Visual NAND Reconstructor 4.0 with GPU ECC correction!
Dear users,
It’s been a while since the last update, we have been very busy trying to eliminate one of the bottlenecks of the chip-off data recovery process. For many years all data recovery engineers around the globe have struggled against the same issue – ECC correction was getting slower and slower due to capacity increase and code complication. It was literally the nightmare to correct ECC on 128GB device and larger. Even with top-notch CPUs. Now it’s in the past.
We have developed the breakthrough technology of ECC correction on GPU! Increase of performance, or better to say, reduction of time required for correction, is almost limitless now! The more powerful your graphic card is the faster ECC correction will be completed.
Here’s the full list of new features and resources:
eMMC-NAND RECONSTRUCTOR
The new chips have been added to database
-
KLMAG2GE2A
-
H26M54003EMR
-
H26M64003DQR
-
H26M78002BFR
-
H9TQ64AAETMC
-
H26M42003GMRA
-
H26M42001FMR
-
H9DP4GK4JJBC
-
H9DP4GG4JJAC
-
H26M21001ECR
-
THGBMAG6A2JBAAR
VNR - GPU ECC CORRECTION
We have developed a well optimized GPU ECC/BCH decoding algorithm with almost linear time decrease required for correction depending on performance of graphic card. In some cases with adequately powerful GPUs it is possible to achieve tenfold performance increase. The feature works on most of the codes and all common GPU vendors are supported. Even integrated GPUs may provide 1.5x performance increase comparing to CPU computing.
Here’s results of our tests calculated on 4GB dumps from different controllers and ECC sizes:
Following data and table shows precise times of correction with some impressive results.
Tested devices/codes
ECC correction time (per 4GB dump)
For the best performance we recommend to use 16/32 Gb of RAM and NVMe SSD.
Usage of the same graphic card for both video output and ECC computation may cause slowing down and artifacts. The best practice is to use integrated card as video output while discrete card is used for ECC computation only.
The features is officially supported and tested on Windows 10 x64 only.
Warning! Make sure that cooling system of your graphic card is adequate and ready for 99% load for hours.
We are not responsible for any damages or consequences caused by hardware overheating while using VNR.
We keep improving our software and hardware, and researching new usages of GPU for future updates!
New NAND chips
-
AD3A15A744 Noname
-
ADB18055AD H27S1G6F2CKA-BM
-
AD5A14F300 Noname
-
AD3A18A664 monoUFD
-
8964643CA1 TK29F32B08MCMFS
-
983C98B376 HVKK0FG
-
454CA9927A microSD
-
454A9AB27E SDYNMMDHSP-016G
-
983E98B376 MonoUFD
-
98C7988276 TH15G3GAFA
-
89D40C32AA 29F0IT2ALC0HI
-
45499BB37E UFD monolith
-
98DE849372 FA32B08UCT1-7F
-
45DE95327A SDZNMMDHER-032G
-
45C3943272 microSD
-
45C585BAF4 mSD
-
45C384BAF0 MicroSD
-
2CBC906654 Noname
New ECC/BCH
-
A2F500M3G FPGA_3638_8.bch
-
AllWinnerA10_8936(ecc84)_8.bch
-
ATJ_8936(ecc70b)_8.bch
-
ATM7021A_17664(ecc70b)_16.bch
-
AU89102DF_18592(ecc124b)_16.bch
-
D2W753-22_8630(ecc51+SAecc21)_9.bch
-
IS902_8704(ecc48+SAecc12)_9.bch
-
Noname_18048(ecc62b)_16.bch
-
No_name_18591_16.bch
-
PS_MicroSD_18592(ecc124b)_16.bch
-
PS_MicroSD_9216(ecc106b)_8.bch
-
PS_MicroSD_9216(ecc119b)_8.bch
-
SM3259_18432(116b)_17.bch
-
SM3281L_18302_18.bch
-
SM3281L_18592(ecc126b)_17.bch
-
SM_mSD_9162(ecc118b)_8+1.bch
-
SSS6131_18432(ecc112b)_16_X.bch
-
TC58TEG6DDK_SD_17664(ecc76b)_16.bch
-
WM8505_4320(ecc13)_9.bch
New scramblers (XOR keys)
-
AllWinner_A10(8936b_128p_xoredSA_ECC)_CF7ED4.xor
-
ATM7021A(16k_256p)_63FD56.xor
-
AU6989SNHL(18256b_256p_ecc112b_SA_xor).xor
-
AU89102DF(18592b_256p_SAxor_ecc124).xor
-
IS902_v5(8k_1p_xored_SA)_713EEC.xor
-
noName(16k_512p)_F4361E.xor
-
Noname(16k_64p)_8855F9.xor
-
Noname(18432b_258p_SA_ECC)_A05F53.xor
-
PS2251-07-V(16k_86p)_FA9E80.xor
-
PS_MicroSD(18592b_256p_124ecc_xoredSA_ECC)_5061FD.xor
-
PS_MicroSD(9216b_256p_106ecc_xoredSA_ECC)_5061FD.xor
-
PS_MicroSD(9216b_258p_SA_ECC)_5061FD.xor
-
PS_microsd(9216b_258p_SA_ECC)_F71636.xor
-
Sandisk(9216b_128p_ecc230b_xoredSA)_C634C7.xor
-
SM3259(16k_384p)_E8018F_v2.xor
-
SM3259(16k_384p)_E8018F_v3.xor
-
SMnoname(4k_256p)_4C912A.xor
-
SSS6695(8832b_256p_1098_xoredSA_ECC)_809E9C.xor
-
SSS6696(8832b_256p_1098_xoredSA_ECC)_809E9C.xor
-
TC58TEG6DDK(17664b_256p_xored_SA).xor
New monolithic chip pinouts
1 Comment