(Fig.1) Quantum mechanical spin hampers and contradicts spintronics, MRAM.
Magnetic (or magnetoresistive ) random access memory (= MRAM ) is a type of non-volatile memory storing data as two ferromagnetic materials' magnetizing direction.
This MRAM uses giant magnetoresistance (= disagreeing with quantum mechanics ) where two parallel (or antiparallel ) magnetizations in two adjacent ferromagnets causing low (high) electric resistance is treated as bit "0" (or "1", this Fig.1 ).
Electron spin (or this ) is known to be unreal, so these magnetizations are caused by electron's orbits, Not by spins.
Overhyped media often baselessly claim MRAM will replace today's mainstream DRAM, SRAM memory (= storing data as classical electric voltage without using quantum mechanical spin ) someday, like the exaggerated graphene and deadend overhyped quantum computers.
But actually, despite extremely-long time researches (= 1996~ ), MRAM (= spin-transfer-torque or STT-MRAM is today's standard MRAM ) remains in niche market forever (= cannot replace the mainstream DRAM, SRAM ), hindered by the unphysical quantum mechanical spin model.
This 6th-paragraph says
"While MRAM pioneers such as Everspin... it remains a niche memory (= forever )"
Because MRAM is energy-inefficient, slower, lower-endurance. more error-prone than DRAM.
The 2nd paragraph of this site says
"There are still plenty of skeptics when it comes to MRAM,.. That has limited MRAM to a niche role over the past couple decades, hampered by high costs, low density, and lower endurance" ← MRAM is far inferior to SRAM without spin.
This 2nd-paragraph says
"However, challenges facing the MRAM market include high manufacturing costs, limited scalability"
This 2nd-paragraph says
"MRAM has Not been able to make much progress outside specialist low volume markets" ← MRAM or (fictitious) spintronics has been deadend, remaining in niche market despite longtime researches.
This 14th-paragraph says
"Technologies like ReRAM, MRAM, etc. will remain niche embedded market alternatives because of this restriction".
This-1. says
"One of the biggest challenges with MRAM technology is making the write error rate (WER) as low as possible"
This 2nd-last paragraph says
"On the one hand, MRAM isn't the Holy Grail of memories because it lacks the raw capacity of DRAM."
↑ So contrary to the overhyped media, the spintronics or MRAM is already deadend with No hope of progress.
(Fig.2) Spin-torque, MRAM, spintronics cannot be predicted by quantum mechanics. Electron spin is unreal.
The present mainstream MRAM is spin-transfer-torque (= STT)-MRAM as seen in Everspin.
This STT-MRAM's mechanism can be explained by giant magnetoresistance (= GMR ) that disagreed with quantum mechanics and unphysical spin.
As shown in the upper figure-①, when electrons' current ( consisting of orbits, Not the unphysical spin ) flowing through the fixed (= thick ) magnetization layer (= magnetization or electron's orbit in the upper direction ) tends to have the same magnetization as the fixed layer, because the electrons' orbits in the opposite magnetization are rebounded by the fixed layer due to increased resistance in magnetoresistance.
As a result, in the case of upper figure-①, the free (= thin ) layer in the right side comes to have the same magnetization as the fixed layer (= which is bit "0" state ) dominated by the electrons' orbits flowing through fixed layer (= having the same magnetization as the fixed layer ) ( this 12~13th-paragraphs, this p.7, this p.41 ).
When electrons' current flows from the right side as shown in the upper figure-② electrons' orbits whose magnetization is opposite to the fixed layer tend to rebound off the fixed layer back to the free layer (= due to magnetoresistance ), and the free layer's magnetization becomes opposite to the fixed layer (= bit "1" state, this p.23 ) occupied by the electrons' orbits rebounded by the fixed layer.
Quantum mechanical (unphysical) spin model just shows abstract artificial model or equation with freely-adjustable spin-torque, damping parameters lacking real particle picture ( this p.3-left-(1)~p.3-right, this p.2-lower~p.3 ). ← No quantum mechanical prediction of spin-torque-MRAM.
Spin-transfer-torque-MRAM remains a deadend niche product inferior to today's mainstream DRAM, SRAM (= without spin ) forever due to this unphysical quantum mechanical model lacking real particle picture.
This p.1-right-1st-paragraph says " This renders STT-MRAM unsuitable for ultrafast applications such as cache memories. Moreover, the single path for both reading and writing makes it challenging to attain reliable reading operations."
STT-MRAM also has the reliability issues or high failure rate ( this p.5-right-2 challenge ).
This p.1-left-2nd-paragraph says
"Despite various advantageous features, STT-MRAM is
facing various reliability challenges including write failure,
decision failure, retention failure and failures due to read
disturb..
a write failure occurs when the bit-cell does Not flip
to its required value during the given write period. This can
happen, since the write process in STT-MRAM is of stochastic (= random, erroneous switch ) nature."
This 1st-paragraph says
"Magneto-Resistive Random Access Memory (MRAM) has so far generally failed to replace SRAM (= using electric current transistors instead of the unphysical spin ) because its Spin Transfer Torque (STT-MRAM) implementation is too slow and doesn't last long enough."
This p.1-introduction-1st-paragraph says
"STT-MRAM faces various challenges along with
its merits such as, the reliability of a tunnel barrier, long write latency and small energy efficiency due to still high
write current."
This-p.4-last~p.5 says
" STT-RAM has Not yet reached sufficient maturity to be introduced in
mass-produced devices. The critical current, required for the switching, is still too high for real-world applications." ← STT-MRAM is energy-inefficient, which is why it cannot be mainstream.
In fact, quantum mechanics, which is unable to predict the mechanism of (spin-transfer-torque)-MRAM, has to artificially fit free parameters of unphysical spin-torque model or equation ( this p.14 ) to experiment ( this p.5-lower, this p.3-TableI, this p.10-right, this p.3-left-last-paragraph ).
This p.2-right-last~p.3 says "We used equation (2) to fit the experimental results by choosing βST, βFT, resonance angular frequency ω0 and spectrum linewidth ∆ as fitting parameters" ← No quantum mechanical prediction of STT-MRAM
This-introduction-1st-paragraph says "A popular method for the description of STT in magnetic multilayers is the model of Slonczewski....
However, it introduces free variables that depend on various system parameters... These parameters are usually determined in a phenomenological fashion "
This p.18-last-paragraph says "Note that some parameters are to date Not known fully. In the term aj an unknown polarization function appears. Furthermore the Gilbert damping term is not exactly known." ← No quantum mechanical prediction of STT-MRAM depending on artificially-chosen damping parameter.
This p.17-last says "No consensus (theoretically and experimentally) over the ratio β/α, which can vary between 0.1 and 10 (= free parameters )"
This p.3-right-last-paragraph says "The effective damping (αeff) of the TaS2/Py bilayer is evaluated by fitting the μ0ΔH versus f data (= parameters adjusted by experiment Not by quantum mechanical theory )."
This p.24-left-3rd~4th-paragraph says "Both show that the agreement between theory and experiment is still incomplete (= wrong prediction )"
This paper ↓
p.39(or p.28)-last-paragraph says "a fit can be obtained only by leaving numerous free parameters: "
p.121(or p.110)-2nd-paragraph says "where B and C are free parameters"
As a result, there is No evidence of spin in spin-torque-MRAM that cannot be predicted by quantum mechanics relying on various free parameters.
(Fig.3) Overhyped spin-orbit-torque (= SOT )-MRAM is far from practical.
Today's MRAM or STT-MRAM is already deadend, stuck in niche market forever.
The latest version of MRAM is called spin-orbit-torque (= SOT )-MRAM.
SOT-MRAM is said to utilize spin-Hall effect to provide the magnetically-polarized electrons' current (= electric current splits into the upper and lower directions depending on its magnetization direction ) to each memory ( this 2nd-paragraph, this 2nd-paragraph ). ← This mechanism is impractical.
But electron's spin is unreal, and relativistic spin-orbit effect allegedly used in spin-Hall effect is paradoxical, wrong.
This spin-Hall effect can be naturally explained by realistic electron's orbit veering into the upper or lower directions due to classical Magnus effect (= friction between the orbiting electrons and their surroundings ).
Despite longtime researches, this "(overhyped) promising" spin-orbit-torque (= SOT )-MRAM is still useless forever ( this-p.11-left-last-paragraph ).
This-last-paragraph says
"it needs high currents for write operations, so its dynamic power consumption is still quite high (= energy-inefficient ). Furthermore, SOT-SRAM cells are still larger than SRAM (= today's mainstream non-spin memory ) cells, and they are harder to make. As a result,.. it is unlikely that it will replace SRAM" ← spintronics MRAM is hopeless.
This p.1-right-last~p.2-left-upper says
"However, the development of
SOT-MRAM for ultrafast applications still remains challenging...
which is an obstacle for practical SOT-MRAM
applications...
Despite some progress, a clear field-free solution
for ultrafast SOT operations remains elusive for real applications"
This introduction-1st-paragraph says
"However, a challenge for SOT-MRAM is that each bit cell requires two access transistors, resulting in a larger unit area (= miniaturization is impossible ), limiting its application in high-density memory scenarios."
This-p.4-lower says
"Lowering the energy demand and enhancing the energy efficiency is an outstanding problem for the SOT-MRAM.
Today, though, the biggest problem with SOT-MRAM is that it only switches about 50% of the time (= too many errors to be practical )."
This 2nd-paragraph says
"an in-plane bias magnetic field (= additional device generating magnetic field is needed for SOT-MRAM ) which has hindered progress in developing practical SOT-MRAM devices ( this-introduction-1st~2nd-paragraphs ). "
The latest research paper on spin-orbit-torque or SOT-MRAM ↓
p.1-rigt says ". This renders STT-MRAM unsuitable for ultrafast applications such as cache memories. Moreover, the single path for both reading and writing makes it challenging to attain reliable reading operations." ← No improvement in the problematic STT(= spin-transfer-torque )-MRAM
"However, the development of SOT-MRAM for ultrafast applications still remains challenging (= both STT and SOT-MRAM are still useless, problematic )"
p.2-Fig.1a shows even this latest SOT-MRAM switch is very bulky, as large as micrometer (= μm ), which is far bigger and bulkier than today's compact DRAM memory of only 20 nanometers.
p.3-Fig.2 shows switching probability (= PSW ) of this SOT-switch is bad and uncertain between 80 ~ 100% (= about 10% error rate ), which is too error-prone to be a practical switch.
p.5-right just artificially chose free parameters without quantum mechanical prediction nor calculation ( this p.2-STT-SOT model, p.3-Table.1,p.7-methods also used input free parameters without quantum mechanical prediction ).
Actually, in this-p.3 6th-paragraph, reviewer #2 says
"On the other hand, the study focuses on micro-sized magnetic dots, and it is well-documented that
reversal mechanisms and energies are very different in µm compared to sub-100 nm dots.
Meanwhile, practical applications are clearly projected for sub-100 nm dimensions (= so this device of micro-meter size is too big to be practical )"
"..In Figure 2, the switching probabilities hardly converge to 100%, but they are not zero either. I raise some doubts about the switching reliability (= this latest SOT-MRAM switch is too erroneous )"
This research paper on SOT ↓
p.2-Fig.1c shows this SOT-switch is too bulky and big = about 10μm (= scale bar )
p.4-Fig.3-D shows this SOT-switch is too slow, one switch took about 3~5 seconds, and stochastic, unreliable ( this p.5-Fig.4-d ).
As a result, spin-orbit-torque or SOT-MRAM is too error-prone, too slow, too bulky to be practical, contrary to the media hype.
Quantum mechanics has to rely on unrealistic negative kinetic energy in tunnel current over extremely narrow space (= ~ nm ).
↑ In such a narrow space, their estimation of the potential barrier is wrong, and electrons can realistically flow over very short distance by thermal fluctuation or de Broglie wave interference pressure (= de Broglie wave has 'power' to push electrons and cause interference pattern ) even without the impossible negative kinetic energy (= so tunneling is Not the occult quantum mechanical phenomenon but realistic electron's current with positive kinetic energy by thermal fluctuation and de Broglie wave interference ).
Feel free to link to this site.
