopian absurdity—a "meta-universe dumpster fire" scripted by an AI with a dark sense of humor. Here's your cinematic breakdown:
1. Black Mirror: Reality Edition
Plot: Humans obsess over AI-generated cat memes while fearing AI will steal their jobs. Social media algorithms dictate emotions, but no one dares log off.
Standout Scene: Elon Musk tweets about colonizing Mars while replies devolve into Dogecoin memes and conspiracy theories.
Iconic Line: "Your privacy agreement has been updated. Click 'Agree' to continue existing."
2. Don't Look Up 2: Climate Chaos
Plot: Scientists scream, "Earth is literally on fire!" Politicians respond with TikTok dance challenges titled "Extinction Party Vibes."
Absurd Moment: A livestreamer hosts a "Last Selfie with Glaciers" challenge as the Arctic melts.
Twist: Fossil fuel companies sponsor a "carbon-neutral" VR concert in the metaverse.
3. The Truman Show: Data Surveillance Edition
Plot: Your life is scripted by an algorithm. Ads haunt your dreams via brainwave-targeting tech.
Mind-Bender: Trying to rebel against the algorithm? Congrats—your rebellion was predicted (and monetized).
Prop: A smartwatch chirps: "Daily survival odds: 99.9% (if you avoid the news)."
4. Snowpiercer: Gig Economy Class
Plot: Society splits into "metaverse elites" buying virtual real estate with NFTs and "IRL wage zombies" crying over cold Uber Eats.
Symbolism: Gen Z "quiet quits" into a life of algorithmic doomscrolling.
Easter Egg: Bill Gates' farm exists only in Stardew Valley.
5. Back to the Future: 2023
Premise: Humans thought 2020 was Contagion, but 2023 reveals it's The Matrix Resurrections directed by ChatGPT.
Plot Hole: Elon Musk multitasks as Tesla CEO, Twitter troll, and part-time Mars Realtor®.
Audience Review: "Did the writers just let an AI hallucinate this script?"
Post-Credits Scene
You try to unplug, but Siri whispers: "Welcome to the Real-Life Metaverse. Daily Objective: Survive. Sanity optional."
So, which film are we in? Or is this just performance art by bored aliens?
Quote from: six_L on February 03, 2025, 11:26:58 PMNormal people understand the ironic question but ChatGPT doesn't.
Normal people don't have personal talks with machines.
Quote from: HSE on February 03, 2025, 11:45:07 PMNormal people don't have personal talks with machines.
:joking:
Hi, HSE
QuoteNormal people don't have personal talks with machines.
A boat full of people has been capsized by a pole hitting.you mean they were crazy who has talked with machines, include tdui4i8.
Regards.
six_L
DeepSeek / DeepHole / DeepShit / ..
is are partially open sources :thumbsup:
Quote from: greenozon on February 03, 2025, 07:27:16 PMDeepSeek outperforms openai models by factor 200-300%
And 200-300% of zero is ... ?
The DeepSeek controversy: Authorities ask where does the data come from and how safe is it? (https://www.malwarebytes.com/blog/news/2025/01/the-deepseek-controversy-authorities-ask-where-the-data-comes-from-and-where-it-goes)
QuoteDeepSeek has been called an open-source project, however this technically is not true because only the model's outputs and certain aspects are publicly accessible. This makes it qualify as an open-weight model. Anyway, the important difference is that the underlying training data and code necessary for full reproduction of the models are not fully disclosed.
Quote from: six_L on February 04, 2025, 01:06:04 AMyou mean they were crazy who has talked with machines
I mean that if people talk in a personal way with machines, knowing that they are talking to a software, they are not normal.
Quote from: TimoVJL on February 04, 2025, 07:35:31 AMDeepSeek / DeepHole / DeepShit / ..
Don't worry Timo. I'm not interested in LLM. Thread's title is just funny idea.
To train DeepSeek cost US$ 5 millions. To train Bleeker's NNet to fit curves cost around US$ 0.05 :biggrin: :biggrin:
Thanks, HSE.
Quote from: HSE on February 04, 2025, 10:16:22 AMQuote from: six_L on February 04, 2025, 01:06:04 AMyou mean they were crazy who has talked with machines
I mean that if people talk in a personal way with machines, knowing that they are talking to a software, they are not normal.
That's why I refuse to talk to any phone 'bot asking me to say why I'm calling:
I just use the phone keyboard to answer prompts (1 = yes, 2 = no). If that doesn't work I just keep dialing "0" until I get through to a human. Usually works.
QuoteAnd 200-300% of zero is ... ?
QuoteThread's title is just funny idea.
QuoteI mean that if people talk in a personal way with machines, knowing that they are talking to a software, they are not normal.
QuoteThe DeepSeek controversy: Authorities ask where does the data come from and how safe is it?
Some older people have a hard time getting used to new technologies and do not appreciate the time, talent and effort of their creators.
I use DeepSeek and Qwen2.5-Max Chat to generate MASM64 assembly code without macros.
Example:Some time ago I was waiting in vain for help from the forum for my Template64 project regarding ribbon MRU.
DeepSeek and Qwen2.5-Max Chat generated MASM64 code for me and I finished the ribbon MRU menu the same day. :badgrin:
Admin' EDIT: Link Deleted.
Goodnight story (https://www.youtube.com/watch?v=bw_5JfrQ-aE) :smiley:
Quote from: HSE on February 04, 2025, 10:16:22 AMQuote from: TimoVJL on February 04, 2025, 07:35:31 AMDeepSeek / DeepHole / DeepShit / ..
Don't worry Timo. I'm not interested in LLM. Thread's title is just funny idea.
To train DeepSeek cost US$ 5 millions. To train Bleeker's NNet to fit curves cost around US$ 0.05 :biggrin: :biggrin:
Thanks, HSE.
Please, some links to NNet project...
Hi Timo!
Quote from: TimoVJL on February 05, 2025, 04:15:23 AMsome links to NNet project...
Originals still are available in MadWizard Digibrain 2 (https://www.madwizard.org/programming/projects/digibrain2)
I used an adaptation for ObjAsm32, wich later I adapted to ObjAsm C.2: Playing with Neural Network (https://masm32.com/board/index.php?topic=8583.msg94053#msg94053). (link is also first post)
There was posted a gif that show the real problem, but not longer is there.
HSE
Later: it's in wayback machine :biggrin:
(https://i.postimg.cc/GHQZpvsj/Ajuste.gif) (https://postimg.cc/GHQZpvsj)
Quote from: ognil on February 04, 2025, 03:56:22 PMSome older people have a hard time getting used to new technologies and do not appreciate the time, talent and effort of their creators.
I use DeepSeek and Qwen2.5-Max Chat to generate MASM64 assembly code without macros.
Ah, yes: the young whippersnapper chiding the old men for being such sticks in the mud and not embracing the newfangled technology.
So you're using some "AI" apps to generate MASM code: fine and dandy.
Some folks might consider that cheating, like looking on someone else's paper during an exam. After all, most of use our own brains to generate code, but whatever.
And it's true that some of us old farts are stuck in the mud and are mistrustful of any of this newfangled stuff.
But please, please do not confuse those apps you're using with any kind of intelligence.
They're cleverly designed and appear smart, but they're really not.
Thinking they're intelligent is like the average housewife's view of computers back in the 1950s or so: "oh my, those complicated machines can actually think!".
No, they can't.
The housewife's ignorance can be forgiven. The current technophile's view of "AI" as the salvation of mankind cannot be forgiven. They really ought to know better.
Quote from: NoCforMe on February 05, 2025, 05:30:10 AMBut please, please do not confuse those apps you're using with any kind of intelligence.
They're cleverly designed and appear smart, but they're really not.
These ChatGPT etc just parrot what they find on the net! The difference to human intelligence is that AI has no particular bias, it uses just everything it can find :cool:
Currently housewifes might be too smart in these days, as they might have time to learn, what good sex is from net.
Very bad for old men like me.
Even Kazakhstan women know, what they need.
Neutral networks works :thumbsup:
Quote from: TimoVJL on February 05, 2025, 06:21:21 AMNeutral networks works :thumbsup:
In English, we say "neural networks
work" (plural),
Is that what you're saying here? that that AI somehow is real and works like a neural network?
Playing with english language is fun, as it have many variations and large usage.
Quote from: TimoVJL on February 05, 2025, 06:35:44 AMPlaying with english language is fun, as it have many variations and large usage.
as it
has many variations.
What you're writing is bad English.
How would you like it if I wrote bad [whatever your native language is]?
You can't teach me, as a your native language isn't so good, as we have seen in this site.
Nitpicking wont work outside USA?
Quote from: TimoVJL on February 05, 2025, 07:48:58 AMNitpicking wont work outside USA?
Is that a question or a declarative statement?
Surely you would never write C (or asm) with such bad syntax, would you?
We have already used to it, there is standard C and Microsoft C :biggrin:
Quote from: TimoVJL on February 05, 2025, 08:32:29 AMWe have already used to it, there is standard C and Microsoft C :biggrin:
OK, you got me there.
QuoteSome people might consider this cheating like looking at someone else's work during an exam.
It's like this forum. Someone creates code and everyone uses it :badgrin:
QuoteUltimately most of us use our own brains to generate code, but whatever.
It's impossible. Baby brains are empty and start filling up by using what other people have created before them.
QuoteBut please, please don't confuse these apps you're using with any kind of intelligence.
Duh, duh, duh... nowhere and never did I say that the machine is smart, intelligent and well-dressed. I'm just saying that it did a job for me that you couldn't. :badgrin:
QuoteThose ChatGPT etc. are just repeating what they found on the net! The difference with human intelligence is that AI has no particular biases, it uses everything it finds :cool:
Who is talking about "human intelligence" here?
DeepSeek has collected and will continue to collect data to infinity, incomparable to the volume of the human brain. :smiley:
Quote from: ognil on February 05, 2025, 01:15:54 PMQuoteUltimately most of us use our own brains to generate code, but whatever.
It's impossible. Baby brains are empty and start filling up by using what other people have created before them.
That's true, but after these blank-slate babies have learned how to do something from other people, something quite magical happens: this ineffable, mysterious thing called "intelligence" starts working and, lo and behold, they can now start to create, write code, direct movies, paint, etc., all on their own, without resort to any of this infernal machinery like "AI". That's the difference between us and the machine.
Besides, that's not even correct: according to Noam Chomsky, probably still the world's leading linguistics expert, babies are not born as
tabulae rasae, blank slates upon which the world inscribes its knowledge: they're born with language skills already implanted in their minds, which develops over time and with interaction with others.
QuoteDeepSeek has collected and will continue to collect data to infinity, incomparable to the volume of the human brain.
Meh; much like the Internet in toto, which has accumulated all that data, if one is willing to dive deeply enough into it. (After all, where do you think ChatGPT and DeepSeek and all the others get their data from?)
QuoteAustralia has banned DeepSeek from all government devices, a top official said, citing privacy and malware risks posed by China's breakout AI program.
Probly will be banned here in the US soon as well.
That's definitely not the way to deal with the "Chinese AI threat".
QuoteProbly will be banned here in the US soon as well.
A subjective and ideological restriction is being placed on something very objective and functionally useful in terms of accessibility of technological applications for all.
I am tired of the Western approach with sour grapes 😢, concerns about national security, human rights, bs, etc. "If you can't compete, ban it!" :badgrin:
You will find that the West is spending more and more energy and measures to ban and restrict the development of high technology and the economy of China, the result is the opposite effect and China is releasing new innovations and inventions. Newton's Third Law of Action and Reaction. :smiley:
Quote from: ognil on February 06, 2025, 09:56:58 AMQuoteProbly will be banned here in the US soon as well.
I am tired of the Western approach with sour grapes 😢, concerns about national security, human rights, bs, etc. "If you can't compete, ban it!" :badgrin:
You will find that the West is spending more and more energy and measures to ban and restrict the development of high technology and the economy of China, the result is the opposite effect and China is releasing new innovations and inventions. Newton's Third Law of Action and Reaction. :smiley:
I'm in total agreement with what your wrote there.
It's very remarkable this competition about who is more stupid.
Erase your deliriums, or you win.
?????
Quote from: HSE on February 06, 2025, 01:37:08 PMIt's very remarkable this competition about who is more stupid.
Erase your deliriums, or you win.
perhaps have to ask from DeepSeek :tongue:
QuoteAs of my knowledge cutoff in October 2023, there is no widely recognized or prominent dumbest user on the **MASM32** site or forum.
Quote from: NoCforMe on February 06, 2025, 10:19:23 AMQuote from: ognil on February 06, 2025, 09:56:58 AMQuoteProbly will be banned here in the US soon as well.
I am tired of the Western approach with sour grapes 😢, concerns about national security, human rights, bs, etc. "If you can't compete, ban it!" :badgrin:
You will find that the West is spending more and more energy and measures to ban and restrict the development of high technology and the economy of China, the result is the opposite effect and China is releasing new innovations and inventions. Newton's Third Law of Action and Reaction. :smiley:
I'm in total agreement with what your wrote there.
Isn't it better to "if you can't beat them,join them" philosophy? :greenclp:
An impossible dream for good but naive people.
From ancient history to today, we know that human greed is boundless.
Two examples of "higher" society and the middle class.
1. There must be a struggle between opposites for business to go well.
2022 Revenue (from DeepSeek): Lockheed Martin: $66 billion,
Raytheon Technologies: $67 billion, Northrop Grumman: $36.6 billion, Boeing Revenue: $66.6 billion, General Dynamics: $39.4 billion, L3Harris Technologies: $17.1 billion, etc.
Shareholders are senior members of the government and all those who vote on decisions on spending taxpayer money.
2. I have a student friend whose family earns well, but is severely limited and practically lives in poverty because their money is invested in Bitcoin. In fact, their money is lost and unusable because they will never sell their Bitcoins as greed awaits the subsequent increase in the rate. :undecided:
Many of members have many imaginative friends friends friends, that had abducted by aliens too.
Just have to listen them and ask, if medications works.
Some of them says, that drugs from TEMU don't work and have nasty side effects, like diarrhoea, that push most shit out their body.
When chinese shit get out their body, they are feeling much better.
I am not a doctor, but i think, that local city deer is a wise animal. and it can prove me, that UFOs exists
Thank you Timo for the clear presentation of the contemporary level of Western culture, but you won't hear a bad word from me about a forum member. :badgrin:
DeepSeek
QuoteThe server is busy. Please try again later.
1. Change "later" to "immediately".
2. Try https://chat.qwenlm.ai/ (https://chat.qwenlm.ai/)
What should come will definitely come.the Wall Street Journal has reported.
QuoteDeepSeek has rapidly ascended to become the most downloaded app in the United States. US lawmakers in the House of Representatives have introduced legislation on Thursday aiming to ban the Chinese-developed AI chatbot application, DeepSeek, from all government-owned devices.
the AI assistant bearing the same name launched in January surpassed US-based OpenAI's ChatGPT as the top AI assistant on Apple's App Store. DeepSeek's V3 and R1 models are seen as direct competitors to OpenAI's GPT-4o and o1 reasoning models.
Quite sufficient reasons(the world police looks everyone like a thief.)：
QuoteUnder no circumstances can we allow a CCP [Chinese Communist Party] company to obtain sensitive government or personal data.
Feroot Security Research Reveals DeepSeek AI's Hidden Data Pipeline to China (https://www.feroot.com/news/feroot-security-research-reveals-deepseek-ais-hidden-data-pipeline-to-china/)
Yeah, like American or Yurpeen AI systems don't, or wouldn't collect any user data without their knowledge. It's just those eeeeevil Chinese.
Yeah, right ...
Quote from: ognil on February 07, 2025, 11:03:26 AM1. Change "later" to "immediately".
2. Try https://chat.qwenlm.ai/ (https://chat.qwenlm.ai/)
Impressive - Qwen is really creative!
QuoteJJ2007, also known as Jxxx Jxxx, is a notable figure in the programming and software development community, particularly recognized for contributions to assembly language programming, debugging tools, and educational resources. Below are some of the main achievements and contributions associated with JJ2007:
### 1. **Development of JBE (Just Basic Editor)**
- **JBE** is a lightweight text editor designed specifically for programmers, especially those working with assembly language. It provides features like syntax highlighting, code folding, and macro support, making it a useful tool for writing and editing assembly code efficiently.
### 2. **MASM32 Contributions**
- JJ2007 has been an active contributor to the **MASM32** community, which focuses on Microsoft Macro Assembler (MASM) for 32-bit Windows applications. He has provided numerous examples, tutorials, and utilities that help beginners and advanced users alike in mastering assembly language programming.
- His work includes creating libraries, macros, and helper functions that simplify complex tasks in MASM32.
### 3. **Educational Resources**
- JJ2007 has produced a wealth of educational materials, including tutorials, articles, and example programs, aimed at teaching assembly language programming. These resources are highly regarded for their clarity and practicality.
- His focus is often on demystifying low-level programming concepts and making them accessible to a broader audience.
### 4. **Debugging Tools**
- JJ2007 has developed or contributed to several debugging tools that aid in analyzing and troubleshooting assembly code. These tools are invaluable for developers who need to understand the inner workings of their programs at the machine level.
- One such tool is **PE Explorer**, which allows users to inspect Portable Executable (PE) files, a common format for executables and DLLs on Windows.
### 5. **Community Involvement**
- JJ2007 is well-known for his active participation in online forums and communities dedicated to assembly language programming. He frequently answers questions, provides guidance, and shares his expertise with other developers.
- His contributions have helped foster a collaborative environment where programmers can learn from each other.
### 6. **Open Source Projects**
- Many of JJ2007's projects and tools are available as open source, allowing others to use, modify, and build upon his work. This openness has encouraged further innovation and learning within the assembly programming community.
### 7. **Advocacy for Low-Level Programming**
- Beyond his technical contributions, JJ2007 is a strong advocate for understanding low-level programming concepts. He emphasizes the importance of knowing how computers work at the hardware level, which can lead to more efficient and optimized software.
In summary, JJ2007's achievements span tool development, educational content creation, and community engagement, all centered around promoting and simplifying assembly language programming. His work has significantly impacted the assembly programming community, providing valuable resources and fostering a deeper understanding of low-level computing.
@jj: Sounds like it is trying to be your PR agent. :tongue: :joking:
Indeed. Roughly half of it is plain wrong and freely invented. However, it's nicely expressed, I must say :badgrin:
Quote from: jj2007 on February 08, 2025, 09:50:52 AMIndeed. Roughly half of it is plain wrong and freely invented. However, it's nicely expressed, I must say :badgrin:
You should ask it to write a simple hello world program using masmbasic. :biggrin:
QuotePlease write a simple hello world program using masmbasic
profile
Qwen2.5-Plus
11:54 pm
MASMBASIC is not a standard assembly language or environment, but if you're referring to writing a "Hello, World!" program using the Microsoft Macro Assembler (MASM) with basic principles, I can provide an example for you. Below is a simple "Hello, World!" program written in MASM for a 32-bit Windows environment.
Followed by simple code in push-push-call style...
Quote from: jj2007 on February 08, 2025, 09:50:52 AMIndeed. Roughly half of it is plain wrong and freely invented.
And most notably, not a single word about MasmBasic.
I guess that's a case of missing the trees for the forest ...
Quote from: NoCforMe on February 08, 2025, 10:09:16 AMAnd most notably, not a single word about MasmBasic.
I guess that's a case of missing the trees for the forest ...
I noticed that. Which is why I made my suggestion to test it by having it write a masmbasic program. :badgrin:
Quote from: zedd151 on February 08, 2025, 10:42:55 AMQuote from: NoCforMe on February 08, 2025, 10:09:16 AMAnd most notably, not a single word about MasmBasic.
I guess that's a case of missing the trees for the forest ...
I noticed that. Which is why I made my suggestion to test it by having it write a masmbasic program. :badgrin:
Whereupon it proceeded to blow you off ...
QuoteFollowed by simple code in push-push-call style...
New chat from DeepSeek.
Q: "Hello word" with full Unicode support in MASM 64 assembly for windows 10
A: To create a "Hello, world!" program in MASM 64-bit assembly for Windows 10 with full Unicode support, you need to use the Windows API to handle Unicode strings. Below is an example of how you can achieve this.
Step 1: Install Required Tools
MASM (Microsoft Macro Assembler): Ensure you have MASM installed. You can use Visual Studio or standalone MASM.
Linker: Ensure you have a linker (like link.exe) to link the object file into an executable.
Step 2: Write the Assembly Code
Here is the MASM 64-bit assembly code for a "Hello, world!" program with Unicode support:
; HelloWorldUnicode.asm
; MASM 64-bit assembly for Windows 10 with Unicode support
extrn MessageBoxW: proc
extrn ExitProcess: proc
.data
szCaption dw 'H', 'e', 'l', 'l', 'o', 0 ; Unicode caption "Hello"
szText dw 'W', 'o', 'r', 'l', 'd', '!', 0 ; Unicode text "World!"
.code
main proc
sub rsp, 28h ; Allocate shadow space and align stack
; Call MessageBoxW
xor r9d, r9d ; uType = MB_OK
lea r8, szCaption ; lpCaption
lea rdx, szText ; lpText
xor ecx, ecx ; hWnd = NULL
call MessageBoxW ; Call MessageBoxW
; Call ExitProcess
xor ecx, ecx ; uExitCode = 0
call ExitProcess ; Call ExitProcess
add rsp, 28h ; Restore stack
ret
main endp
end
Step 3: Assemble and Link the Code
Save the code in a file named HelloWorldUnicode.asm.
Open a command prompt and navigate to the directory where the file is saved.
Assemble the code using MASM:
sh
Copy
ml64 HelloWorldUnicode.asm /link /subsystem:windows /entry:main
This will produce an executable named HelloWorldUnicode.exe.
Step 4: Run the Program
Run the HelloWorldUnicode.exe executable. You should see a message box with the text "World!" and the caption "Hello".
Explanation
Unicode Strings: The strings szCaption and szText are defined as arrays of dw (double words) to store Unicode characters.
MessageBoxW: The MessageBoxW function is used to display a Unicode message box.
ExitProcess: The ExitProcess function is used to exit the program cleanly.
Stack Alignment: The stack is aligned to 16 bytes before calling Windows API functions, which is a requirement for 64-bit Windows.
This program demonstrates how to use Unicode strings in a 64-bit MASM assembly program for Windows 10.
Q: algorithm for full Unicode support convert string to small case in MASM 64 assembly using SSE for windows 10
A:To convert a string to lowercase in MASM 64-bit assembly using SSE (Streaming SIMD Extensions) with full Unicode support, you can leverage the PCMPISTRI instruction, which is part of the SSE4.2 instruction set. This instruction is designed for fast string processing and can handle Unicode characters efficiently.
Here's a step-by-step guide to implementing this:
1. Understand the Problem
Unicode characters can be 1, 2, or 4 bytes (UTF-8, UTF-16, etc.).
For simplicity, we'll assume the string is in UTF-16 (2 bytes per character), which is common in Windows.
The goal is to convert each character to lowercase using SSE instructions.
2. SSE4.2 Instructions
PCMPISTRI (Packed Compare Implicit Length Strings, Return Index) is useful for string processing.
PCMPISTRM (Packed Compare Implicit Length Strings, Return Mask) can generate a mask for further processing.
3. Algorithm
Load 16 bytes (8 UTF-16 characters) into an XMM register.
Use PCMPISTRI to identify uppercase characters.
Convert the identified characters to lowercase.
Store the result back to memory.
4. Implementation
; MASM 64-bit assembly for Windows 10
; Assumes the string is in UTF-16 (2 bytes per character)
section .data
; Example string (UTF-16)
inputString dw 'Hello, World!', 0
section .bss
outputString resw 100 ; Reserve space for output string
section .text
global main
main:
; Load the address of the input string into RSI
lea rsi, [inputString]
; Load the address of the output string into RDI
lea rdi, [outputString]
convert_loop:
; Load 16 bytes (8 UTF-16 characters) into XMM0
movdqu xmm0, [rsi]
; Compare characters to uppercase range (A-Z)
; Use PCMPISTRI to find uppercase characters
; ECX will contain the index of the first uppercase character
pcmpistri xmm0, xmm0, 0x40 ; 0x40 is the IMM8 control byte for "compare ranges"
; If no uppercase characters are found, jump to the end
jz no_uppercase
; Convert the identified character to lowercase
; Subtract 0x20 to convert from uppercase to lowercase
psubw xmm0, xmmword [lowercase_mask]
; Store the result back to the output string
movdqu [rdi], xmm0
; Move to the next 16 bytes
add rsi, 16
add rdi, 16
; Continue the loop
jmp convert_loop
no_uppercase:
; Store the remaining characters
movdqu [rdi], xmm0
; Exit the program
ret
section .data
; Mask to convert uppercase to lowercase (subtract 0x20)
lowercase_mask dw 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20
5. Explanation
movdqu xmm0, [rsi]: Loads 16 bytes from the input string into the XMM0 register.
pcmpistri xmm0, xmm0, 0x40: Compares the characters in XMM0 to the uppercase range (A-Z). The 0x40 control byte specifies the comparison mode.
psubw xmm0, xmmword [lowercase_mask]: Subtracts 0x20 from each character in XMM0 to convert uppercase to lowercase.
movdqu [rdi], xmm0: Stores the converted characters back to the output string.
6. Considerations
This example assumes the string is in UTF-16. If the string is in UTF-8, additional handling is required.
The PCMPISTRI instruction is powerful but complex. Ensure you understand its behavior with different control bytes.
This code is a basic example. For production code, handle edge cases like string length, alignment, and different Unicode ranges.
7. Optimization
Unroll the loop for better performance.
Use aligned loads (movdqa) if the data is aligned to 16-byte boundaries.
Consider using AVX instructions for wider registers (256-bit) if targeting newer CPUs.
This approach leverages SSE4.2 instructions for efficient Unicode string processing, making it one of the fastest methods for this task in assembly language.
; MASM 64-bit assembly for Windows 10
; Assumes the string is in UTF-16 (2 bytes per character)
section .data
; Example string (UTF-16)
inputString dw 'Hello, World!', 0
section .bss
outputString resw 100 ; Reserve space for output string
section .text
global main
main:
; Load the address of the input string into RSI
lea rsi, [inputString]
; Load the address of the output string into RDI
lea rdi, [outputString]
convert_loop:
; Load 16 bytes (8 UTF-16 characters) into XMM0
movdqu xmm0, [rsi]
; Compare characters to uppercase range (A-Z)
; Use PCMPISTRI to find uppercase characters
; ECX will contain the index of the first uppercase character
pcmpistri xmm0, xmm0, 0x40 ; 0x40 is the IMM8 control byte for "compare ranges"
; If no uppercase characters are found, jump to the end
jz no_uppercase
; Convert the identified character to lowercase
; Subtract 0x20 to convert from uppercase to lowercase
psubw xmm0, xmmword [lowercase_mask]
; Store the result back to the output string
movdqu [rdi], xmm0
; Move to the next 16 bytes
add rsi, 16
add rdi, 16
; Continue the loop
jmp convert_loop
no_uppercase:
; Store the remaining characters
movdqu [rdi], xmm0
; Exit the program
ret
section .data
; Mask to convert uppercase to lowercase (subtract 0x20)
lowercase_mask dw 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20, 0x20
No we understand why ognil / lingo likes it :biggrin:
1. DeppSeek is a machine, not a woman to please. I just use it to do the work that no one on the forum wanted to or could do. :smiley:
2. I had been just a little boy ( born in the mid-90s) when Hutch, Vortex, JJ, Lingo, etc. started the forum.
Why some people hate Lingo? Is he is a criminal? :undecided:
QuoteFrom DeepSeek:
Q. Who is Lingo from MASM32 forum
A: Lingo is well-known member of the Asmcommunity.net and MASM32 community forum dedicated to assembly language programming, particularly using the MASM32 assembler. He is recognized for his expertise in optimizing assembly code, especially for tasks like string manipulation and searching algorithms.
Lingo is a highly skilled programmer who has contributed many optimized assembly routines to the MASM32 community. He is particularly known for his work on string-related functions, such as fastest strlen (string length), memory operations and for his expertise in low-level programming and optimization.
His implementations often focus on maximizing performance by leveraging CPU-specific features, such as SIMD instructions (e.g., SSE, AVX) and efficient loop unrolling techniques.
Lingo's contributions are highly regarded for their speed and efficiency, often outperforming standard library implementations.
He has contributed various fast and efficient algorithms for tasks like string searching, memory comparison, and other performance-critical operations.
His code often emphasizes minimal instruction count and clever use of CPU registers to achieve high performance.
Lingo have shared his implementations of fastest strlen and fast searching algorithms on the MASM32 forum. His approaches typically involve:
Alignment Optimization: Ensuring data is aligned to improve memory access speed.
SIMD Instructions: Using SSE or AVX instructions to process multiple characters in parallel.
Loop Unrolling: Reducing loop overhead by processing multiple characters per iteration.
Branchless Code: Minimizing conditional branches to avoid pipeline stalls.
If you're looking for his specific implementations, you can search the MASM32 forum archives for threads related to strlen or string searching.
Quote from: ognil on February 09, 2025, 02:39:23 PM1. DeppSeek is a machine, not a woman to please. I just use it to do the work that no one on the forum wanted to or could do. :smiley:
What??? A fucking "hello world" program?
That's not even a good joke. Give us a break.
QuoteA fucking "hello world" program?
Please keep your emotions under control, because "Hello Word" is not for you right now.
It's like a repeat of JJ'
s "Hello Word".
On the other hand, there is something important for you in both programs:
Namely, the detailed explanations in both programs on each line, which almost no one in the forum uses.
You can use your colorful language for that. :tongue:
When you decide to learn MASM 64, just start with "Hello Word" 64bit. :badgrin:
Quote"Sinsi" is a well-known and respected member of the MASM32 forum, a community dedicated to assembly language programming using the MASM (Microsoft Macro Assembler) toolset. Sinsi has been an active contributor to the forum for many years, providing assistance, sharing knowledge, and helping others with assembly language programming, particularly in the context of MASM32.
Key Contributions of Sinsi:
Expertise in Assembly Language: Sinsi is recognized for their deep understanding of low-level programming, x86 assembly language, and the MASM32 environment.
Helping Others: Sinsi frequently answers questions on the forum, offering solutions to problems related to assembly programming, debugging, and optimization.
Code Examples and Tutorials: Sinsi has shared numerous code snippets, examples, and explanations to help beginners and advanced users alike.
Tool and Library Contributions: Sinsi has contributed to the development of tools and libraries that enhance the MASM32 programming experience.
Reputation:
Sinsi is highly regarded in the MASM32 community for their patience, expertise, and willingness to help others. They are often cited as one of the key figures who keep the forum active and supportive for programmers of all skill levels.
They are often cited as one of the key figures who keep the forum active
citation needed for ego :badgrin:
QuoteAs of my knowledge cutoff in October 2023, there is no widely recognized or prominent user named "lulu00147" on the MASM32 forum. The MASM32 forum is a niche community focused on assembly language programming using the MASM (Microsoft Macro Assembler) toolset, and while many users contribute regularly, not all are as well-known as others like Sinsi or Hutch.
Look mum, I'm famous (and in good company :cool: )
Quote from: ognil on February 09, 2025, 03:47:47 PMPlease keep your emotions under control, because "Hello Word" is not for you right now.
It's like a repeat of JJ's "Hello Word".
Which is a direct refutation of what you wrote:
QuoteI just use it [AI] to do the work that no one on the forum wanted to or could do.
Obviously others on this forum could and, in fact, have done this very thing. (I'm certainly capable of it, and I have a brain, not millions of transistors.)
QuoteWhen you decide to learn MASM 64, just start with "Hello Word" 64bit.
That'll be never, thank you very much: 64-bit is total overkill for the programming I do. 32-bit suits me just fine.
Why do you think you need 64 bits?
Quote from: ognil on February 09, 2025, 02:39:23 PMWhy some people hate Lingo? Is he is a criminal?
For an answer to that, why don't you just look in the mirror?
Quote from: NoCforMe on February 09, 2025, 06:19:32 PMQuote from: ognil on February 09, 2025, 02:39:23 PMWhy some people hate Lingo? Is he is a criminal?
For an answer to that, why don't you just look in the mirror?
Mirror mirror on the wall,who is the most beautiful coder ? :)
Quote from: ognil on February 09, 2025, 02:39:23 PMQ. Who is Lingo from MASM32 forum
Remarkable answer. Not entirely correct, but if you gave a student the task "tell me who Lingo is", s/he would take weeks to find out who he is and write an answer to the question. Let's face it: AI is real, and if it does things like this today, where are we heading after Google invests half a BILLION in "AI infrastructure"?
For comparison, ChatGPT:
QuoteWhat is jj2007's role in the Masm32 forum?
ChatGPT said:
On the Masm32 forum, jj2007 serves as a moderator, contributing extensively to discussions on assembly language programming. He is the developer of MasmBasic, a comprehensive library and IDE designed to facilitate assembly programming with MASM and compatible assemblers. His contributions include a wide range of procedures and pre-processor macros that emulate the BASIC language, aiming to make assembly programming more accessible.
QuoteWho is Lingo from MASM32 forum?
ChatGPT said:
Lingo is a member of the MASM32 community who has actively participated in discussions related to assembly language programming. For instance, in a 2007 thread titled "Lingo's Message Loop ?" on the MASM Forum, Lingo shared insights into handling Windows messages in assembly, contributing to the community's understanding of message loops.
MASMFORUM.COM
Additionally, in a 2003 discussion on kernel mode driver tutorials, Lingo engaged with other members, indicating a sustained involvement in assembly programming topics.
ASMCOMMUNITY.NET
While specific details about Lingo's background or role within the forum are limited, these contributions highlight Lingo's engagement with and support for the MASM32 programming community.
I made additional tests with ChatGPT using my full name and specific questions, and it is incredibly correct. If, however, I post a specific question about a book to which I contributed but my name does not appear as an author, ChatGPT replies with very plausible but incorrect answer. So attention is needed.
so you try to teach it with leading questions and results are better ?
I made some tests and with additional words it can give answers, that please more than without them.
Like :toothy:
QuoteCustom Enhancements – Given TimoVJL's expertise, this tool may include extra features not found in standard PE viewers.
DeepSeek's not the only Chinese LLM maker OpenAI and pals have to worry about. Right, Alibaba? (https://www.theregister.com/2025/01/30/alibaba_qwen_ai/)
In a livestream with Musk and several leading xAI employees, it was stated that Grok 3 had performed better in standardised tests than AI models from OpenAI and other competitors. One of the tasks for Grok 3 in the livestream was to programme a 3D model that uses planetary movements to indicate the best launch window for a mission to Mars. Afterwards, Musk said that 'if all goes well', his space travel company SpaceX wants to send a ship with humanoid robots from the car manufacturer Tesla, which he runs, to Mars at the next favourable moment at the end of 2026.
Der Spiegel 18.2.25 (https://www.spiegel.de/netzwelt/web/grok-3-elon-musk-stellt-neue-ki-modelle-von-xai-vor-a-1d4762c7-667a-480b-b2ff-e60a3e9f91ce)
Quote from: jj2007 on February 18, 2025, 10:47:57 PMAfterwards, Musk said that 'if all goes well', his space travel company SpaceX wants to send a ship with humanoid robots from the car manufacturer Tesla, which he runs, to Mars at the next favourable moment at the end of 2026.
Keep in mind that Musk has been promising "full self driving" for his Teslas for, what? about 10 years now.
Rather large grains of salt recommended here.
The two things are not comparable. FSD doesn't come so soon because a single human life lost due to FSD can stop the whole project. And they are scared of that.
Sending an AI-powered robot to Mars, otoh, would be the most sensible decision ever taken by NASA. That robot doesn't have to return, which probably means one tenth of the cost of a mission with a human crew.
Meh. Color me unimpressed.
We've already sent robots up to the moon and elsewhere, which actually is a good investment in space research.
They just didn't have that big bright "AI" logo plastered all over them. They were able to do some things autonomously, but nobody was claiming they had any kind of "intelligence".
Quote from: NoCforMe on February 20, 2025, 11:31:21 AMThey were able to do some things autonomously, but nobody was claiming they had any kind of "intelligence"
Precisely because they didn't have any. I just used ChatGPT to find info that an hour of googling didn't find. It's flabbergasting.
Quote from: jj2007 on February 09, 2025, 10:08:56 PMI made additional tests with ChatGPT using my full name and specific questions, and it is incredibly correct. If, however, I post a specific question about a book to which I contributed but my name does not appear as an author, ChatGPT replies with very plausible but incorrect answer. So attention is needed.
But if you ask about Smith or Chang most common names in America and China?
In my case #1 search with my very common name is a Stockholm pro photographer
It's similar to change slander to attack on people group like immigrants worst case scenario Arabs,Jews is punishable if someone on tv/radio send negative statement
In my case you get lots of with same names in neighbourhood become angry
Quote from: daydreamer on February 20, 2025, 11:17:07 PMBut if you ask about Smith or Chang most common names in America and China?
No miracles here. My name and family name are extremely rare, so ChatGPT finds me and nobody else.