adrian_b

3 ms·

1.
by adrian_b 5d ago
Yes, I had the same experience while using various Sun computers between 1995 and 2005. During all that decade the Sun computers were reliable, but excessively slow. Around the beginning of 2006, I had a laptop with a 64-bit AMD Turion CPU
2.
by adrian_b 5d ago
"Whiskey", in the International Radiotelephony Spelling Alphabet (NATO phonetic alphabet), is easier to pronounce than "double u".
3.
by adrian_b 5d ago
While in French there is also a great discrepancy between the written language and the spoken language, when reading French the corresponding pronunciation is almost always very easy to predict. Only the reverse can be difficult, because ma
4.
by adrian_b 6d ago
The guarantees could be provided, if there would be a way to ensure that the compiler uses the same standard math library that is used by the executable program that is created. This could be done, for instance, if the standard math library
5.
by adrian_b 6d ago
About NaNs, you are right, because the standard does not mandate the use of a specific NaN for each kind of invalid operation. Nonetheless, if you use a comparison function for which all NaNs are equivalent, different CPU architectures that
6.
by adrian_b 6d ago
In China and in places influenced by Chinese culture, due to homonymy between "4" and death.
7.
by adrian_b 7d ago
It should be noted that subnormals have appeared for the first time in standards a decade before Intel 8087 (the ancestor of the IEEE standard), in the standards for digital telephony, and that happened in a certain form of FP8. When teleph
8.
by adrian_b 7d ago
All the tested CPUs implement the standard and they implement it in the right way, except for Intel, who has chosen to save some bucks even if this decision might cause unpredictable financial losses for naive customers, who might choose to
9.
by adrian_b 7d ago
Scientific computing can be done only in 2 ways, either with subnormals or by enabling the underflow exception and writing a suitable exception handler for it. If the use of subnormals is disabled with FTZ/DAZ that is guaranteed to gen
10.
by adrian_b 7d ago
On early computers, any underflow generated an exception that would crash the program if not handled. This was very good, because underflows completely break the assumptions about floating-point arithmetic on which numeric algorithms are ba
11.
by adrian_b 7d ago
It is non-deterministic mainly due to compilers that generate different code or when there are concurrent computations that are executed in an unpredictable order. A single sequence of floating-point operations executed on an IEEE 754 compl
12.
by adrian_b 7d ago
This is true only on most hardware that does not care about computational errors, i.e. which is intended mostly for games or for AI.
13.
by adrian_b 7d ago
Which is certain to produce big errors at every instance where a subnormal result would have been generated or propagated otherwise. Such options exist only for games or for experts in numeric analysis, not for general-purpose programming.
14.
by adrian_b 7d ago
Most people who believe that they do not need subnormals do not actually know whether they need or not subnormals and it is very likely that they need subnormals. The options mentioned above are guaranteed to generate big errors, so they
15.
by adrian_b 7d ago
Nonetheless, there is some truth in the headline. When an organ is known to be generated from 2 distinct embryonic preforms, which become merged and integrated at a later stage of development, there is a likelihood that in some distant ance
16.
by adrian_b 8d ago
In my opinion, conceptualizing memory consistency in terms of acquire/release is wrong and it confuses many programmers. One problem is that 2 kinds of ordered loads and stores are not enough, 4 kinds are needed. Some algorithms need n
17.
by adrian_b 8d ago
Free preprint: https://www.biorxiv.org/content/10.1101/2025.07.02.662771v1
18.
by adrian_b 9d ago
Technological progress is bottlenecked by the fact that there are many mathematical problems for which currently there are no known practical methods of solution. Because even with supercomputers the equations that describe many physical sy
19.
by adrian_b 9d ago
To help you with the citations: Garbage collectors: "Recursive Functions of Symbolic Expressions and Their Computation by Machine, Part I", John McCarthy, Communications of the ACM, 1960-04, pp. 184-195 (open access at ACM). Stack
20.
by adrian_b 9d ago
Their CPUs are faster at inference than many GPUs. These are server CPUs with 144 cores and wide memory interface. Fujitsu has added some ISA extensions for AI inference to the standard Arm ISA, e.g. instructions for inference with FP8. Eve
21.
by adrian_b 9d ago
For many years the Fujitsu SPARC implementations were much faster than the Sun SPARC implementations. During the decade 1995-2005, until the 64-bit AMD Opteron servers with Linux offered an alternative that crushed any kind of SPARC CPU wit
22.
by adrian_b 9d ago
No it is not at all Neoverse V2, which is obsolete and has much lower performance. It is a custom Armv9.3-A design (same ISA like the Arm C1 CPUs from the flagship smartphones of 2026), but it has double-width execution units for SVE2 (i.e.
23.
by adrian_b 9d ago
They did a presentation at Hot Chips, with all these details. The core die is made with TSMC-N2P, while the SRAM die and the peripheral die are made with TSMC-N5. Their CPU is designed in Japan, like the AMD CPUs or the Intel CPUs are desig
24.
by adrian_b 9d ago
True, but besides wars there are other dangers, e.g. USA does not declare war when they apply "sanctions" randomly, under the legal theory that they can dictate how anything that includes some piece of US origin may be used by its
25.
by adrian_b 9d ago
They made a detailed presentation at Hot Chips. All the component chips are made at TSMC, in various processes. "Developed in Japan" means the same thing like "developed in USA" means for NVIDIA or AMD or Intel. You can
26.
by adrian_b 9d ago
They did a detailed presentation at Hot Chips. See e.g.: https://chipsandcheese.com/p/hot-chips-2026-fujitsus-monaka-... They have provided far more details than companies like Apple or Qualcomm.
27.
by adrian_b 9d ago
Yes, I do not know what this paragraph wants to say: > "The first memory allocation scheme started with a stack-based memory allocation. Next came the dynamic-based memory allocation scheme where linked-list and bucket-heap mechanis
28.
by adrian_b 9d ago
This is partially mitigated by the fact that all the formats that quantize to 4 bits, or other such low values, partition the weights into small blocks and they also keep scale factors for each block of 4-bit weights. This works well when t
29.
by adrian_b 10d ago
Yes, while for now there is no known method for useful interstellar flight, spaceships with nuclear reactors could reduce the time required for traveling through the Solar System to acceptable values.
30.
by adrian_b 10d ago
Earth is by far the best we can get, but on big enough spaceships it is possible to create living conditions that are better than on any other body of the Solar System, except perhaps on a ship floating in the clouds of Venus. The experimen

More ›