Fast Hash (ASIC) by Helion Technology High Performance SHA and MD5 Hashing Cores

[slippyrocks]

IP Name          Fast Hash (ASIC)
   
   Provider         Helion Technology

   Description         High Performance SHA and MD5 Hashing Cores

   Categories         IP Catalog : Digital Core IP : Security : Decryption
IP Catalog : Digital Core IP : Security : Encryption
IP Catalog : Digital Core IP : Security : Other

   Portability          ASIC

   Process Node          90nm

   Type          Soft IP

   Maturity         Please login or register to view this data

   Overview          The Helion Fast Hash core implements the NIST approved SHA-1, SHA-224, SHA-256, SHA-384 and SHA-512 Secure Hash Algorithms to FIPS 180-3 and the legacy MD5 hash algorithm to RFC 1321. These are high performance cores that are available in either single-mode or multi-mode versions and have been designed specifically for ASIC. The hash algorithms take as input a message of arbitrary length, process the message as a series of 512 or 1024 bit blocks, and produce as output a compressed representation of the message data in the form of a message digest, the length of which varies with hash algorithm. Applications for the hashing cores include implementations of the standard Keyed-Hash Message Authentication Code (HMAC) described in FIPS 198-1. They are commonly used in the IPsec and TLS/SSL protocols, as well as Digital Signature applications, where a hash function is required to ensure both data integrity and origin authentication.

   Features         Implements one or more of SHA-1, SHA-224, SHA-256, SHA-384, SHA-512 & MD5 hash algorithms
Fast operation â€" one clock per hashing algorithm round
Performs automatic message length calculation and padding insertion
HMAC wrapper available for quick and easy implementation
Optional user initialisation of IVs for efficient HMAC support
Optional state unload/reload feature for handling fragmented messages

   Deliverables         Fully synthesisable Verilog RTL
Self-checking testbench with FIPS test vectors
Comprehensive user documentation
Example simulation scripts

   Market Category          Automotive, Communications, Consumer Electronics, Data Processing, Industrial and Medical, Military/Civil Aerospace

   Datasheet         Please login or register to view this data

   Gate Count         Please login or register to view this data

   QIP Rating        This IP is not yet QIP rated.    
        
     Related IP from Helion Technology you may be interested in...

   IP Name         Description
   AES-XTS Fast (Xilinx)         Fast AES-XTS/CBC Core with Ciphertext Stealing
   Fast Hash (ASIC)         High Performance SHA and MD5 Hashing Cores
   ModExp (Xilinx)         Modular Exponentiation Accelerator for RSA and Diffie-Hellman
   Fast Hash (Xilinx)         High Performance SHA and MD5 Hashing Cores
   AES-CCM (Xilinx)         AES-CCM Authenticated Encryption Cores

http://www.chipestimate.com/ip.php?id=24018

http://www.heliontech.com/fast_hash.htm

[1715480677]

1715480677

[1715480677]

1715480677

[1715480677]

1715480677

[1715480677]

1715480677

[shibaji]

90 nm process node now is completely outdated - IMHO

[dani]

Scam.

I'm just saying because I dont want to see the first 3 replies to be "wow im interested" or "I will order pre-order.." blabla

Edit: This does not look like any miner out there, can anyone tell me what exactly this is? 

[shibaji]

Scam.

I'm just saying because I dont want to see the first 3 replies to be "wow im interested" or "I will order pre-order.." blabla

Edit: This does not look like any miner out there, can anyone tell me what exactly this is? 

They are trying to deliver soft IP. Which means, take their HDL code, and go build your own ASIC chip, and that too on a 90nm process node. Others are trying to come up with 28nm process node.

[slippyrocks]

http://www.google.com/search?&q=Multi-Project_Wafer

http://www.google.com/search?&q=mosis

BitSyncom:  Avalon is TSMC 110nm G  @ https://bitcointalk.org/index.php?topic=79637.msg1289344#msg1289344

[||bit]

Scam.

I'm just saying because I dont want to see the first 3 replies to be "wow im interested" or "I will order pre-order.." blabla

Edit: This does not look like any miner out there, can anyone tell me what exactly this is?  

It's not a scam. It's an asic chip that can do SHA256 hashing. I've asked questions about these same chips more than once before. But haven't yet read an answer on why this type of chip hasn't already been designed, by some bitcoin techies in this field, in hybrid fashion to work with FPGA's (or GPU's) to make something that mines better than FPGA's and GPU's. My bet is you could get something five to ten times faster and more efficient than either technology alone, but it wouldn't be quite as fast & efficient as pure asic technology.

[papamoi]

Scam.

I'm just saying because I dont want to see the first 3 replies to be "wow im interested" or "I will order pre-order.." blabla

Edit: This does not look like any miner out there, can anyone tell me what exactly this is?  

It's not a scam. It's an asic chip that can do SHA256 hashing. I've asked questions about these same chips more than once before. But haven't yet read an answer on why this type of chip hasn't already been designed, by some bitcoin techies in this field, in hybrid fashion to work with FPGA's (or GPU's) to make something that mines better than FPGA's and GPU's. My bet is you could get something five to ten times faster and more efficient than either technology alone, but it wouldn't be quite as fast & efficient as pure asic technology.

as per your calculation,what speed we could get with an asic based on this?

[shibaji]

http://www.google.com/search?&q=Multi-Project_Wafer

http://www.google.com/search?&q=mosis

BitSyncom:  Avalon is TSMC 110nm G  @ https://bitcointalk.org/index.php?topic=79637.msg1289344#msg1289344

Yeah, and they have already delivered finished product - not starting with it. Good luck with 90nm.

[aTg]

It is published here is a "library" with you can to design a structured ASIC, there are thousands of libraries for SHA256, but try to put yourself and make a chip that works for mining from that.
This is not really anything.
http://www.actel.com/products/ip/search/detail.aspx?id=627
http://www.ipcores.com/sha_ip_core.htm
https://www.alma-technologies.com/index.php?item=products&pid=30203
http://www.gdatech.com/ip/sc_IP_SHA256.shtml
and more ...

[slippyrocks]

according to Toshiba 90nm "logic densities up to 400,000 gates/mm2 or more and low-k dielectric"
http://www.toshiba.com/taec/Catalog/Line.do?familyid=1&lineid=7231

Helion Fast Hash IP is SHA-512 @ 2400 Mbps using 41500 gates

2400 Mbps / 512 = 4.688 MHashs/s per unit; Bitcoin is 256+256

20mm x 20 mm = 400 mm2 x 400k gates = 160,000k gates / 41.5k gates = 3855 units

3855 units x 4.688 MHashs/s = 18.072 GHashs/s per chip @ 200Mhz

[papamoi]

according to Toshiba 90nm "logic densities up to 400,000 gates/mm2 or more and low-k dielectric"
http://www.toshiba.com/taec/Catalog/Line.do?familyid=1&lineid=7231

Helion Fast Hash IP is SHA-512 @ 2400 Mbps using 41500 gates

2400 Mbps / 512 = 4.688 MHashs/s per unit; Bitcoin is 256+256

20mm x 20 mm = 400 mm2 x 400k gates = 160,000k gates / 41.5k gates = 3855 units

3855 units x 4.688 MHashs/s = 18.072 GHashs/s per chip @ 200Mhz

seriously?

but would it work for mining?
cause usually all this are made for archiving etc

[||bit]

according to Toshiba 90nm "logic densities up to 400,000 gates/mm2 or more and low-k dielectric"
http://www.toshiba.com/taec/Catalog/Line.do?familyid=1&lineid=7231

Helion Fast Hash IP is SHA-512 @ 2400 Mbps using 41500 gates

2400 Mbps / 512 = 4.688 MHashs/s per unit; Bitcoin is 256+256

20mm x 20 mm = 400 mm2 x 400k gates = 160,000k gates / 41.5k gates = 3855 units

3855 units x 4.688 MHashs/s = 18.072 GHashs/s per chip @ 200Mhz

seriously?

but would it work for mining?
cause usually all this are made for archiving etc

Keep in mind that the chip that does the SHA256 hash is only one step in the overall mining algorithm. For each hash, you will have to transmit that data to some operation outside of the asic (those kinds of speeds may not work readily without engineering around some interelemental interactance issues). Also, it will take 'X' number of clock cycles to compare the output to some target value, and if it doesn't meet your target do some kind of data shift on the digest value, then resend that digest back to the hasher. Something like that. So, expect less than 18GH/chip. Finally, what is the cost fo one chip?

All that said, I'd still like to see what can be made with it.

[||bit]

Scam.

I'm just saying because I dont want to see the first 3 replies to be "wow im interested" or "I will order pre-order.." blabla

Edit: This does not look like any miner out there, can anyone tell me what exactly this is?  

It's not a scam. It's an asic chip that can do SHA256 hashing. I've asked questions about these same chips more than once before. But haven't yet read an answer on why this type of chip hasn't already been designed, by some bitcoin techies in this field, in hybrid fashion to work with FPGA's (or GPU's) to make something that mines better than FPGA's and GPU's. My bet is you could get something five to ten times faster and more efficient than either technology alone, but it wouldn't be quite as fast & efficient as pure asic technology.

as per your calculation,what speed we could get with an asic based on this?

I didn't really make a calculation. My guess on orders of speed or efficiency were largely speculation.

If you mean asics in general, consider the BFL Jalapeno, you see 4.5GH advertise (the Jalapeno will have one asic chip). But they likely have done a lot of engineering around the chip to support that speed. Just look at their FPGA boards compared to other sources of FPGA boards. BFL has more extra-component support.