IP Library › Granted Patent US 10,083,034
Granted Patent B1
US 10,083,034 · App. 15/713,526 · Granted Sep 25, 2018

Method and apparatus for prefix decoding acceleration

Inventors: Sudhir K. Satpathy (Hillsboro, OR); Vinodh Gopal (Westborough, MA)
Assignee: Intel Corporation
G06F9/3016G06F9/3004G06F9/30149
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 10,083,034
App. No.
15/713,526
Granted
Sep 25, 2018
Kind
B1
Abstract

In one embodiment, an apparatus comprises a memory, a processor and a prefix decoder engine to access a plurality of code lengths of a header associated with a compressed data block; determine a number of instances of each code length of at least some of the plurality of code lengths; and operate a plurality of decode streams in parallel, a first decode stream of the plurality of decode streams to iterate through a first portion of the plurality of code lengths and determine codes corresponding to the first portion of the plurality of code lengths, a second decode stream of the plurality of decode streams to iterate through a second portion of the plurality of code lengths and determine codes corresponding to the second portion of the plurality of code lengths.

Claims (38)

1. An apparatus comprising:

a memory;

a processor; and

a prefix decoder engine to:

access a plurality of code lengths of a header associated with a compressed data block;

determine a number of instances of each code length of at least some of the plurality of code lengths; and

operate a plurality of decode streams in parallel, a first decode stream of the plurality of decode streams to iterate through a first portion of the plurality of code lengths and determine codes corresponding to the first portion of the plurality of code lengths, a second decode stream of the plurality of decode streams to iterate through a second portion of the plurality of code lengths and determine codes corresponding to the second portion of the plurality of code lengths.

2. The apparatus of claim 1 , wherein the plurality of decode streams comprises a first decode stream that is to iterate in a forward direction through the first portion of the plurality of code lengths and a second decode stream that is to iterate in a reverse direction through the second portion of the plurality of code lengths.

3. The apparatus of claim 1 , wherein the plurality of decode streams comprises a first decode stream that is to iterate in a forward direction through the first portion of the plurality of code lengths and a second decode stream that is to iterate in a forward direction through the second portion of the plurality of code lengths.

4. The apparatus of claim 1 , wherein the plurality of decode streams comprise three or more decode streams.

5. The apparatus of claim 1 , wherein the prefix decoder engine is to:

determine that a plurality of consecutive code lengths of the header are the same value; and

simultaneously write the code lengths to a plurality of wordlines of the memory.

6. The apparatus of claim 5 , wherein the prefix decoder engine is to provide an address and a corresponding tag to the memory to enable the memory to determine a plurality of wordlines to be enabled during a single write operation to write the plurality of consecutive code lengths to the plurality of wordlines of the memory.

7. The apparatus of claim 1 , wherein the prefix decoder engine is to disable a bank of the memory based on a determination that a plurality of consecutive code lengths of the header each have a value of zero.

8. The apparatus of claim 1 , further comprising a plurality of decompression engines each operable to decompress a payload based on codes decoded by the prefix decoder engine.

9. The apparatus of claim 1 , wherein the codes are Huffman codes used to encode a DEFLATE payload.

10. The apparatus of claim 1 , further comprising a battery communicatively coupled to a processor, a display communicatively coupled to the processor, or a network interface communicatively coupled to the processor.

11. A method comprising:

accessing a plurality of code lengths of a header associated with a compressed data block;

determining a number of instances of each code length of at least some of the plurality of code lengths; and

operating a plurality of decode streams in parallel, a first decode stream of the plurality of decode streams to iterate through a first portion of the plurality of code lengths and determine codes corresponding to the first portion of the plurality of code lengths, a second decode stream of the plurality of decode streams to iterate through a second portion of the plurality of code lengths and determine codes corresponding to the second portion of the plurality of code lengths.

12. The method of claim 11 , wherein the plurality of decode streams comprises a first decode stream that is to iterate in a forward direction through the first portion of the plurality of code lengths and a second decode stream that is to iterate in a reverse direction through the second portion of the plurality of code lengths.

13. The method of claim 11 , wherein the plurality of decode streams comprises a first decode stream that is to iterate in a forward direction through the first portion of the plurality of code lengths and a second decode stream that is to iterate in a forward direction through the second portion of the plurality of code lengths.

14. The method of claim 11 , further comprising:

determining that a plurality of consecutive code lengths of the header are the same value; and

simultaneously writing the code lengths to a plurality of wordlines of a memory.

15. The method of claim 11 , further comprising disabling a bank of a memory based on a determination that a plurality of consecutive code lengths of the header each have a value of zero.

16. At least one machine readable storage medium having instructions stored thereon, the instructions when executed by a machine to cause the machine to:

access a plurality of code lengths of a header associated with a compressed data block;

determine a number of instances of each code length of at least some of the plurality of code lengths; and

operate a plurality of decode streams in parallel, a first decode stream of the plurality of decode streams to iterate through a first portion of the plurality of code lengths and determine codes corresponding to the first portion of the plurality of code lengths, a second decode stream of the plurality of decode streams to iterate through a second portion of the plurality of code lengths and determine codes corresponding to the second portion of the plurality of code lengths.

17. The at least one medium of claim 16 , wherein the plurality of decode streams comprises a first decode stream that is to iterate in a forward direction through the first portion of the plurality of code lengths and a second decode stream that is to iterate in a reverse direction through the second portion of the plurality of code lengths.

18. The at least one medium of claim 16 , wherein the plurality of decode streams comprises a first decode stream that is to iterate in a forward direction through the first portion of the plurality of code lengths and a second decode stream that is to iterate in a forward direction through the second portion of the plurality of code lengths.

19. The at least one medium of claim 16 , the instructions when executed to cause the machine to:

determine that a plurality of consecutive code lengths of the header are the same value; and

simultaneously write the code lengths to a plurality of wordlines of a memory.

20. The at least one medium of claim 16 , the instructions when executed to cause the machine to disable a bank of a memory based on a determination that a plurality of consecutive code lengths of the header each have a value of zero.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Sep 22, 2017
From: SATPATHY, SUDHIR K.; GOPAL, VINODH
To: INTEL CORPORATION
Reel/Frame 043978/0571 →
Cited By (3)
US 12,438,557 US 12,620,050 US 12,694,571