One hundred and ninety-one
"I know who you are!" yelled Hennessy, almost startling for a long time, "it's amazing that the workstation that has been in the limelight lately is your work." You also say that you sell computers, you are so dishonest!"
"I'm a computer seller, and I'm not lying. "It's an honor for me to be remembered by Professor Hennessy." β
Hennessy smiled: "We bought one of your alpha, it's really a whimsical idea, especially those two processors, it's really beyond my expectations, 'sex' can be compared to Patterson's chips, I don't know which company's product you introduced." β
Long laughed: "That processor was made by the Americans." He doesn't want to say that it's his own credit, it's better to be a low-key person.
Hennessy was full of praise: "Very advanced, it is simply the perfect architecture I imagined, although the process is a bit backward, but it can be imagined that with the high integration of manufacturing technology, the architecture of this chip is unscrupulous." β
To be so praised by the future engineering academicians and industry leaders, I have been a comfortable mess for a long time, but I am also a little embarrassed, after all, the architecture of K32 absorbs the 'fine' Hua of these predecessors, and it is quite a bit of a feeling of victory.
I was waiting to be modest for a long time, but I heard Professor Hennessy say: "Originally, I still wanted to study it, but a few days ago I saw this processor was a little interested, other people's products are already very mature, I believe that it is the most prominent in this year's microprocessor research forum, and I want to 'communicate' with the inventor of this processor." β
I was ashamed for a long time, and it would be a big sin to let this master give up research because of him, so he hurriedly persuaded: "Dr. Hennessy, this is not good, my company is very welcome to you, there is a project that needs a great power like you to preside, *&%#Β₯Β₯#......"
Hennessy smiled: "You want to be bad, I didn't want to 'get' BM to talk to me a few years ago, but I refused, I'm just an academic staff and don't want to get involved in the business field." Originally, they agreed to fund me to work on RISC architecture, but now it doesn't seem to make much sense. β
Chang Chang understood a little, and immediately changed the conversation: "There is nothing wrong with this, I have some very difficult things to solve here, I think you will be interested." β
Hennessy laughed, "Listen, it's not good if you don't challenge 'sex'. β
"That's right. "For a long time, a little closer," microprocessors are just a development direction, since Neumann proposed the storage processing architecture, the processor, a component that is always important, has evolved from generation to generation, and now it is a bit of a branch. β
Professor Hennessy nodded in agreement: "That's good, the development of integrated circuit technology has made this possible, and now there are microprocessors, microcontrollers, embedded and so on. But what's the difference, when it comes to the basic structure, it's still the same. β
Chang Chang explained: "Yes, in my opinion, there are only two development directions, one is the general-purpose chip represented by the microprocessor, and the combination of software will make this architecture bigger and bigger, the 'sexuality' can be higher and higher, and more and more work can be done." β
Professor Hennessy patted his head: "That's true, what you said should be a special chip, this is not very easy to do, only for some companies, the market capacity is not large, it seems that there is no prospect." β
Chang smiled and said: "No, no, no, special ICs are not only custom-made. β
"And?" Hennessy frowned, "are you talking about DSPs?" β
"Worthy of a professional. "I gave a thumbs up for a long time, a little pat, and I still admire its erudition, you must know that it was not until 82 that the first commercial DSP was just launched by Texas Instruments, which is the most cutting-edge technology of our time.
"This ah......" Professor Hennessy looked at it for a long time, tapping rhythmically with his fingers, "Texas Instruments products, it can be regarded as a pioneering achievement, GPP (General Purpose Processor) can handle almost everything, but it is a little powerless to calculate the science and technology, but DSP has found another way, mainly for the number that represents the signal, especially good at dealing with mathematical formulas, how can you be interested in this thing?"
"I just think it's very useful, hehe. "When I was developing the graphics system of the workstation, I had a headache with these graph geometric transformations, and it was really difficult to solve them with ordinary methods, so I had to integrate the multiplier, and then I found that there was a DSP thing, which made me feel like I was blocked." β
"Graphics, I'm not good at that. Professor Hennessy thought about it for a moment and said, "But what you said is a bit of an interest to me, and it is a way forward, although I don't like Texas Instruments developing DSPs for the military, but I have great expectations for other uses of this thing." β
"Yes, analog signals are too difficult to deal with and costly, so it would be nice to represent all human signals in numbers." The TV cable is no longer transmitted by analog signals,Then there will be absolutely no clutter" waves,Snow "flakes"There will be no more,High definitionγ β
Professor Hennessy was stunned: "If the image signal is all digital, how much information will be needed, and it can't be processed." β
"That's why we developed a DSP, and it's absolutely fine to use a special 'gate' to process this kind of signal by solidifying various formulas into the circuit. "This is not only convenient, but also has a very high utilization rate of frequency bands, especially suitable for optical cable transmission, I can foresee that in the future all the networks will be simplified and integrated, and each household only needs a small optical cable in addition to the power line to complete all the tasks." β
"What about the phone?" asked Professor Hennessy.
"The phone is just transmitting sound signals, and this thing is also modulated into data, and then restored after transmission?"
Professor Hennessy's heart was pounding, he understood people, naturally knew that this was absolutely possible, and was about to speak, but he heard a voice coming.
"Mr. Cao, so you're here! Professor Hennessy, you're here too!"
Why is this so familiar, and when I looked up for a long time, it turned out to be an old acquaintance Grove, who also came to participate in this drinking party, and it seems that today is a gathering of great gods.
"Mr. Andy!" Hennessy stood up and hugged Grove, "I heard you already have a new product, why didn't you give me a copy of it?" β
Grove smiled: "There are fewer samples, and I will hand them to you shortly." You are so happy that you are hiding here with two old smoking guns, and many people are looking for you. β
Hennessy said: "If you don't go, it's better to be here." With those people, you will only quarrel, not as good as this Mr. Cao's chat. β
Grove said, "What are you talking about?" As he spoke, he greeted the 'waiter' and divided three glasses of wine.
"Just moisturizing the throat. "I was amazed by Professor Hennessy's knowledge, and we were discussing the direction of future processors. β
"It's RISC again. As soon as Grove heard it, he understood that Hennessy was a loyal follower of Patterson, "The 'fine' architecture is indeed excellent, but the 'sexual' performance of CISC products is not far behind, and you will know what this means when you see the finished product of 386." β
"That's your secret project, how do I know what 'sex' is?" said both Chang and Hennessy.
"It's simple, it's like this. Grove began to introduce 386's 'sex' and various 'sex', and at the end he praised his own products in passing.
However, Chang Chang didn't intend to make him so comfortable, and said loudly: "It's really strong, but I heard that Lovo has formed IBM's own team, and I don't plan to continue to use your chips." β
Grove was furious: "That 'bastard' of Lowo is too deceitful, this kind of thing will definitely not let him succeed! Now the market is not up to IBM alone, what do you say?"
laughed for a long time without answering, and Hennessy watched the two play dumb riddles inexplicably.
Grove didn't want to talk much in front of outsiders, so he changed the topic and said, "John, do you have any projects now?"
"I'm investigating, but it seems to be a bit of an eyebrow now, and I've been doing research on memory architecture recently, and I'm nearing the end. Hennessy replied.
Talking about this, Grove was full of energy, and said to Chang Chang: "John didn't say that I forgot, 286 has paid too much sacrifice in order to be compatible with 8086, and the memory used by the AT function has been limited to one megabyte, and it can no longer be used." β
For a long time, his heart moved, and he saw that the old boy's eyebrows were dancing and he must have made some new discoveries: "It's none of our business, IBM's designers have stipulated the management of memory space to death, godson can only use 640k, and the rest is left to the bottom of the BIOS and so on." If it weren't for these limitations, I think 286's 'sexual' energy would have been brought into greater play. β
"You're right, I suspect IBM did it on purpose to pave the way for their own processors," Grove said. However, they miscalculated, and Intel's engineers finally had a solution after B's unremitting efforts. β
I knew for a long time that the play 'meat' was coming, so I pretended to be surprised and said: "Oh, is that really the case? I thought VS was the first to solve this problem, but I didn't expect the hero to see the same thing!"
Grove was shocked, this 'door' trick to improve the memory management space technology is Intel's unexpected achievement, originally they wanted to let 386 integrate various modes to replace memory management, but the existence of IBM makes the AT machine do not have more than one trillion memory management capabilities, so as a last resort, Grove turned this technology out.
It's actually very simple, Intel just opens up a 64K space in the memory, which is used to store directories, reflect to other spaces, and with certain drivers, the system can access more than one trillion of memory.
The reason why Grove kept it secret was to make IBM suffer a little, but he never expected that he would say something amazing for a long time, and he actually claimed that VS also had similar technology, and hurriedly asked what it was.
I explained the ins and outs to Grove for a long time, and it turned out that when the AT machine was released in 84, he felt that the memory of this machine was a huge shortcoming, and the original 286 had 24 bit address lines, which was enough to support 16 trillion memory.
But compatibility is a double-edged sword, although it makes the 286 popular, and it is also in order to be compatible with the previous 8086 series, so I have to humiliate the price and continue to make do with the 640k memory limit.
IBM is also 'insidious', and the design of the AT machine also does not solve this problem, so that all kinds of programs are squeezed into such a small space, which is why the spreadsheet program released by Microsoft cannot do vs-office.
In order to solve this problem, the software of VS continued to grab a time gap, and it was decided to try a new technology in the new version of the chipset, which is derived from the memory of the extended memory EMS standard.
The so-called EMS originally had a name called LIM in my long memory, which means the first letter of the names of intel, lotus and microsoft, because this is advocated by these three software and hardware giants.
It's just that now lotus is not the king at all, and Microsoft has been driven into the mac field for a long time, and dos no longer exists, so EMS has been proposed by Intel alone today.
Although the specific specifications of this technology are not clear for a long time, the principle is very simple, and after some research and development, it has now blossomed, although it has not been integrated into the chipset due to process problems, but there are specific finished products.
After listening to the introduction for a long time, Grove didn't know what it was like, because this technology for a long time was similar to Intel, but it was nothing more than some details of the specification, and the essence was still the same.
"What a wonderful Creator, the results of our two families are so similar. Grove smiled stubbornly, a little embarrassed, after all, as allies and 'operational' system and hardware providers, the two actually did not communicate with each other, although there are interests for reasons, but it can also be seen that the alliance is not monolithic.
"It's a shame that Intel has already started this work. "This is a 'waste' of wisdom and financial resources, especially in the face of IBM, a common enemy, it seems that we still need a lot of 'exchange'." β
"That's, that's. Grove took the opportunity to step down, proposing, "Why don't we communicate with each other and work together to develop a set of specifications for the use of memory expansion, and allies can't kill each other." β
"Great goodness!"