Just to follow up with Coby's comment today on Crossbar Inc, he mentioned that Crossbar Inc did involved some non-standard material in the process and that might be one of the reasons their product did not make a big splash in the field.
Did a brief check on them and Crossbar pulled a collaberation deal with SMIC in China in 2016 and they achieved sample chips in 2017. However, that was in the news for a little while and then just like Coby have said, things disappeared. There is no much of follow up after that.
It turned out with Crossbar's production, silver is involved in the processing. I guess that is not as friendly as the WBT product.
Really, WBT is not the pioneer for producing ReRAM. HP, Panasonic, Fujutsu....... they all have achieved commercialisible level of ReRAM chips way before WBT, but none of them is using the material based on silicon oxide. The popularity of those product is not seen by far in the market and I guess that is for a very good reason.
Then Coby mentioned about the high endurance of WBT chip under high temperature, better than a lot other products. I think again, this type of silicon oxide based material might not be the thoughest one, i.e., MRAM can achieve much higher endurance or high tempeature resistance by its nature but also comes with a very high cost.
In semiconductor industry, it is always about the sweet spot. For large usage product, that has to be cheap enough and easy enough to manufacture. Just as Coby has mentioned that the embedded chip WBT can offer into a car, the life span of the chip itself is even longer than the car. So lets say, if this chip can make a car running reliably for 20 years, why one would even worry about using the best of the best chip to make the car running for 200 years?
Therefore, this makes WBT in a very unique postion and sweet spot as they got the chip parameters fall in the high end but manufacturing requirements stay in the low end. Just for that reason, among the current ReRAM market, maybe that is right that there is no competetor with WBT.
Then another unique feature with WBT's future discrete memory is this selector/resistor combination. Many companies make ReRAM claim that they can produce large size array for data center, but the memory array they do is structuring transistor with resistor. In embedded memory, with low density of bit cell, each of the single transistor can control single bit cell and that is very good for analog / digital conversion or logic controlling. Then when it comes to stand alone memory, the bit cell density is high and transistor size is big, it is not feasible to have each single individual bit cell under controlled by each transistor. Then many of the good features of ReRAM is comprimised, like read/write speed, energy consumption, endurance..........
WBT again, has achieved the first one selector/one resistor structure for their memory array development, to capture both small sizing and single bit cell controlling. I think this is so called ground breaking.
Expand
(20min delay)
